WP WebRTCForBusinessPeople Kandy 1114

WebRTC
for Business People
Tsahi Levent-Levi
tsahil@bloggeek.me
Tsahi Levent-Levi
WebRTC for Business People
Sponsored by
Table of Contents
Table of Contents................................................................................................................................................... 1
List of Figures ......................................................................................................................................................... 4
List of Tables .......................................................................................................................................................... 4
Introduction........................................................................................................................................................... 5
Who Am I Anyway? ...................................................................................................................................................5
Why WebRTC for Business People? ...........................................................................................................................5
Why is this Report Open? ..........................................................................................................................................6
What is WebRTC? .................................................................................................................................................. 7
What WebRTC? .........................................................................................................................................................8
Free ...........................................................................................................................................................................9
VoIP Developers ......................................................................................................................................................10
ORTC and WebRTC ..................................................................................................................................................11
WebRTCs Job to be Done .................................................................................................................................... 12
Barriers of Entry for New Vendors ..........................................................................................................................12
Reducing End User Friction .....................................................................................................................................13
The Innovators Dilemma and WebRTC ..................................................................................................................13
Browser Support .................................................................................................................................................. 15
Interoperability between Chrome and Firefox ........................................................................................................17
Handling Safari and IE.............................................................................................................................................18
Dealing with Mobile ................................................................................................................................................19
WebRTC on Devices.................................................................................................................................................21
WebRTC API Components .................................................................................................................................... 22
GetUserMedia .........................................................................................................................................................22
PeerConnection .......................................................................................................................................................22
DataChannel ...........................................................................................................................................................23
Networking .......................................................................................................................................................... 24
P2P ..........................................................................................................................................................................25
Signaling .................................................................................................................................................................27
Tsahi Levent-Levi
Sponsored by
Firewall and NAT Traversal .....................................................................................................................................28

Security ...................................................................................................................................................................30
Requested Features ............................................................................................................................................. 31
Multipoint ...............................................................................................................................................................31
Recording ................................................................................................................................................................32
Interoperability .......................................................................................................................................................33
WebRTC vs. the World ......................................................................................................................................... 34
Anatomy of VoIP Solutions......................................................................................................................................34
H.323 and SIP ..........................................................................................................................................................39
Flash ........................................................................................................................................................................41
"Proprietary" ...........................................................................................................................................................41
Vidyo .......................................................................................................................................................................42
WebRTC Hype ...................................................................................................................................................... 43
The WebRTC Ecosystem ....................................................................................................................................... 46
The 5 Uses of WebRTC ............................................................................................................................................46
WebRTC Use Cases by Verticals ........................................................................................................................... 55
Tooling ....................................................................................................................................................................55
Video Conferencing .................................................................................................................................................59
Telecommunications ...............................................................................................................................................63
Customer Service & Support ...................................................................................................................................67
Expert Marketplace .................................................................................................................................................72
Telehealth ...............................................................................................................................................................75
Gaming....................................................................................................................................................................79
Social Networking and Consumer VoIP ...................................................................................................................84
Streaming and Content Delivery .............................................................................................................................87
Other .......................................................................................................................................................................91
The Missing Verticals ..............................................................................................................................................95
Recommendations ............................................................................................................................................... 96
Appendix A: Online Resources ............................................................................................................................. 98
Blogs .......................................................................................................................................................................98
Tsahi Levent-Levi
Sponsored by
Communities .........................................................................................................................................................100
Appendix B: Choosing a WebRTC API Platform report ....................................................................................... 101
Appendix C: Finding out More ........................................................................................................................... 102
Appendix D: About Kandy .................................................................................................................................. 103
Appendix E: Kandy Usage Scenarios ................................................................................................................... 104
Tsahi Levent-Levi
Sponsored by
List of Figures
Figure 1: The innovators dilemma and WebRTC ......................................................... 14
Figure 1: Major milestones of WebRTC support in browsers ........................................ 16
Figure 2: Browser downloads on Android...................................................................... 19
Figure 3: Chat session between Web browsers with and without WebRTC .................. 25
Figure 4: WebRTC media traffic with and without a TURN server ................................ 26
Figure 5: Anatomy of VoIP - the 4 core components of any VoIP solution .................... 34
Figure 6: Developers interest in WebRTC over time ..................................................... 44
Figure 7: Social network interest in WebRTC over time ................................................ 45
Figure 8: Types of vendors in the WebRTC ecosystem ................................................ 46
Figure 9: Growth pattern in web APIs............................................................................ 53
Figure 10: The WebRTC value chain and the 2nd market .............................................. 54
List of Tables
Table 1: Browsers support for WebRTC........................................................................ 16
Table 2: Comparison of WebRTC to other prominent VoIP protocols (SIP and H.323) 39
Tsahi Levent-Levi
Sponsored by
Introduction
Who Am I Anyway?
My name is Tsahi Levent-Levi. I am a developer at heart. I have been working in the
telecom and VoIP/UC industries of the past 15 years (and counting) in various roles: from
a developer, to project manager, product manager and CTO. Most of that time, I was
dealing with signaling and media products that were licensed to other developers who
built their own products with them. This gives me a broad view of the market and an
understanding of the challenges and opportunities that exist in the domain of VoIP.
I came across WebRTC when it was first announced by Google and saw the potential in
it. Since then, I have been watching the WebRTC space closely, consulting about it and
writing about it on my blog: BlogGeek.me. From a hobby it became a "profession".
Why WebRTC for Business People?

WebRTC is going to be transformative to the way we think about digital communication,
and there are already a lot of resources out there. But most of the information is geared
towards developers with very little provided to inform the entrepreneur or the product
manager on what can be done with WebRTC.
The fact that WebRTC is evolving with every passing day, coupled with the FUD (Fear,
Uncertainty and Doubt) that is spread by incumbent players makes it hard to make
business decisions related to WebRTC.
This research paper comes to serve this purpose of explaining what WebRTC is, what it
can and can't do, outlining the ecosystem around WebRTC and provide a healthy dose
of vendors who have already taken the plunge with WebRTC and have built a business
around it. At the end of each chapter, you will find links for extra reading material some
from my own writing and some from bright minds from around the web. At times, the same
link will appear in more than a single chapter.
Tsahi Levent-Levi
Sponsored by
Why is this Report Open?

Last year I took the time to write and publish this report. At the time, I decided to place it
as a paid report on my website.
People purchased it and the feedback I received was good, but many stated it was too
pricy for them. I have been contemplating this problem ever since. On one hand, this
report contains valuable information, so Id like to get paid for the work done. On the other
hand, I really believe in the transformative nature of WebRTC and want to get the
concepts it contains to as large an audience as I can.
Which led me to Kandy, who were kind enough to sponsor it so that anyone can pick it
up and read it freely.
Since it has been over a year from the initial publication of the report, things had to change
and updated a bit both to fit the changes in the market as well as to fit better my own
experiences and understanding of WebRTC. This report that you are now reading has
been updated, and hopefully it is even better than the original one.
Id like to thank the Kandy team for sponsoring this report.
Tsahi Levent-Levi
Sponsored by
What is WebRTC?
WebRTC stands for Web Real-Time Communication. In essence, it is the fusion of two
separate branches in technology: VoIP and the web.
VoIP, short for Voice over IP, is a set of technologies and techniques that enable sending
media (usually voice and video) over an internet connection. Up until the introduction of
WebRTC, VoIP lived within its own ecosystem silo, next to the booming Internet we are
accessing daily via web browsers.
WebRTC comes to connect VoIP into our browsers, and by that, into websites and mobile
apps. It does that by offering a thin layer of Javascript APIs that are implemented by
modern web browsers, and are part of the HTML5 specification.
This means that now every web developer can add real time communication
capabilities to his website or web application.
Further Reading: 001 bloggeek.me/ref/biz001/
The history of WebRTC inside Google [Quora]

Introduction to WebRTC [VisionMobile]
The hurdles involved in using WebRTC
Two important aspects of WebRTC:
It is free
It changes the definition of VoIP developers
Together, they lower the barrier of entry for communication services to a level that enables
use cases that were always required but never fulfilled by previous technologies.
Tsahi Levent-Levi
Sponsored by
What WebRTC?
The term WebRTC in itself is a bit misleading, as it refers to two separate things at the
same time:
1. WebRTC the specification
2. WebRTC the open source project
Both notions for WebRTC are used interchangeably most of the time and it is important
you know the distinction between the two.
1. WebRTC the Specification

WebRTC is a specification that is being standardized at the W3C and the IETF.
The specification deals with:
What goes on the wire what is sent through the network, so that two WebRTCcapable applications can interact with each other
What are the APIs that utilize the functionality itself the defined Java Script API
that is located on top of WebRTC
The standardization of the first version of WebRTC hasnt finalized yet, which detracts
incumbents and telecom vendors from adopting it. That said, there are still hundreds of
vendors who are using WebRTC today, ignoring the transient state WebRTC is in today.
2. WebRTC the Open Source Project

WebRTC is also an open source project. When Google first announced WebRTC, it was
just an open source project located in its own website.
Google made it clear their intention was to standardize WebRTC and embed it into
browsers. To that end, they donated the code to the community under a very permissive
open source license.
As an open source project, WebRTC is quite a powerful piece of code that anyone can
adopt and use in any way he sees fit with or without relation to the fact that there is a
WebRTC specification.
Ericsson recently announced its own WebRTC open source project called openWebRTC.
How popular this project will be is yet to be seen.
Tsahi Levent-Levi
Sponsored by
Free
WebRTC is free in every possible aspect.
1. It is packaged as an open source project and licensed under the permissive BSD
license. Past implementations of similar technologies were either provided under
a proprietary license that had to be purchased or under more restrictive open
source licenses which wasn't palatable for a lot of vendors.
2. It supports and promotes free codecs which don't require any royalty payments
for patent licenses. Codecs usually require multiple payments to use, starting
from a license fee of a vendor and royalties based on quantities sold. This poses
a financial barrier for adoption as well as a hassle when each quarter licensing
fees needs to be recalculated and paid for.
WebRTC's Job to be done [UC Strategies]

Codec wars in WebRTC around the mandatory codec
Is there a future for optional codecs in WebRTC?
The future video codec: H.265 versus VP9
This allows developers to repurpose WebRTC and its components in every possible
mean:
Porting it to operating systems and environments where similar media processing

capabilities are necessary
Using it in large scale deployments where the number of endpoints is considerable
and the payment from each customer is low (most Over-The-Top solutions today)
Adopting smaller modules out of WebRTC and using them alone. This is true
especially to WebRTC's echo canceller and codec implementations
Later chapters in the area of use cases will show how different vendors and entrepreneurs
are wielding the power of WebRTC.
Tsahi Levent-Levi
Sponsored by
VoIP Developers
Developing VoIP capabilities is far from easy. It requires engineers in specific disciplines.
VoIP projects are usually built by teams with 10 developers or more, dealing with aspects
of codec integration, media engine development and fine tuning, operating system porting
and interoperability testing. While some of these capabilities can always be outsourced
or licensed, there are costs and time involved.
WebRTC is not just a specification. It was created from company acquisitions that Google
has made recently. As such, it isn't just a "pet project" of a couple of developers but a
commercial grade package.
With that in mind, you can see how WebRTC enables more developers to use VoIP
related capabilities, but it doesn't stop there. The most important aspect of it is that
addition of Javascript APIs and integration of WebRTC into the browser.
This changes the game for communication developers considerably:
It changes the type of communication developers from C/C++ and Java developers
to web and Javascript developers. This opens up the field to a considerably larger
audience of developers, but also changes the mindset of these developers. You
can think of it as a transition from waterfall development to agile development
techniques
It makes the browser the endpoint or engine that gets used. This reduces the
hassles of issues such as security patches, upgrades and interoperability
nonfunctional aspects of development that takes time and money to deal with
properly
The change in developers also means a change in focus from the need to focus on how
to implement a solution with high media quality, to a focus on the service being developed
and the user experience it provides.
10
WebRTC is for web developers and not VoIP developers [NoJitter]
Tsahi Levent-Levi
Sponsored by
ORTC and WebRTC

You might have heard of CU-RTC-Web or about ORTC.
CU-RTC-Web was a competing specification proposed by Microsoft. It was ignored in
many ways by the IETF during the standardization work done on WebRTC. It later on
merged with another proposal known as ORTC Object RTC.
There is not enough clarity of the market as to the difference between the two: ORTC and
WebRTC.
ORTC is a proposal of a new API faade for WebRTC. One that provides lower level
capabilities for developers, enhancing the types of use cases that can be covered with
WebRTC. It is also an API faade that lets go of the notion of SDP and Offer-Answer
procedures that are used in WebRTC today and are part of the legacy baggage WebRTC
brings with it due to its VoIP related origin.
ORTC is aimed towards making it easier for web developers to use WebRTC and
manipulate the media streams. It is planned as an addition to WebRTC and not a
competing specification.
It is assumed that once ORTC is finalized, existing browsers supporting WebRTC will add
ORTC support.
ORTC is handled by the IETF in its own ORTC Community Group, where the lead is taken
today by Microsoft and Google along with 70 other vendors and individuals.
For planning purposes, you can view ORTC as WebRTC 1.1 an evolution of the
WebRTC specification.
11
Tsahi Levent-Levi
Sponsored by
WebRTCs Job to be Done

If we look at WebRTC from Clayton Christensens Job to be Done theory in book The
Innovators Solution, then what we are aiming for is to think less about market segments
and more about the jobs customers want to do.
In the case of WebRTC, the customers can be viewed as developers and services who
need real-time communication capabilities. For them, the job to be done can be split into
two aspects:
1. Reducing their costs their barrier of entry
2. Reducing the friction for end users using their communication capabilities
Barriers of Entry for New Vendors

As an open source project with a permissible license, WebRTC offers a huge head-start
for developers. It brings them immediate value of around 50,000-200,000 USD the cost
of a commercial media engine.
Coupled with availability through browsers on multiple operating systems and across
mobile devices1, this brings the development and testing costs down to a point where it
makes sense to start using it.
This reduction in barrier of entry brings with it 3 distinct changes in the market:
1. More vendors and developers are looking at using real time communications
2. Use cases that had no reason economically to exist before, now make business
sense with an ROI behind them
3. New business models are being experimented with
While the WebRTC open source project may not offer a nicely packaged mobile SDK, the code itself is
almost fully ready for both Android and iOS use already, with noticeable improvements with each new
release.
12
Tsahi Levent-Levi
Sponsored by
Reducing End User Friction

As a browser capability, WebRTC enables the web to be much more interactive and
capable. Digital storefronts and support services can now include the ability to
communicate and dial right from within the website. This reduces the friction with the
end users, and increases their chance of using the service more.
There are many different ways in which vendors are exploiting this capability:
Adding and acting from context coupled with the request to communicate right from
the website
Enabling users to stay in the website without going for their phone to dial for service
Adding capabilities to social interactions from within the same service, not losing
the users to other services such as Skype
Empowering contact center agents to work from homes and caf shops without
the need for any additional installations or setup
This reduction in friction has the potential of increasing the value of the service employing
WebRTC.
The Innovators Dilemma and WebRTC

WebRTC fits into the classic Innovators Dilemma. Wikipedia sums the Innovators
Dilemma rather nicely:
First published in 1997, Christensen's book suggests that successful companies
can put too much emphasis on customers' current needs, and fail to adopt new
technology or business models that will meet customers' unstated or future needs;
he argues that such companies will eventually fall behind.
Displayed as a graph, the innovators dilemma is represented as the markets demand
versus the progress of a sustaining as well as a disruptive technology.
13
Tsahi Levent-Levi
Sponsored by
Performance
WebRTC and the Innovators Dilemma
Time
Figure 1: The innovators dilemma and WebRTC
In our case, WebRTC is the disruptive technology.

WebRTC is similar to current day VoIP, but at the same time very different:
It targets a different type and set of developers

It reduces barrier of entry for developers and reduces friction for end users
It relies on a different set of technologies in the backend
Incumbents faced with WebRTC, wont see the big difference between it and their current
technology. This has the danger of making them complacent.
14
A WebRTC JTBD (Job to be Done) Weekend

WebRTC JTBD: Reducing Barriers of Entry
WebRTC JTBD: Reducing End Customer Friction
Whos a WebRTC Vendor? The BlogGeek.me WebRTC Litmus
Test
Tsahi Levent-Levi
Sponsored by
Browser Support
WebRTC is an evolving standard. It started in 2011 and since then, it has progressed
faster than any earlier VoIP protocol both in its specification as well as in availability of
implementations. That said, there are differences in which parts of WebRTC browsers
implement, if at all, and availability of WebRTC in mobile browsers is patchy at best.
The diagram below gives a short historic view into the progress of WebRTC
implementation across the main browsers supporting it: Chrome, Firefox and Opera
May 2011: WebRTC Announced

Google releases WebRTC source code for
the first time under a permissive BSD
license
2012
Jan 2013: Firefox 20 adds WebRTC

First release of Firefox supporting WebRTC.
Comes with GetUserMedia support only,
which gives access to the local camera
Nov 2011: Chrome 23 adds WebRTC

No optional flag is required. Data channel
capabilities not supported
2013
Feb 2013: Interoperability
Initial interoperability between Chrome
and Firefox browsers achieved. This is still
early on in the process, so things still don't
work as expected, but this is an indication of
things to come
Jun 2013: Firefox 22 released

First Firefox release that officially supports the ability to
make video calls as well as use the Data channel API
Aug 2013: Chrome for Android
Chrome 29 for Android now fully supports WebRTC
Jul 2013: Chrome for Android Beta with WebRTC

The beta version of Chrome 29 on Android supports
WebRTC
Sep 2013: Firefox for Android supports WebRTC
The beta version of Chrome 29 on Android supports
WebRTC
Source:
15
Tsahi Levent-Levi
Sponsored by
Oct 2013: Opera 18 Beta intros WebRTC

First Opera release based on Chromium,
providing immediate WebRTC support
2014
Feb 2014: Opera for Android Beta with WebRTC
First Android release for Opera supports WebRTC
Mar 2014: Opera for Android with WebRTC

Opera 20 for Android has WebRTC in GA
May 2014: Microsoft promises to support GUM

Microsoft indicates in its IE status page that it plans
to support GetUserMedia APIs in its next version of
Internet Explorer
Oct 2014: Microsoft announced ORTC support
Microsoft officially announced plans to support ORTC
(WebRTC 1.1) in a future release of Internet Explorer
Source:
Figure 2: Major milestones of WebRTC support in browsers
As you can see, browsers are updated frequently, filling in gaps in the support of certain
WebRTC aspects. The table below provides the current state of affair:
Chrome
Firefox
IE
Safari
PC
Android
PC
Android
Get User Media
Peer Connection
Data Channel
Table 1: Browsers support for WebRTC
The table indicates two important aspects of WebRTC adoption:

1. The dominant players are Google Chrome and Mozilla Firefox2. Apple Safari and
Microsoft Internet Explorer are not taking part in the adoption of WebRTC,
though Microsoft announced plans to adopt it in the future
2. Support is mobile is painfully lacking for iOS
The Opera browser for desktop and Android also has WebRTC support
16
Tsahi Levent-Levi
Sponsored by
In the next subchapters I will indicate what solutions, if any, are available for developers.
Up to date browser support of WebRTC [&yet]

Overview of browsers support of WebRTC [NoJitter]
WebRTC: Who Will Surrender First? Apple or Microsoft?
Decrypting Googles Roadmap for WebRTC
Interoperability between Chrome and Firefox

Not everything is equally implemented in both Chrome and Firefox. That is to be expected
this early in the game, with new browser version being released every other month. It
should also be noted that there are two important aspects that WebRTC brings that
weren't available before:
1. Interoperability is less of an issue prior to WebRTC, 10's of different
implementations needed to be tested against each other for interoperability. I
attended such interoperability events they are a lot of fun and usually packed
full with vendors testing their applications and devices against each other. With
WebRTC, we are down to a handful of browsers that need to deal with the real
headache of interoperability. So consider this one mostly "done", while expecting
the same incompatibilities you will find with other HTML5 features across
browsers.
2. Interoperability bugs get fixed faster with browsers being updated regularly
and automatically these days, you can expect known interoperability issues to be
solved by browser vendors rather fast, and reach end users faster than any other
VoIP solution out there.
While the basics of WebRTC are supported well by both Chrome and Firefox, any
deviation from the basics may get to the realm of an unimplemented feature of WebRTC
in one of the browsers or a simple mismatch between them. Plan for it in advance and
make sure to try any out of the ordinary aspect of your service as early as possible on
both browsers.
17
Tsahi Levent-Levi
Sponsored by
Handling Safari and IE

Apple and Microsoft are staying at the sidelines when it comes to WebRTC. While both
companies are showing interest in WebRTC, nothing is publicly known of their plans.
To make things even harder to deal with, Google has discontinued their Chrome Frame
plugin for Internet Explorer this year. Chrome Frame enabled the use of a single plugin
for IE that enabled any version of the IE browser to support the latest HTML5 features,
WebRTC included. With this discontinuation, the ability to use IE with an official Google
support for WebRTC on it has become impossible.
This situation leaves developers at a disadvantage if they select WebRTC. If you plan on
starting a WebRTC-based service, you will need to take this into account with one of 4
options available to you:
1. Support only Chrome and Firefox. Anyone trying to access your service via IE
or Safari can receive an indication that they need to switch a browser. To see if
this fits your service, you need to research into the browser use in the region you
plan to launch your service and your specific target audience. Large enterprises,
for example, might not allow running anything other than IE on their employee
machines.
2. Write your own plugin. You can wrap WebRTC as a plugin and have it installed
when users access your service from IE or Safari. This adds some friction to
using your service, but might be a better solution than deciding to ignore such
users. Again, there are cases where the use of plugins will be forbidden, mainly
in large enterprises.
3. Use a free plugin. There are several plugins available on the market today. The
two most known ones are Temasys free plugin for IE and Safari; and Priologics
open source plugin for IE. Same restrictions as the option of writing your own
plugin apply here.
4. Dont use WebRTC. WebRTC is great, but it might not be for you. If your service
mandates the use of IE without use of plugins, then WebRTC cannot be used at
this point in time. Just select another technology, a different target market or a
different service.
18
How to get support into old browsers? [Disruptive Analysis]

The future of WebRTC in Internet Explorer
WebRTC Plugin: Temasys or easyRTC? Free or Open Source?
8 Things to Consider when Selecting a WebRTC Plugin
Tsahi Levent-Levi
Sponsored by
Dealing with Mobile

Mobile poses an even larger challenge with WebRTC:
On mobile, browsers are not the main channel for user consumption. Applications
are
Only Chrome and Firefox in their mobile versions support WebRTC, and that on
Android devices only no iOS devices are supported at this point
For most of the use cases on mobile, the requirement will be to wrap WebRTC
inside an application instead of using the browser as the point of access.
Downloads in millions
If you look at the download numbers of these two browsers and compare them to the
number of activations of Android devices and other downloadable mobile browsers, the
following picture emerges:
1200
Browsers
Other apps
1000
800
600
400
200
Maxthon
Dolphin
UC Browser
Firefox
Opera
Opera Mini
Chrome Candy Crush Whatsapp
Facebook
Popular Android browsers supporting WebRTC

Popular Android browsers NOT supporting WebRTC
Popular Android applications
Figure 3: Browser downloads on Android
19
October 2014
Tsahi Levent-Levi
Sponsored by
While browser downloads on Android have significantly improved in the past year, there
is still a long way to go. The figure above brings with it several insights:
Chrome browser is hugely popular on Android because it is being installed by

default on devices since the Android KitKat release
Other browsers arent fairing as well people dont usually install browsers on their
devices
In June 2014, Google announced 300 million monthly users of Chrome on mobile.
That includes both Android and iOS. This is below over 500 million downloads of
Chrome on Android, showing that people dont necessarily use the browser on
their device
Games and communication apps are more popular than browsers something
true for apps in general
That in itself isn't bad news, depending on the service you have in mind. Current behavior
patterns on mobile show that consumption of online services comes from apps and not
directly through the web browser: Flurry Analytics estimates that 86% of the time spent
on iOS and Android is on Apps, while only 14% is spent in web browsers. Most amount
of the Apps time is spent on Gaming, Facebook and Social Messaging.
An additional advantage of creating an application for mobile as opposed to using the
browser directly, is the opportunity to address challenges posed by the mobile
environment. These include network changes (switch between WiFi and cellular and no
coverage, camera rotation, interruptions from incoming native calls, etc.
Using the controlled environment of an application can assist in overcoming these issues,
but require the care and attention of the developer or the 3td party SDK used for the task.
Check how your service gets consumed on mobile. If users are expected to "bump" into
it when browsing the internet or by specifically going to your service via an app. Going via
an app means you can use WebRTC in its ported form, wrapped inside an application
something that several vendors have already done. Assuming you want WebRTC to be
embedded in your app, there are three different options open for you:
1. Port and integrate WebRTC on your own for the mobile devices you plan to
support. Vonage, for example, has taken this route
2. License a ported WebRTC stack from vendors offering such a service. There are
multiple WebRTC outsourcing outfits today, and some of them offer porting
services
3. Use a 3rd party WebRTC API platform that already have their offering ported to a
mobile SDK.
20
Tsahi Levent-Levi
Sponsored by
Predictions on annual app downloads on app stores [Gartner]

Where is WebRTC positions on mobile platforms
The challenges of porting WebRTC to mobile devices
Flash to WebRTC: Why Gruveo Finally Converted
Choosing a WebRTC API Platform report
WebRTC on Devices
An additional opportunity of utilizing WebRTC is in other devices set-top boxes, video
room systems and Chromecast-like dongles are some examples.
In this case, a device or component that needs to be able to process and send/receive
media can achieve that by a porting WebRTC and embedding it within the device.
There are two main reasons to select WebRTC for these devices:
1. WebRTC has a very permissive open source license, making it easy to use as
the basis of a generic media processing component. It has also been proven to
be portable numerous times already.
2. By using WebRTC embedded in the device makes it easy to interact with these
devices from other mobile devices, laptops and browsers.
While some may complain that WebRTC isn't a stable standard yet, using it may prove
the best approach for those who want to limit vendor lock-in for their media processing
components and systems.
21
Chromecast as a predicate of M2M use of WebRTC

Motion detecting baby monitor with WebRTC [webrtcH4cKS]
Comcast Getting it Right with WebRTC: Opts for Video Streaming
over Video Chatting
Tsahi Levent-Levi
Sponsored by
WebRTC API Components

WebRTC APIs are designed around 3 main objects. These are:
GetUserMedia, enabling access to the cameras and microphones of the device

PeerConnection, taking care of all media transport activities in WebRTC
DataChannel, dealing with proprietary data transmission between clients
GetUserMedia
The Get User Media APIs enable capturing the input devices: camera and microphone,
giving access to them via JavaScript. To get over privacy issues, when these APIs are
called in a website, the user is asked for permission to give such access.
This API can be viewed as external to WebRTC in a way, it doesn't deal with
communication at all. That said, Get User Media opens up possibilities for a wide variety
of services from photo-booth experiences to capturing a user's photo when onboarding
a new service (this is what MailChimp does when you register a new account or edit your
profile).
HTML5 demos making use of GetUserMedia [Shiny Demos]

Face detection with GetUserMedia [Raymond Camden]
What if Microsoft was Really Serious about WebRTC
PeerConnection
At the heart of WebRTC lies the Peer Connection. This is where most of WebRTC's VoIP
implementation lies.
The PeerConnection API holds inside it everything: SDP negotiation, media sending and
receiving, handling network issues such as packet losses, codec implementations, NAT
traversal, etc. The API itself wraps it all in a simple package that enables adding media
streams into a peer connection, negotiating capabilities and NAT traversal.
Most vendors focus on adding value by using the PeerConnection APIs.
22
An example of using PeerConnection [Big Nerd Ranch]
Tsahi Levent-Levi
Sponsored by
DataChannel
The Data Channel can be considered the "jack in a box" of WebRTC. It is a generic
interface that can carry any type of data media or otherwise between 2 WebRTC
devices.
While doing that in applications is rather straightforward, WebRTC offers a solid
infrastructure that usually isn't available for developers, along with the fact that it offers it
inside a web browser in a real P2P fashion:
1. Ability to send data either reliably or unreliably, giving the flexibility for developers
to choose what fits their needs without adding additional applicative layers
2. Same security and NAT traversal capabilities used for sending media is extended
to the data channel as well
3. P2P support within the web browser for any generic data
4. Adheres to the network characteristics by limiting or extending bandwidth to what
is available dynamically
These aspects make the Data Channel an unknown factor in what developers can do.
Some of the early implementations that have adopted the data channel include:
Assisted P2P video streaming, similar to how BitTorrent is working

CDN (Content Delivery Network) P2P, where multiple users browsing the same
website share the resources and images they download from web pages amongst
themselves directly
Online file sharing
Gaming and chat applications, where low latency is required
23
Introduction to broadcasting over IP

Sending files using the DataChannel
WebRTC's secret weapon is the data channel [Chris Kranky]
7 Creative Uses of WebRTCs Data Channel
Tsahi Levent-Levi
Sponsored by
Networking
WebRTC deals with networking much in the same way that other VoIP solutions treat
them. The main difference you will find is that WebRTC leaves a lot less options for
developers in certain areas; and in the case of WebRTC, fewer options means better
solutions.
VoIP standardization has suffered for over a decade from over-complexity. Put simply, if
something can be done with a specific VoIP protocol, it can probably be done in multiple
ways, making development, testing and interoperability a real headache.
WebRTC took the route of as little options as possible when it comes to issues of how to
implement the basics. This chapter covers several areas where such decisions were
made, providing a glimpse into the mindset behind WebRTC.
While WebRTC is based on existing specifications, it did divert from the tried and true
of VoIP. In some cases, it selected to support specifications that werent popular or in
wide use in the industry. In other cases, it tweaked and improved the existing
specifications to fit its own needs.
The topics covered here are:
24
P2P the type and nature of the peer to peer that WebRTC provides
Signaling what role does signaling take in WebRTC
Firewall and NAT Traversal the mechanisms of NAT traversal supported by
WebRTC
Security what is the security paradigm employed by WebRTC
Tsahi Levent-Levi
Sponsored by
P2P
WebRTC is said to be a P2P protocol, by which it means that 2 peers connected via
WebRTC get connected directly to each other. While that may well be true, there are
several aspects of WebRTC's P2P nature that need clarification:
1. WebRTC is no different than most VoIP solutions

WebRTC provides only the media path, and many of its predecessors use much the same
techniques to send media directly between peers. In a way, there's nothing new in
WebRTC's P2P nature when compared to SIP-based deployments or other proprietary
deployments. As with other VoIP protocols, WebRTC can work P2P or via servers it is
up to the implementer (and the network conditions) to decide which architecture is used.
2. WebRTC is the only P2P technology in web browsers

The main difference of WebRTC's P2P nature is within the web browser itself. Web
browsers have been designed to allow communication between the web browsers to a
server. Up until WebRTC, there was no way for one web browser to talk directly to anther
web browser without the intermediation of a server. With WebRTC, that becomes possible
Chat session between web clients
sans WebRTC
Chat session between web clients

with WebRTC
Web Server
Web Server
Signaling
Signaling
Chat
Web Browser
Web Browser
Web Browser
with WebRTC
Chat
Web Browser
with WebRTC
Source:
Figure 4: Chat session between Web browsers with and without WebRTC
25
Tsahi Levent-Levi
Sponsored by
3. WebRTC may require media relay

While WebRTC offers a P2P solution, at times, it requires relaying the media itself via
intermediary TURN servers. More on that in the next subchapter on NAT Traversal.
WebRTC Media Traffic With and Without a TURN Server
TURN
Server
Web Server
Web Server
Signaling
Signaling
Media
WebRTC
Client
Media
WebRTC
Client
WebRTC
Client
WebRTC
Client
Source:
Figure 5: WebRTC media traffic with and without a TURN server
4. WebRTC requires external signaling, which isn't P2P

While WebRTC offers a P2P solution, in order to reach another peer, there needs to be
some centralized decision making processes that occur on the external signaling protocol
that each service choose to adopt.
26
4 facts about WebRTC and P2P

The role of the server in a WebRTC P2P service [vLine]
Tsahi Levent-Levi
Sponsored by
Signaling
Signaling was specifically left out of WebRTC. There are those that see this as a positive
decision while others believe that signaling needs to be added under the fold of WebRTC.
The reasons for not having signaling as part of WebRTC are various:
1. Adding signaling to the WebRTC specification would drag the process through
standardization there are more existing options for VoIP signaling than there is
for media processing
2. WebRTC can be hooked up to use different signaling protocols today. Selecting
one mechanism, either existing or new, would mean reducing the use cases and
deployments where WebRTC can bring value
3. VoIP signaling is good for VoIP. WebRTC adds a web component that was always
missing in VoIP, where a lot of services already have some kind of signaling. Being
able to accommodate these services made sense
The only component within WebRTC that does deal with signaling is SDP. SDP stands
for "Session Description Protocol" and is used for describing multimedia session
capabilities and negotiating them. It is a critical first step towards connecting media
streams.
SDP is used in SIP, and as with other VoIP related issues, there are those that believe it
isn't suitable for WebRTC while others see it as sufficient.
SDP can be considered as a "necessary evil" which will be part of WebRTC for years to
come until a better/different solution is negotiated in the standardization bodies.
A more detailed discussion on the various signaling options can be found in the next
chapter.
27
The death of signaling as we know it

Why signaling isn't important anymore
Node.js at the service of WebRTC
Why we need signaling for WebRTC [WebRTC.is]
Signaling options for WebRTC applications [webrtcH4cKS]
SDP: Session Description Protocol RFC [IETF]
Tsahi Levent-Levi
Sponsored by
Firewall and NAT Traversal

To get media flowing from one browser to another, the media packets may need to pass
through NAT devices. NATs are Network Address Translators. They are deployed
everywhere in our home DSL boxes, enterprise boundaries and in our service providers'
backend. They have various roles and the come in different shapes and sizes. To that,
enterprises and home devices add Firewalls, preventing unauthorized access to the
network these are unaware of VoIP and WebRTC related traffic and block it by default.
In most daily scenarios, our devices have local IP addresses ones that cannot be used
globally. When we access the web, the NAT device replaces the local IP address of our
requests with a public IP address. This process is done in different ways by different NAT
devices. This causes headaches for those who need to pass NATs to communicate
especially when that communication is bidirectional in nature.
The main challenge with NAT traversal in WebRTC is the fact that it is a new challenge
for web developers, who are accustomed to having data pass through NATs and firewalls
seamlessly for web traffic. They are unaware to the intricacies, challenges and cost of
dealing with NATs in the real-time communications domain.
The selected solution for WebRTC in the area of NAT traversal is the use of ICE, which
in turn makes use of STUN and TURN. These are the same threesome of protocols used
by SIP. As with other such similarities, here too WebRTC takes the approach of using the
latest possible drafts and ideas and mandating them for use with WebRTC.
Trying not to dive into details, here is a short explanation on each of the 3 specifications
used in WebRTC for NAT traversal:
1. STUN provides means for the WebRTC client to find its own public IP address
simply by asking an external STUN server. Knowing this information in many cases
is enough to get a WebRTC session to connect
2. TURN when STUN fails, TURN can be used to relay all media between the
WebRTC clients. This requires more bandwidth and processing power from the
backend, but it gets more calls to connect
3. ICE ICE is in charge in orchestrating this whole show deciding which route and
NAT traversal mechanism best fits a given moment in time for the session
28
Tsahi Levent-Levi
Sponsored by
At the end of the day, STUN and TURN servers need to be deployed; which is a hassle.
Developers use multiple solutions for such a case:
1. Decide to deploy only with STUN support, and use publicly known STUN server
addresses (hoping they will not be moved or removed)
2. Deploy with STUN and TURN, installing and maintaining such servers as part of
the operation
3. Use third party SaaS provider for NAT traversal, who provide paid STUN and
TURN servers access
It should be noted, that there are cases where none of the techniques used by WebRTC
can successfully get a session to connect. This usually happens when the endpoints in
question are within enterprises with strict security rules. There are multiple drafts
proposed to the IETF to address this issue mainly by tunneling media traffic over web
sockets.
29
Introduction to NAT Traversal in WebRTC [webrtcHacks]

An overview of NAT Traversal by AudioCodes [SlideShare]
ICE always tastes better when it trickles
ICE: Interactive Connectivity Establishment RFC [IETF]
Tsahi Levent-Levi
Sponsored by
Security
Security is part and parcel of WebRTC. Where other VoIP protocols have taken the route
of allowing security as an optional feature (usually as an add-on years after the original
standard was specified), WebRTC has no means of sending media without having it
encrypted.
This approach reduces the number of options and decisions developers have to make
about the way they use WebRTC and results in a robust solution suitable for modern
communication.
There are several important aspects to remember when dealing with WebRTC security:
1. WebRTC uses SRTP and AES encryption for the media being sent. These are
both widely used and mature specifications
2. While media communication is secure with WebRTC, the signaling part isn't. It is
up to the developer doing the integration with a specific signaling solution to take
care of protecting that path it can be as "easy" as placing that communication
over HTTPS/TLS
3. WebRTC in the browser have an additional factor of security any security holes
found within WebRTC are handled by the browser vendors. While this reduces the
control of implementers, browser vendors have been very responsive to date in
closing known security holes much faster than vendors in other domains
4. With the current awareness around security and privacy due to the Snowden Affair
and talks about the NSA practices with both carriers and web vendors, WebRTC
is sometimes touted as a possible solution due to its secured nature. That is usually
overstated as WebRTC is a technology that can be used in decentralized
scenarios but also in centralized ones.
Never take security for granted, but keep in mind that WebRTC brings with it top notch
capabilities in security that are not readily available elsewhere.
30
Why WebRTC is the most secure VoIP technology

How security and privacy may be threatened by WebRTC
What can WebRTC tell you about your local network
Tsahi Levent-Levi
Sponsored by
Requested Features
WebRTC is a relatively new technology. It has been with us for 2 years only and has
already amassed a nice number of vendors around it. There are a lot of gaps in this
technology for those who wish to employ it for various use cases. These can be seen as
hurdles to some and as opportunities to others.
In this chapter, I will outline the main pain points vendors are expressing today with
WebRTC.
What developers have to say about their WebRTC needs

WebRTC infrastructure is tricky
8 Gaps Left by Google in WebRTC
Multipoint
Support for multipoint video calling is an issue for developers. For those in need for such
a feature in their use case, make sure to find a solid architectural solution.
Multipoint support means the ability to have more than 2 participants within the same
session. The common practice today with WebRTC services is to build video conferences
of up to 4 participants by directly connecting each browser to all other browsers. This
practice requires processing power and bandwidth in the endpoints, which means it
doesn't scale well.
The other option is to use a central routing/switching/transcoding unit in the backend that
takes care of all media mixing for the participants. Such solutions vary in size and price,
so if you need to go that route, start by writing down the use case and flow you need to
support before searching for a solution that fits.
31
5 part series on multipoint topologies in WebRTC

WebRTC Multipoint: Mix, Mesh or Route?
Tsahi Levent-Levi
Sponsored by
Recording
Many communication related use case require the ability to record the session.
While recording can be done on the browser side to some extent, its real value and
potential can only be realized when done on the server side. Most vendors and services
today don't offer a recording capability, but have expressed interest in adding one.
Recording in WebRTC on the server side can be done in one of two ways:
1. Route the session through a server instead of P2P and have the server side handle
the recording. This requires more processing power and bandwidth on the server
side to manage, but reduces the upload bandwidth as well as processing power
required from the client side
2. Have the client's JavaScript code send the media to an additional participant which
is the recording server. This makes it easier to decouple the recording capability
from the signaling and media paths of the sessions, but requires more upload
bandwidth as well as processing power from the client side, so it might not fit all
use cases and all geographies
It is important to remember that recording doesn't come alone: it also requires the ability
to archive and store the media sessions as well as retrieving them. The retrieval part can
be rather tricky, as it will depend on the devices and methods in which you plan on offering
playback capabilities for the sessions enabling retrieval on Internet Explorer or direct
streaming to mobile devices will require transcoding the sessions either in advance or on
demand.
32
Why something as simple as recording is complex [NoJitter]
Tsahi Levent-Levi
Sponsored by
Interoperability
WebRTC can live in a flux, where a specific use case can live in its own island. In such
cases, there is no external ecosystem you need to connect with. In some other use cases,
connectivity to other domains is mandatory things like integrating with an existing PBX
system or to the global PSTN are examples of this.
Such a requirement around connectivity is usually referred to as interoperability.
Interoperability in WebRTC happens in three separate domains, each with its own
challenges:
1. Signaling adding signaling on top of WebRTC isn't always possible and has its
own nuances. SIP and H.323, for example, are challenging as WebRTC signaling
protocols:
a. Accommodating SIP isn't supposed to be hard, but it does have its nuances:
the SDP is managed slightly different, and many of the RFCs and features
that are mandated in the media streams for WebRTC are considered new
and optional for SIP. Bridging the gap requires careful planning and
awareness to the problem
b. H.323 cannot be glued on top of WebRTC within the browser itself. This
may be true with other signaling protocols as well. Providing connectivity
between WebRTC clients to an H.323 deployment is not trivial
2. Networking WebRTC comes with mandated security as well as support for the
latest RFCs (ability to send multiplex media on the same network port, etc.). It
means that existing and legacy deployments can't directly interoperate with
WebRTC without some mediation in-between.
3. Codecs Some of The codecs selected for WebRTC are different than those used
in other VoIP deployments. This means that connectivity can occur only by
transcoding the media between WebRTC and external networks.
These challenges are usually solved by mediation mechanisms, where the most common
ones used are gateways and session border controllers (SBC). There are numerous
vendors offering this equipment in different shapes and sizes, so finding one that fits your
needs without the hassles of developing it from scratch is possible.
33
Multiple vendors announce gateways for WebRTC

The necessity of gateways
Tsahi Levent-Levi
Sponsored by
WebRTC vs. the World

WebRTC doesn't exist in a void. It is a relatively new entrant to an established VoIP
market. It bears a lot of resemblance to other technologies. If you are in search for a
technology to fit a service you have in mind to create, it is important to understand the
commonalities as well as the differences that WebRTC brings with it.
First step in this understanding is to know the anatomy of VoIP solutions.
Anatomy of VoIP Solutions

A VoIP product is usually based on 4 core components:
Infrastructure
Signaling
Media Processing
Codecs
Source:
Figure 6: Anatomy of VoIP - the 4 core components of any VoIP solution
This section will review each of these components and how WebRTC handles them.
34
Tsahi Levent-Levi
Sponsored by
Codecs
A full HD video requires 1920x1080 pixels per frame with 30 frames per second. If each
pixel requires 24 bits of color, this gets you to about 186MB of data per second an
impossibly large amount of data even for 1Gb networks.
To that end, codecs (code-decode) are used. Their sole purpose is to encode
(=compress) video and voice signals prior to sending them and to decode (=decompress)
these signals once received. For VoIP, these codecs are lossy they lose information
when they encode to save on space, usually by making assumptions around the human
eye and the human ear and also by external restrictions such as the quality required or
the bandwidth available.
WebRTC comes with 2 mandatory codecs:
1. G.711, the baseline voice codec. Offers narrowband voice quality, without any
compression or resiliency as part of the codec. Ubiquitous in its availability in
existing products.
2. Opus, a top of the line wideband codec that made its first appearance in
WebRTC and is now trickling to other VoIP products as well.
There is an ongoing debate over the mandatory video codec in WebRTC.
Google has taken the stance of supporting VP8 and is working on adding VP9 to
Chrome. These are "proprietary" video codec of Google
Microsoft and Apple are gravitating towards H.264, the standard video codec used
everywhere today
Cisco has released an open H.264 implementation, paid for in terms of royalties.
It has aligned with Mozilla who are supporting both VP8 and H.264
The main reason for VP8 and VP9 royalty free codecs (as opposed to H.264 which is a
royalty bearing codec).
Each VoIP product and protocol makes its own selection of mandatory and optional
codecs, usually having multiple codecs supported to enhance the potential of
interoperability.
35
Tsahi Levent-Levi
Sponsored by
Codecs bring 3 main challenges to developers:

1. Which codecs to select for a given product?
2. Who to license the codecs from to support a given use case and device types?
3. How to handle licensing terms and royalty payments, fitting it into the ROI
equation
WebRTC reduces these challenges in a lot of the use cases, but not in all of them.
36
The need for a mandatory codec in WebRTC

Does the future of WebRTC lie in VP9 or H.265
H.265 may never happen [NoJitter]
In favor of H.264 as a mandatory video codec [WebRTC.is]
On Opus voice codec
iPhone 6 FaceTime Supports H.265. Where is VP9 for WebRTC?
Microsoft, IE, ORTC, WebRTC, Skype & H.264: Where to now?
Tsahi Levent-Levi
Sponsored by
Media processing
While codecs do a decent job, they are not enough. There are other aspects that need to
be taken care of when it comes to sending media across networks. The two main ones
are:
1. Networks aren't created equal, and what is being sent might not really be
received on the other end, so someone needs to handle these things
2. There are local transient issues, such as echo between the microphone and the
speakers of a device that needs to be cancelled, varying volumes from changing
position of the speakers, etc.
These can all be wrapped into what I call "media processing". To that end, vendors will
either self-develop or license "media engines" to do the work. Part of this module is
straightforward implementations while other are based on best practices, heuristics and
"black magic" that developers have placed into their code.
In the case of WebRTC, this is what that Google has acquired from GIPS and ended up
open sourcing as parts of what WebRTC is. There are developers who simply rip media
processing algorithms from WebRTC and integrate it into their own products where they
see fit.
There are three main approaches to WebRTC when it comes to media processing, where
different approaches may be selected when dealing with the client side part and the server
side part of media processing:
1. Use WebRTC as is, relying on whatever the web browser (and in effect Google at
this point in time) to deliver the quality necessary and handle the network
impairments on its own
2. Replace the innards of the media processing components of WebRTC to
provide a better experience in the given use case. This is done either on the server
side or in closed applications on mobile or desktop where WebRTC can be
controlled and modified. The things that people replace in these cases are:
a. Rate control mechanism, which handles the decision of how much
bandwidth is available at any given point in time
b. The audio algorithms things like echo cancellation, noise suppression and
audio gain control
3. Use WebRTC's media processing components in a totally different setting
treating it just like pieces of code to be repurposed elsewhere. More on this in the
ecosystem chapter
37
Tsahi Levent-Levi
Sponsored by
Signaling
Now that we are able to send media from point A to point B, the question then becomes
how A knows about B and how do they signal to each other their intents. That is done by
using signaling protocols.
In the domain of VoIP, there are essentially 4 options:
1. H.323, an ITU protocol that is used mainly in video conferencing deployments
with dwindling support for it
2. SIP, the IETF equivalent of H.323. Used everywhere these days
3. XMPP, another IETF protocol that started with presence and IM focus but has its
own extensions for voice and video chat (Jingle)
4. Proprietary, where each vendor to its own. A good example to it is Skype
WebRTC decided not to deal with signaling at all and left that to implementers. The
selection of the signaling protocol used will affect the capabilities you will be able to
achieve and will usually be linked to the types of devices and ecosystems you wish to
connect.
The death of signaling as we know it

Why signaling isn't important anymore
Node.js at the service of WebRTC
Why we need signaling for WebRTC [WebRTC.is]
Signaling options for WebRTC applications [webrtcH4cKS]
Infrastructure
Infrastructure is where VoIP service diverges from each other and where real
differentiation happens. It is how VoIP clients get connected to back ends, what additional
capabilities are provided just because there's infrastructure involved.
What WebRTC did was commoditize codecs, media processing and signaling making
them all a lot less relevant and enabling vendors to focus on their infrastructure and use
cases. This is why there are so many new use cases that WebRTC vendors are
implementing that were technically possible before WebRTC but never touched they
required too much effort for the return they provide.
38
Tsahi Levent-Levi
Sponsored by
H.323 and SIP

A lot of times when VoIP is mentioned, it refers to the H.323 and SIP protocols. While
these protocols come to solve the same problem set and have similar philosophies and
worldviews in a lot of areas, their markets are slightly different.
Protocol
WebRTC
SIP
H.323
Market
Websites and
web apps
Voice and IMS

deployments
Enterprise video
conferencing
Media
SRTP
RTP or SRTP
RTP or SRTP
STUN, TURN, ICE
STUN, TURN, ICE
H.460.x
SDP
SDP
H.245
Voice codecs
G.711, Opus3
G.711, other
G.711, other
Video codecs
VP84, H.264
H.26x
H.26x
API
Java Script
Unspecified
Unspecified
NAT traversal
Media negotiation
Table 2: Comparison of WebRTC to other prominent VoIP protocols (SIP and H.323)
H.323 controls a lot of the enterprise video conferencing market deployments, although
its market share is probably diminishing; and SIP exist practically everywhere else,
including in IMS5 deployments.
Both of these protocols use RTP to carry their media and optionally use SRTP SRTP is
mandatory in WebRTC. The media transport protocol of choice for WebRTC. On the NAT
traversal and media negotiation front, WebRTC bears great resemblance to SIP while
H.323 has its own protocols.
In the voice and video codecs, WebRTC differs greatly from SIP and H.323. While SIP
and H.323 offer a wide variety of choice, that choice causes a lot of headache in
interoperability of high quality codecs. For WebRTC, Google took the route of selecting
specific codecs with the main focus being latest available technology coupled with royalty
Some browsers offer additional optional voice codecs as part of their WebRTC implementation
All browsers supporting WebRTC today use VP8. Firefox is in the process of adding H.264 support.
Microsoft announced support for only H.264 at this point
5 IMS stands for IP Multimedia Subsystem and is the standardized solution for carriers. Its VoIP protocol of
choice is SIP and you can find it in VoLTE, RCS and other IMS related specifications.
4
39
Tsahi Levent-Levi
Sponsored by
free business model. There is an ongoing debate in the industry whether a popular royalty
bearing codec (H.264) should be used instead of VP8. While that debate isnt over, the
majority of WebRTC services are using VP8 today.
On the API front, WebRTC offers a specific API that is used by web developers while SIP
and H.323 are provided with different APIs in different languages from the various
vendors.
H.323 and SIP don't live within the browser and using them requires applications to be
developed, compiled, packaged and delivered as part of the service. That said, SIP is
moving towards the web browser by introducing SIP over WebSockets; making it a
suitable companion for WebRTC.
While WebRTC doesn't compete with these protocols, and can even be used and
interoperate with them, it does pose a threat to them, simply because it doesn't mandate
the use of any of them. Many developers today opt for using WebRTC with their own
proprietary signaling implementation to reduce the complexities associated with SIP and
H.323 signaling. By doing so, they shun the traditional VoIP world and its use cases to
implement their narrow use case. This approach doesnt force sessions in WebRTC to
adhere to the decade old telephony system and enables anyone, with or without VoIP
experience, to use WebRTC with ease.
40
WebRTC is shifting signaling to XMPP and JS

The rise of Java Script SIP implementations
Tsahi Levent-Levi
Sponsored by
Flash
Flash was the only mechanism in the web browser where real time communication could
occur prior to WebRTC. Flash brought with it several problems:
1. For a long time it supported only H.263 video codec, which was inferior. The end
result was poor video quality
2. Flash only worked with a server for all media transmission. This proves to be a
barrier for many use cases and for experimentation by developers
3. It wasn't HTML. The tools required for development weren't as common place and
as accessible as WebRTC is today
4. Its echo canceller was bad, which resulted poor voice quality in conversations
conducted over the PC speakers
5. To a large extent, it was non-existent in smartphones and tablets. While that is due
to Apple, its implementation on Android wasn't good enough
Many of the Flash based companies are now migrating towards WebRTC either doing
it purely by switching the technology or by providing both options: using WebRTC when
possible and downgrading to Flash when WebRTC can't be used.
How WebRTC and Flash differ [SaleMove]

Would you use WebRTC to Chat with Mom?
Flash to WebRTC: Why Gruveo Finally Converted
"Proprietary"
There are and will always be proprietary implementations. The most notable one today is
Skype. For the past decade, Skype has provided VoIP solutions based on a proprietary
signaling protocol. Even their selection of voice codecs wasn't reliant on best practices in
the industry.
This approach has worked well for Skype, enabling them to overcome technical
challenges that were debated in the standardization bodies. This gave them a head start
on most other competitors who relied on standardized SIP or H.323.
Opting for the proprietary route comes with its own set of challenges mainly the inability
to tap to the external ecosystems to license components while focusing on the core
business.
41
Tsahi Levent-Levi
Sponsored by
Vidyo
Vidyo is mentioned here due to its recent announcement with Google around adding its
SVC technology into VP9. VP9 is the predecessor of VP8. Google have already
integrated VP9 decoder into Chrome for YouTube videos. They plan on introducing VP9
to WebRTC during 2015.
Vidyo is based on its own technology, where SVC takes a key part. SVC stands for
Scalable Video Coding and it is a technique in which a video stream is encoded in layers
each layer adding more quality on top of the previous one. This layered approach allows
sending different layers to different participants, based on the needs of the conference
and the capabilities of the network and the devices. The end result is the ability to build
more efficient multipoint video bridges.
This kind of technology makes sense for use cases where there are more than 2
participants. Managing the backend in such scenarios isn't trivial and this is why Vidyo
went for the deal it hopes to bank on the ability to use WebRTC in the browser directly
for its customers and deployments without the need to install an application. For Google,
this means a better video codec and less server computations for its Google Hangouts
service.
42
Challenges and opportunities of Vidyo following the VP9 deal

How Vidyo's deal affects the rest of the UC market [Talking Pointz]
Google is targeting Skype with SVC deal [NoJitter]
How Google is Disrupting the Enterprise Video Conferencing
Market
Tsahi Levent-Levi
Sponsored by
WebRTC Hype
There is no hype when it comes to WebRTC.
Pundits have been debating this for quite some time now, with varying opinions. I believe
there isn't and wasn't any hype around WebRTC whatsoever.
When I started blogging about WebRTC about a year ago, I was alone in the field. Very
little information and thought leadership could be found about WebRTC on the web. Since
then, that has changed with individuals, vendors and media outlets now covering this
space.
Any apparent hype being felt can be attributed to many reasons:
1. WebRTC in a new technology. As such, the process of understanding it that
takes place these days in public on blogs look like hype for some: you will find few
observations these days about where SIP belongs within the domain of VoIP, but
a lot of writing about WebRTC's place in VoIP
2. WebRTC deals with VoIP and web. While VoIP belongs to a small set of
developers within the communication market, WebRTC opens up that market to a
distinctly larger set of developers. This increase in availability of communication
technology seem to some like hype
3. WebRTC reduces the barrier of entry. Use cases that had no ROI up until
WebRTC are now being introduced, enhancing the ways in which VoIP is use
In the past 8 months, I have collected data points indicating the interest around WebRTC.
43
Tsahi Levent-Levi
Sponsored by
Developers interest in WebRTC over time

4500
4000
3500
3000
2500
2000
1500
github projects
1000
Stack Overflow questions

Discuss WebRTC group
500
0
Jan-13
May-13
Aug-13
Nov-13
Mar-14
Jun-14
Sep-14
Dec-14
Source:
Figure 7: Developers interest in WebRTC over time
The graph above illustrates three domains that are prime destinations for developers
interested in WebRTC:
1. The Google group called discuss-webrtc
2. The number of github projects related to WebRTC
3. The number of questions around WebRTC on StackOverflow
44
Tsahi Levent-Levi
Sponsored by
LinkedIn profiles mentioning WebRTC

4500
4000
3500
3000
2500
2000
1500
1000
500
0
Jun-14
Jul-14
Aug-14
Sep-14
Figure 8: LinkedIn profiles related to WebRTC
The graph above details the number of LinkedIn profiles that have WebRTC somewhere
in their description. While this number is rather low, it is easy to see an upwards trend.
The results of these graph show a linear increase in interest this is an indication that we
are either before the real hype of WebRTC or that no hype will happen with WebRTC
at all.
45
There is no hype around WebRTC [NoJitter]

Why WebRTC is overhyped [NoJitter]
Why the hype around WebRTC doesn't matter [NoJitter]
Gartner Hype Cycle, WebRTC Not Included [The New Dial Tone]
WebRTC on the Rise? Geek Questions Hold the Answer [NoJitter]
Tsahi Levent-Levi
Sponsored by
The WebRTC Ecosystem

The WebRTC ecosystem is built of a large set of vendors, startups and individuals; all
with their own agendas. This chapter tries to analyze the WebRTC ecosystem from
several angles, each providing a different prism into WebRTC.
The 5 Uses of WebRTC

If I had to split the ecosystem and categorize vendor on the types of usage they make of
WebRTC, I'd be splitting them into 5 broad categories: Supporting, Using, Porting,
Repurposing and Adopting.
2nd Market
Vendors
Repurpose
Tooling
WebRTC
Core
Figure 9: Types of vendors in the WebRTC ecosystem
Before I go into detail about the categories there are a few things to keep in mind:
46
A vendor can be in multiple categories at the same time

Similar vendors chose a different category to belong to. The choice changes their
core competencies and basis of competition to some extent
Tsahi Levent-Levi
Sponsored by
WebRTC Core
The supporting role at the heart of WebRTC belongs to vendors who contribute to it
directly. Here you will find Google and Mozilla, who are embedding WebRTC into their
browsers, but also companies like Vidyo who are adding SVC technologies into future
WebRTC releases.
The amount of vendors in this area is small, as they require extensive knowledge into the
inner-workings of WebRTC. This role comes with little upside in the way of direct revenue
but gives a lot of power in ensuring technical superiority versus other vendors:
Mozilla and Google may use their support of WebRTC to attract additional users
from Apple and Microsoft, increasing their market share
Vidyo may use their assistance to WebRTC's technology stack in maintaining their
first mover advantage in multipoint video conferencing capabilities
Ericsson with their latest openWebRTC stack, rivaling Googles own WebRTC
implementation
Cisco who open sourced their H.264 video codec in an effort to sway the decision
of WebRTCs mandatory video codec towards H.264
These vendors usually don't intend on banking from WebRTC directly and have indirect
business models for WebRTC; at times the business model isn't easily apparent.
Tooling
Vendors in this position offer tools that make development of WebRTC-based services
much easier. These vendors have varying business models focused around B2B, where
the target market is other vendors who wish to deploy communication services.
These vendors can be split into several domains:
47
API vendors
Services vendors
Infrastructure and SDK vendors
Projects vendors
Open source vendors
Tsahi Levent-Levi
Sponsored by
WebRTC API Vendors

These are vendors who provide either hosted or on-premise platforms that offer an
abstraction on top of WebRTC as well as closing the infrastructure gaps of WebRTC.
They may offer features such as connectivity to PSTN, mobile SDK, downloadable
plugins or Flash support for non-WebRTC environments.
Such solutions are adopted by vendors who want to focus on their use case, mainly when
that use case requires communication to be a feature that is stitched into their business
processes. This has been coined CEBP (Communication Enabled Business Processes)
and is a strong trend in the uptake of WebRTC.
There are many vendors in this domain, some who started as Telco API vendors and
have added WebRTC to their arsenal of capabilities, and others who started as pure
WebRTC API vendors.
A shortlist of these vendors includes but not limited to: 2600hz, AddLive, Apidaze, Forge,
Kandy, OpenClove, TokBox, Respoke, SigntCall, Sinch, Tropo, Twilio and VoxImplant.
Who provides telco APIs

API platforms adding video with WebRTC
Choosing a WebRTC API Platform report
Services Vendors
There are pieces in WebRTC deployments that can be "outsourced" to others. These
pieces are offered as service, where the SLA, global footprint and focus of the vendor on
the specific service is the main offering.
In this domain, you can find XirSys as a notable NAT traversal providers, offering TURN
server hosting as a service to those who don't want to deal with media relay in WebRTC
and focus on their own service logic.
Another interesting set of vendors here are PubNub, Firebase and Pusher who offer low
latency messaging infrastructure for developers. All have added WebRTC into their
arsenal of tools; where they specialize in adding signaling to it.
The customers of the services vendors are usually those who are comfortable in
developing directly on top of WebRTC but view the hosting of WebRTC components as
less important to their core offering.
48
Tsahi Levent-Levi
Sponsored by
Whats Your SaaS for WebRTC Signaling?

An interview with XirSys CEO
An interview with MNS CMO
Infrastructure and SDK Vendors

These vendors offer either equipment or licensed software that makes it easier for
developers to build their solution. Here you will find a "breakdown" of features you'd
expect to consume from APIs and services vendors in packaged format building blocks
that can be used to create higher level solutions.
The target market of these vendors is developers who are looking for more control of their
solution. This includes large service providers and enterprises, and to some extent
vendors who decided to manage their own data centers or manage their own cloud
deployment and just need to bolt on a few features/capabilities.
Another interesting set of such solutions are PBX and soft switch vendors, who started
offering WebRTC connectivity. As they offer an engine of sorts, it can provide a good
starting point for many communication services. Since there are some solid open source
offerings in that area, they are also very popular with developers.
You will find in this group of vendors offering gateways, media servers, NAT traversal
server software, etc. Notable names include Dialogic, Eyeball Networks, GENBAND,
Digium and FreeSWITCH.
49
An interview with Digium's Director of Strategic Programs
Tsahi Levent-Levi
Sponsored by
Projects Vendors
As with any vibrant community, there are those who offer their services to assist
companies in building whatever it is they require. WebRTC include large and small
software houses of this type ones that take on development tasks related to WebRTC.
Some of these vendors offer a breadth of capabilities coming either from a broad
outsourcing and system integration heritage, others are more focused into the VoIP and
video processing space and the rest are starting off with WebRTC or can be considered
as "heads for hire", usually focusing on projects to local vendors.
WebRTC being a new domain, there is a real need for such vendors as well as a rise in
the number of vendors who are publishing their expertise in WebRTC and even investing
in showcasing it in.
An interview with Oxagile's CTO

Blacc Spot Media reference customer [Blacc Spot Media]
An interview with Blacc Spot Media CEO
Open Source Projects

WebRTC is building around it a large and healthy set of open source projects. At the time
of writing these lines (October 2014), there are over 2,800 github projects related to
WebRTC. This number is growing consistently and they show a vibrant community.
These projects usually sprout from the following reasons:
1. A single person, student or other, looking into the technology and just introducing
the results online for everyone to use and share. This is usually done to attract a
potential employer
2. Someone with a pain point that can be solved with WebRTC. OpenVRI is a good
example of this case a hard of hearing person who decided to build a solution he
will be happy to use on a daily basis
3. A vendor who believes in the open source movement and contributes components
back to the community while building his own services and offerings
How much these projects will be used as the baseline of future projects and products
outside of the scope of their original developer is an open question at this point in time.
My assumption is that developers who are building WebRTC-based services end up
50
Tsahi Levent-Levi
Sponsored by
"bumping" into them while Googling for things they need to do and decide in an ad-hoc
basis if and how to adopt them. This practice needs to be handled with care, as the
licenses attached to these open source packages may be limiting the business models in
which the adopting vendor will be able to operate in the future.
Examples of WebRTC open source modules

WebRTC Demo: How to Explain WebRTC to People
Vendors
Vendors are those who end up offering WebRTC-based communication services to the
end users. Their market may be B2B or B2C, depending on who they serve.
It is hard to cluster these vendors together, as they cater different verticals and use cases.
It is also where business models diverge greatly.
Vendors in this category may use a Tooling vendor to rely on, or just build everything from
scratch on their own. Reasons for selecting different routes here depend on a large variety
of parameters, unique to each case.
The use cases chapter will try to take a closer look as some of the markets to which these
vendors belong to.
51
Tsahi Levent-Levi
Sponsored by
Repurpose
Repurposing WebRTC can be viewed as the gray market of WebRTC. There are active
vendors in this space, where most of them aren't widely known their use of WebRTC is
such that is left behind curtains.
WebRTC, the open source project, as opposed to the specification, is a state of the art
media engine. It is the result of GIPS' commercial offering and On2's proprietary video
codec, coupled with Google's ongoing investment in its quality.
To put things in perspective, when GIPS got acquired by Google, many vendors who were
GIPS' customers immediately started searching for alternatives, which were hard to find
at that level of quality and capability. Among GIPS customers you could have found IBM,
Google, Yahoo, WebEx, Nortel, AOL, Citrix, Avaya, Samsung and QQ.
When Google open sources WebRTC under a very permissive BSD license, they in effect
provided anyone who needed a media engine free access to a high quality solution only
thing missing was an SLA.
Vendors have taken note, and some have pried bits and pieces of WebRTC that fits their
needs and embedded them into their own products and services. This requires little
publication of the fact, so most of these activities are left unknown in the market. The
most known and notable such move was the one made by Vonage. Their story is outlined
in the use cases chapter.
52
Changes in the media engine market landscape

How Vonage repurposed WebRTC for their own needs
An interview with Jan Linden of GIPS prior to its acquisition
[Wirevolution]
Tsahi Levent-Levi
Sponsored by
2nd Market
The 2nd Market of WebRTC is something quite rare but its existence is important to note.
WebRTC fuses between VoIP and the web. This brings two important traits together:
1. VoIP gains an open and modern delivery system the web browser
2. The web gains P2P capabilities and real time communication
The web is based today on loosely coupled web sites and technologies that are mashed
up together to build new solutions and services. The ProgrammableWeb is an important
directory of web APIs used for mashup6 purposes. It documents APIs, mashups of them
and stories around them.
At the moment, there are over 10,000 documented APIs and 7,000 mashups on the
ProgrammableWeb:
Figure 10: Growth pattern in web APIs
A mashup is a web page or web application that uses content/functionality from multiple sources across
the web. When one web service offers open APIs, other services can consume that API (as well as other
APIs) to generate a new service offering
53
Tsahi Levent-Levi
Sponsored by
This trend of mashup and open API publication is also something you see with WebRTC,
and not only with the API vendors: many of the WebRTC services players are starting to
offer APIs of their own for various purposes:
1.
2.
3.
4.
Integration into enterprise directory services

Integration with CRM/ERP systems of enterprises
Widgets and call buttons to place on websites
White labeling and branding via APIs
To some extent, this creates a 2nd market of companies who integrate with these services
vendors and still see themselves as part of the WebRTC ecosystem.
The most distinct example of this is Intuitive Solutions, who are running a managed
contact center service for customers. Intuitive Solutions in this case, have even publicized
this initiative in a press release and see themselves as a WebRTC player to some extent.
Figure 11: The WebRTC value chain and the 2nd market
Intuitive Solutions run their service on top of LiveOps, which provides contact center SaaS
systems for its customers. LiveOps have built their own solution on top of Twilio, a
WebRTC API vendor.
54
TNW Academy announcement [TheNextWeb]

The new value chain that WebRTC enables [UC Strategies]
WebRTC 2nd Market: The Story of Intuitive Solutions
Tsahi Levent-Levi
Sponsored by
WebRTC Use Cases by Verticals

This chapter holds the main body of this research paper. It consists of various verticals
and use cases of real vendors within those verticals who are using WebRTC in one way
or another.
The purpose of it is for entrepreneurs and product managers to get acquainted with the
market dynamics of WebRTC and to see how different vendors are tackling the technical
and business issues associated with the introduction of WebRTC.
Things to Remember
The vendors listed in this chapter were selected not due to their technical merit or
the type of support they provide these were not part of the analysis done. The
vendors listed here were selected as examples of the different services and
business models that vendors are offering in the WebRTC ecosystem.
Many of the vendors in this section have been interviewed for BlogGeek.me in the
past. You can find more about them here: http://bloggeek.me/webrtc-interviews/
Tooling
Tooling vendors are at the heart of the WebRTC ecosystem. They are the enablers filling
the gaps that the WebRTC specification doesn't solve this essentially mean a lot of
backend capabilities.
These vendors come in different shapes and sizes. I have selected 3 such vendors:
55
Digium the vendor behind the popular Asterisk SIP communication platform,
who has added WebRTC to Asterisk's capabilities
GENBANDs Kandy a Telco API vendor, enabling anyone to build a use case
on top of their platform instead of directly on top of WebRTC
XirSys offering a NAT traversal service in SaaS form
Tsahi Levent-Levi
Sponsored by
Digium
Digium offers a popular and widely used
open source communication framework
called Asterisk. Founded in 1999, Digium
have been in the VoIP arena a lot longer
than WebRTC.
Asterisk added support for WebRTC,
which enables anyone who need a PBX or
other SIP server to be able to connect calls
to it over WebRTC from a browser. This
support is being used by multiple software
vendors already.
Solution Type
PBX
Target Audience
Developers
Business Model
Dual License
Country
US
Twitter
@digium
Media Types
Voice, Video
Use of WebRTC
Direct integration
Asterisk isn't alone in this domain. There are other such open source VoIP platforms that
have added or are in the process of adding WebRTC support.
The advantages of adopting an existing VoIP platform that added WebRTC are:
Easier integration with existing VoIP deployments. Main domains that are already
using this approach are contact centers and unified communication solutions
Rich framework and architecture, with an existing ecosystem of developers and
outsourcing vendors
Recently, Digium have introduced a new WebRTC API Platform called Respoke.
Things to Remember
56
This route is usually selected by "hard-core" VoIP developers, especially when

connecting a legacy VoIP or PSTN system to the web
These platforms tend to have open source licenses some more permissive than
others. Check the license to see that it fits your business needs
This support of both WebRTC and legacy brings with it limitations to what can be
achieved with WebRTC things like video support, or Opus voice support comes
to mind here
Tsahi Levent-Levi
Sponsored by
GENBANDs Kandy
GENBAND is a vendor specializing in
carrier grade communication products.
In 2014, it launched a new Platform as a
Service called Kandy, which delivers its set
of capabilities via API and SDK to
developers, enterprises and carriers.
Kandy offers the ability to create real time
communication based services with little
knowledge of WebRTC or communication
related technologies. It does so by offering
REST APIs, SDKs for Web and mobile and
a cloud-based communication platform.
Solution Type
Telco API
Target Audience
Developers
Business Model
Subscription
Country
US
Twitter
@KANDY_io
Media Types
Voice, Video
Use of WebRTC
Direct integration, porting,

plugin
Kandy is targeting a variety of developer archetypes:
Service Providers, who wish to deploy the solution within their own controlled
environment and use it to build their own services, or white label it altogether
Enterprises, where an SLA and the ability to customize and offer a solution on top
of an API platform is necessary at times
The long tail, where any developer can on board and start developing immediately
with little friction
The things Kandys team is most proud of is their ability to simplify the development effort
required, taking out of the equation a lot of the challenges associated with developing
WebRTC and communication related services for web and mobile.
Things to Remember
57
Many vendors are currently competing in this space, each with his own
differentiating business or technology factors
Kandy provides an end-to-end Communication-as-a-Service solution
This space is dynamic, with several players getting acquired and taken off the
market. The selection of a vendor to work with is an important and non-trivial one
Tsahi Levent-Levi
Sponsored by
XirSys
XirSys offers a SaaS for WebRTC NAT
Traversal. In many cases, developers don't
wish to deal with the hassles of installing,
setting up; integrating and maintaining
STUN and TURN with their WebRTC
service.
In such cases, they search for vendors who
can provide this solution under an SLA.
The advantages of outsourcing this to an
external vendor includes:
Solution Type
SaaS
Target Audience
Developers
Business Model
Subscription
Country
US
Twitter
@XirSys
Media Types
Voice, Video
Use of WebRTC
Direct integration
Wider coverage around the globe

Ability to focus on user experience of the solution being developed
Flexible support for spikes in service use
Predictability of costs and reduction in ongoing development and maintenance
costs
A NAT Traversal SaaS vendor is also capable of providing additional insights to its users
via reporting and analytics areas where most developers will not invest until later into
the service lifecycle.
Things to Remember
58
SaaS can include other types of services. NAT traversal is the most obvious one,
but signaling, media processing and recording are all areas where developers
have indicated the need for assistance
When building a service, SaaS solutions can greatly reduce time to market with
the initial minimum viable product. In the longer run, it can simplify your work
Tsahi Levent-Levi
Sponsored by
Video Conferencing
Video Conferencing is a market that exists for over a decade. Its focus over the years has
always been Enterprise Video Conferencing, sold to multinational companies with large
IT budgets.
In recent years, companies have moved to offer Cloud based video conferencing
offerings, in an attempt to shift IT spending from CAPEX to OPEX and to start selling
video services instead. WebRTC is now part of this trend and can be viewed as an
accelerator of it.
Many of the initial WebRTC vendors tried to penetrate this market, with the main value
being a reduction in the end cost to the customer due to their lower development and
operating costs.
The vendors provided in this subchapter as example use cases were selected due to their
business models and geographical variety:
Pexip Norwegian startup company focused on cloud enterprise video

conferencing
Sqwiggle San Francisco based startup focused on collaboration between
remote workers in the same team
Veeting Rooms Swiss startup putting its focus on data privacy
59
WebRTC isn't only about gateways [UC Strategies]

Two ways for UC vendors to look at WebRTC [UC Strategies]
Tsahi Levent-Levi
Sponsored by
Pexip
Pexip can be viewed as the next
generation
of
enterprise
video
conferencing solutions. It comes with the
same world view of unified communication
solutions of multiple video participants in a
single meeting, but diverges from common
practices in the following ways:
Solution Type
Cloud Enterprise Video

Conferencing
Target Audience
SMB, Enterprise
Business Model
Subscription
Country
Norway
Everything is software based

@PexipInc
nothing
required
proprietary Twitter
hardware
Media Types
Voice, Video
The solution is fully virtualized a Use of WebRTC
Direct integration
new trend for enterprise video
conferencing solutions that require real time response from the operating system
As with other cloud video conferencing platforms for the enterprise, Pexip markets its
solution as connecting to any system be it H.323, SIP or WebRTC coming from any
device.
Pexip is a Norwegian company, based out of the same talent pool that started off
Tandberg a video conferencing vendor that got acquired by Cisco a few years ago.
Unlike the other vendors in this section, Pexip does not offer a hosted service, but rather
licenses its software to run on-premise in its customer's data centers.
Things to Remember
Video conferencing in the enterprise is hard. A lot of issues there revolve around
interoperability and transcoding. Only engineers with such domain expertise can
build a solution for that market
For Pexip, WebRTC is just another device connecting into their network it isnt
at the heart of their architecture
60
Overview of Pexip Infinity video MCU [NoJitter]
Tsahi Levent-Levi
Sponsored by
Sqwiggle
Sqwiggle is trying to solve the collaboration
challenges of remote workers in the same
team. It does so by redefining video
conferencing in enterprises and making
use of ambient video technologies.
Sqwiggle places the whole team in a single
web page, where still images of the team
members get updated several times a
minute. This creates a type of presence
that bridges the distance within the team.
Solution Type
Video
Conferencing,
Collaboration
Target Audience
SMB
Business Model
Subscription
Country
US
Twitter
@sqwiggleinc
Media Types
Voice, Video
Use of WebRTC
Direct integration
At a click of a button, a video session can

start between two peers in the team, without the need for the called peer to answer or
decline the call. This ends up feeling just like raising your head from the computer to ask
a colleague in the same room a question.
Interestingly, most of Sqwiggles video calls are really short. This is the opposite of
common enterprise use cases where video calls span 30-60 minutes on average (the
length of a scheduled meeting in the calendar).
Things to Remember
61
Video communications in the enterprise comes in different shapes and sizes. The
fact that legacy vendors have focused on a specific use case doesnt mean other
use cases dont exist
Presence can be a lot more fluid than online, offline, busy and away. Sqwiggle
redefined it by making taking a picture of its users and making them available to
the team
Sqwiggle doesnt aim to connect to any existing video conferencing equipment. It
lives as a standalone solution and works well in that context
Tsahi Levent-Levi
Sponsored by
Veeting Rooms
Veeting Rooms provides a web meeting
service enabling multiple participants to
attend a video conference.
While many provide similar capabilities,
Veeting Rooms differentiates itself from
the rest of the pack by putting emphasis on
data privacy.
Everything in Veeting Rooms from the
data center used to hosting their service, to
the third party SaaS providers for analytics
are located within Switzerland, where
strict data privacy rules exist.
Solution Type
Cloud Enterprise Video

Conferencing
Target Audience
SMB, Enterprise
Business Model
Subscription or Pay as you

go
Country
Switzerland
Twitter
@VeetingCom
Media Types
Voice, Video
Use of WebRTC
Direct integration
Veeting Rooms have built their system in a way that enables them to white label and even
replicate the solution to another geography if necessary.
Things to Remember
62
The Enterprise Video Conferencing market competed in the past decade mainly
on port density, resolutions and video quality. With good enough solutions, it is
becoming hard to differentiate
Small startups like Veeting Rooms are finding gaps and needs in the market that
are ignored by the incumbents and offer tailored solutions. Data privacy is such a
gap for the competing cloud providers of video conferencing
As with the case of Sqwiggle, Veeting Rooms refrains from addressing the existing
video products, instead making use of WebRTC in the browser
Tsahi Levent-Levi
Sponsored by
Telecommunications
Carriers have their own interest in WebRTC. This stems from the basic notion that
WebRTC provides communication capabilities that can be either wielded by OTT
vendors, competing on voice and SMS services, or by the carriers themselves.
The domain of carriers is larger than what can fit in such a paper, with multiple directions
that can be taken and are taken today. In this section, several non-obvious approaches
of carriers into WebRTC are described:
appear.in an ad-hoc video conferencing service based on WebRTC

Comcast video streaming to a set top box
Vonage broadband service provider in the US that is using WebRTC to "go
mobile" and introduce video
63
What WebRTC means to telecoms [Alan Quayle]

Network effects, Telcos and WebRTC [Baby is 60]
What are Carriers Doing with WebRTC? Everything But IMS
Comcast Getting it Right with WebRTC: Opts for Video Streaming
over Video Chatting
Tsahi Levent-Levi
Sponsored by
appear.in
appear.in define themselves as a startup
within the incubator environment of
Telenor Digital. They offer a simple, adhoc, multipoint video conferencing service.
The service started as an interns project.
It gained traction and popularity, which
caused Telenor Digital to invest more
resources into it.
Today, appear.in offers a decent video
conferencing service that can work with up
to 8 participants, along with an iOS app.
Solution Type
Video Conferencing
Target Audience
Consumers
Business Model
Exploration
Country
Norway
Twitter
@appear_in
Media Types
Voice, Video
Use of WebRTC
Direct integration, porting
Things to Remember
64
There are similar services out there. This is the only one developed and managed
by a Telecom vendor
It is an experimental service for Telenor, as there is no business model apparent
at the moment behind it
Vendors (especially incumbents) need to tinker and play with WebRTC to
understand its capabilities, and decide from there where it fits
There is more to do with WebRTC in Telecom than simply make it an access point
to the IMS network
Small teams in large companies can develop great services if you let them
Tsahi Levent-Levi
Sponsored by
Comcast
Comcast is the largest cable provider in the
US. Its XFINITY X1 Platform is used by
many of its customers as their Pay TV
platform.
Recently, Comcast announced its plans of
adding video streaming capabilities which
will work similar to how Chromecast does:
it will enable their customers to live stream
personal video content from their mobile
devices to the television via the X1 set top
box.
Solution Type
Video streaming
Target Audience
Consumers
Business Model
Subscription
Country
US
Twitter
@comcast
Media Types
Voice, Video
Use of WebRTC
Direct integration, porting
This capability is added by way of WebRTC.

On top of that, Comcast has invested in the Jitsi Videobridge, which provides multipoint
video calling capabilities. The reasons and target service for this investment has not been
made public.
Things to Remember
65
While the obvious addition of WebRTC to a set top box would have been video
calling, Comcast has taken the route of enabling better video streaming user
experience with WebRTC
There are many ways to use WebRTC besides enabling a peer to peer video call
WebRTC use cases are not limited to the browser or even to a mobile device. The
X1 platform is an embedded device that has WebRTC support to it
Telecom Carriers are large companies with multiple services. There are many
ways in which they can leverage WebRTC. Comcast seem to be doing it both as
a video streaming capability on their X1 platform and as an unannounced
multipoint video conferencing solution (based on their Jitsi Videobridge
investment)
Tsahi Levent-Levi
Sponsored by
Vonage
Vonage is a voice communication provider
that used WebRTC to build a mobile client
and to add video calling.
Vonage has been operating since 2001,
offering its ~2.3 million subscribers with
unlimited voice calling to over 60 countries
over its VoIP infrastructure.
Vonage was running predominantly on IP
phones over wireline connections, and
wanted to expand their offering in two new
trajectories:
Solution Type
Consumer VoIP
Target Audience
Consumers
Business Model
Subscription
Country
US
Twitter
@vonage
Media Types
Voice, Video
Use of WebRTC
Porting, repurposing
1. Mobile introducing a BYOD mobile application that can run on Android and iOS
devices, enabling Vonage customers to use their service anywhere
2. Video add video calling capabilities on top of their extensive voice service
To do that, Vonage took the following steps:
Looked at WebRTC as a pure open source technology stack

Adopted and ported the WebRTC code base to run on Android and iOS
Replaced the voice codecs used in WebRTC from G.711 and Opus to G.711, iLBC
and iSAC this removed the need to transcode media when calling between their
existing wireline service to their new mobile application. It has since then added
Opus support as well
Used VP8 video codec that comes with WebRTC to offer video calling between
their mobile applications
Added SIP signaling on top of WebRTC in their native application, to support the
same signaling of their existing wireline service
This approach enabled Vonage to meet their requirements with little change to their core
infrastructure. Today, the Vonage application has been downloaded over 1 million times
from the app stores.
Things to Remember
66
Purists will not see this as a WebRTC initiative, as it treats WebRTC as any other
generic media engine
Such a development project was rather expensive at the time Vonage started
(2012), but costs a lot less than it used to prior to WebRTC
The fact that WebRTC can be plagiarized is one of its unique values
Tsahi Levent-Levi
Sponsored by
Customer Service & Support

One of the areas where many see the value of WebRTC is customer service and support.
The main example given is usually the ability of a customer to surf a website, click a call
button on the website and get connected to an agent.
This being said, WebRTC has many different uses in the domain of customer service and
support some of which may prove to be a better fit for many enterprises.
Use cases in this section include:
Amazon Mayday Amazon's new video support service. While its use of
WebRTC isn't confirmed, it may change the way contact centers are going to be
structured
FreeCRM a CRM service that adopted WebRTC to enable calling contacts right
from within the CRM service, without the need to pick up a phone and dial
Intuitive Solutions a Texas managed contact center company that changed the
way they make decisions by using WebRTC
Vacasa Rentals a vacations provider using WebRTC internally to run their day
to day operation
67
WebRTC disruption in the contact center [UC Strategies]

Where Amazon Mayday's service is headed [Disruptive Wireless]
Challenges of WebRTC and customer service [WebRTC World]
What the Amazon Fire Phone Predictable Failure can Teach us
About WebRTC?
Why Invest in WebRTC When Everyone Goes Self Service? [UC
Strategies]
Tsahi Levent-Levi
Sponsored by
Amazon Mayday
Amazon Mayday is a service offered by
Amazon for free to its Kindle Fire tablet and
Fire Phone customers. With it, a customer
can press a Mayday button on his device
and get connected with a service agent
within 10 seconds.
A lot has been written about this service on
the internet. The most interesting aspects
of this service are:
Solution Type
Contact Center
Target Audience
Consumers
Business Model
Differentiation
Country
US
Twitter
@amazon
Media Types
Voice, Video
Use of WebRTC
Porting, repurposing
It is the first time Amazon provides

live agent support. Up until now they had only email support. They took the extreme
of offering live video support instead of voice only
Their target is having all calls answered within 15 seconds, something unheard of
it in today's contact centers. They achieved an average wait time of less than 10
seconds
The solution is targeted at their tablet and smartphone buyers only, as a
differentiating point from other mobile device vendors
The service supports one way video, two way voice, the ability of the agent to take
control of the tablet remotely and the ability of the user to share everything he is
doing on the tablet in real-time
While the technical details were never provided, it is safe to assume that such a solution
is based on WebRTC: Amazon's Kindle Fire tablets and Fire Phones are based on the
Android operating system. WebRTC is developed by Google with support for Android, so
having something readily available that can be used and integrated without paying for it
seems like a good bet.
Things to Remember
68
The deep integration done by Amazon with the ability to maintain a live call while
using the tablet and sharing the screen across applications isn't trivial. It was made
possible by the fact that Amazon controls the chipset, operating system and
services on their tablets
This move may have a large impact on contact centers, customer support and
mobile device vendors alike. Since Amazons announcements, many companies
(startups and established) have been introducing their own angle to the Amazon
Mayday button
Tsahi Levent-Levi
Sponsored by
FreeCRM
FreeCRM offers a CRM system in a
freemium model. One of its paid features is
the ability to call contacts from the CRM
directly.
The contacts don't need to be using
WebRTC in any way they receive their
call on the regular telephony service
(PSTN).
Solution Type
CRM
Target Audience
SMB, Enterprise
Business Model
Subscription
Country
US
Twitter
@FreeCRM
Media Types
Voice
In the past, FreeCRM has been integrated

Direct integration
with both Skype and Vonage, but recently Use of WebRTC
they have added the ability to use WebRTC. This provides a transparent experience to
the user, who does not need to have an account at any other service. It also makes it
easy to track the calls, their progress and even content and have that automatically
reflected in the CRM system.
FreeCRM have built their own solution based on the Asterisk communication framework,
connecting calls to the PSTN via a SIP gateway.
The service offers a voice only solution and is available on supporting browsers, which is
a legitimate approach for a CRM solution targeted at SMBs and enterprises.
Things to Remember
69
Seamless integration of calling capabilities into the contact center makes it easier
to handle and provides a single vendor solution
In this domain, PSTN connectivity is important
Such a solution can be achieved by self-development or by utilizing API vendors
Tsahi Levent-Levi
Sponsored by
Intuitive Solutions
Intuitive Solutions is a managed contact
center service located in Texas,
specializing in the pizza restaurant
business.
Intuitive Solutions was looking to grow, but
had several challenges:
1. Office space
2. VoIP software client for agents that
was hard to manage and maintain,
especially with remote workforce
3. 4 different vendors handling their
contact center
Solution Type
CRM, Contact Center
Target Audience
SMB, Enterprise
Business Model
Subscription
Country
US
Twitter
None
Media Types
Voice
Use of WebRTC
3rd party
Intuitive Solutions searched for a single vendor to offer all of their IT needs, be it CRM or
contact center related. They found such a third party platform that provided them CRM
capabilities along with the ability to receive incoming calls using WebRTC from within the
browser.
This solution offered Intuitive Solutions the following benefits:
Hassle free deployment and management of VoIP clients, as they are now a part
of the Chrome browser
Ability for contact center agents to work from home or a coffee shop, freeing them
from the brick and mortar office space, and allowing more flexibility in work hours
Better decision making due to a clear view of their day to day operations, as the
new vendor had integrated CRM and telephony reports
Things to Remember
70
Intuitive Solutions isnt a technology provider, and yet it sees itself as a company
directly benefiting from WebRTC
Fusing VoIP with the web enables tighter integration across software applications,
bringing with it better visibility and insights for the business
As companies mature and try to grow out of their niche, with CRM vendors win the
contact center market, or will VoIP PBX and contact center vendors win over the
CRM market?
Tsahi Levent-Levi
Sponsored by
Vacasa Rentals
Vacasa Rentals is a US based company
providing a vacation homes rental service.
They handle around 500 homes spread
across 6 different states in the US. They do
this by operating solely on the cloud using
SaaS services. Besides dealing with
physical laptops and smartphones, all the
rest runs off virtual services and that
includes their telephony system as well.
Solution Type
Contact Center
Target Audience
Internal consumption
Business Model
None
Country
US
Twitter
@vacasarentals
Media Types
Voice
rd
3 party
Vacasa Rentals debated as to the Use of WebRTC
approach to use between VoIP and softphones to WebRTC, and ended up using
WebRTC. This provided them more flexibility in how calls are routed to the relevant
Vacasa Rentals agents, who at the end of the day received the calls via WebRTC on their
browser.
Vacasa Rentals planned on doing the whole development work on their own, but ended
up using Plivo, one of the WebRTC API Platform vendors.
Things to Remember
71
This is a contact distribution solution, where incoming calls from PSTN get routed
via WebRTC to the relevant agent
While not a VoIP company or a contact center vendor, such use cases exist out
there vendors who do their business mainly via the phone and need intelligent
routing. They may end up developing it on their own, using a third party WebRTC
API Platform vendor, using a contact center solution or using traditional VoIP
Tsahi Levent-Levi
Sponsored by
Expert Marketplace
There is a market forming about expert services people with knowledge in a specific
domain who wish to share that information. Think of this market as the personal gym
trainer, only online and yes, there is more than a single vendor who offers this kind of a
personal gym trainer service online via WebRTC.
But the expert marketplace is larger than your typical gym, with experts from a large
variety of domain selling their time online. If you are a domain expert, then why not reach
out to customers via the web? While this can be achieved via Skype on the technical
level, these platforms offer the means to discover experts, understand their skills, see
ratings and comments of their customers, have the ability to interact with them, schedule
the session online and pay for the service all from within the same
website/webpage/service, with a single identity and without the hassle of installing any
software that has real value.
It isn't surprising then that there are WebRTC vendors in this market already, and two of
them are:
LiveNinja offering their solution on top of TokBox API platform

popexpert who developed a homegrown solution of their own
Apart from the two vendors detailed here, there are other vendors who offer marketplaces
for smaller niches.
72
Review of the experts market and its opportunities
Tsahi Levent-Levi
Sponsored by
LiveNinja
LiveNinja is one of the vendors vying into
the expert marketplace.
It started off running on top of TokBox
platform in its Flash days and shifting with
TokBox into WebRTC.
Solution Type
Experts market, Webinars
Target Audience
Soho
Business Model
Revenue sharing
LiveNinja offers a platform for experts of Country

US
any kind to host their service online from Twitter
@LiveNinjaDotCom
the creation of a profile page, to comments
Media Types
Voice, Video
and ranking capabilities, scheduling and
Via 3rd party
payment. All this is wrapped under a Use of WebRTC
discovery engine enabling users to find a suitable expert for their need.
The business model around expert support is based on revenue sharing, where the user
scheduling a session with an expert pays for the session and LiveNinja takes a
percentage of that amount as processing fees.
As the platform of LiveNinja evolved, so did their solutions and focus. Today, LiveNinja
offers a webinars platform, based on one-time fees for the session; along with OEM'ing
the platform for integration into other solutions.
A new and interesting service from LiveNinja called Katana offers a video contact center
that can be embedded into existing websites.
Things to Remember
73
While expert advice over the internet has existed prior to WebRTC, WebRTC offers
a simpler user experience where the whole business process is handled by a
single vendor, under a single roof, from a single website
What enables LiveNinja to move fast and offer new capabilities such as webinars
and events is their reliance on an API vendor, who provides these capabilities as
part of the service
For its Katana service offering, LiveNinja have opted for building its infrastructure
on its own, due to limitations of the API vendor it used in the features necessary in
Katana
Tsahi Levent-Levi
Sponsored by
Popexpert
popexpert is similar to LiveNinja in many
ways. It too offers a platform for experts to
build their profile pages, schedule and sell
their time online with the same type of
business model revenue sharing.
Solution Type
Experts market
Target Audience
Soho
Revenue sharing
Unlike LiveNinja, popexpert have taken the Business Model
US
route of developing the solution from the Country
@popexpert
ground up on their own including all of the Twitter
WebRTC related parts. While popexpert Media Types
Voice, Video
did start out by using an API vendor, they Use of WebRTC
Direct integration
switched along the way to their own
implementation this gave them an independence from the feature set (and bugs) of an
API vendor.
Things to Remember
74
An expert marketplace solution is about marketing and service. To that end, the
ability to customize the experience end to end is important. As with LiveNinja, a lot
of the attention of popexpert goes to the user experience domain
In the same industry, two different vendors choose different directions with
WebRTC some will go for tooling vendors for assistance while another would
prefer independence and self-development. The fact that getting started with
WebRTC is simple, these two options for every use case can almost always be
found
Tsahi Levent-Levi
Sponsored by
Telehealth
Telehealth is an interesting market segment when it comes to WebRTC. The challenge
with healthcare services is dealing with regulation and the liabilities involved in the
interactions created. US based vendors, for example, need to adhere to HIPAA
compliance, which stands for Health Insurance Portability and Accountability Act. HIPAA
includes directives around privacy of patients amongst other issues areas requiring
special attention and care from the vendors offering Telehealth solutions.
WebRTC is uniquely positioned as a very suitable technology. The reasons for that
include:
WebRTC is secured and private by default, whereas other systems can optionally
be made secure
WebRTC is suitable for easy incorporation into business processes and
Telehealth is all about business processes (from a regulatory point of view)
Development of Telehealth requires a lot of knowledge and experience of
healthcare something that is hard to come by in the VoIP domain, so having
something like WebRTC that makes it easy to develop VoIP use cases, makes it
possible to develop Telehealth services
The vendors discussed in this section provide different Telehealth solutions that are built
differently and have very different business processes:
Clrais Healthcare connecting seniors to their loved ones

Regroup Therapy provides single and group therapy sessions remotely
TruClinic provides solutions to everyone in the healthcare industry
75
How Regroup is bringing the shrinks couch online [VentureBeat]
Tsahi Levent-Levi
Sponsored by
Claris Healthcare
Clearis Healthcare develops and sells the
Claris Companion, an Android-based
tablet for seniors.
Their role is to enable seniors to increase
their connectivity with their loved ones.
This is achieved by offering a specialized
tablet device with a user interface and
applications capable of supporting specific
activities and needs of seniors.
Solution Type
Healthcare
Target Audience
Consumers
Business Model
Device selling,
subscription
Country
US
Twitter
@clarishealth
Media Types
Voice, Video
The device can communicate with other

Use of WebRTC
Via 3rd party TBD
family members by using WebRTC as the
technology of choice, making it suitable for smartphones, tablets and browser
communications.
Things to Remember
76
While not a healthcare service per se, the Claris Companion is targeted at an
audience with specific needs focused around healthcare
WebRTC and the rise of tablets enabled Claris Healthcare to offer this type of
service while focusing on the user experience
Claris Healthcare are combining a business model of selling a device with a
monthly subscription fee
Tsahi Levent-Levi
Sponsored by
Regroup Therapy
Regroup Therapy enables the creation of
personal therapy sessions or group
therapy sessions.
While the solution may seem similar to the
services provided by LiveNinja or
popexpert, it is actually quite different; the
reason being the need to handle the
healthcare regulation.
Solution Type
Healthcare
Target Audience
Vertical
Business Model
Revenue sharing
Country
US
Twitter
@RGTherapy
Media Types
Voice, Video
Therapists who wish to provide remote

Via 3rd party
sessions to patients, can leverage Use of WebRTC
Regroup Therapy and open up their account online. Regroup Therapy takes care of
onboarding these therapists and checking their credentials and ability to treat patients.
Those looking for therapy, can find online a therapist who fits their needs; or even visit a
therapist in his office and later on shift to doing online sessions from the comfort of his
home.
Regroup Therapy takes its share of the cost of any given session.
Things to Remember
77
Developing such a solution require as much knowledge and experience (or even
more) in group therapy as it does in communication technologies this knowledge
spans down from legal aspects, through bureaucratic/political aspects and up to
behavioral aspects
Healthcare comes in different shapes and sizes. Group therapy is one such
domain. Many other care related services can be provided by utilizing WebRTC
Tsahi Levent-Levi
Sponsored by
TruClinic
TruClinic provides a healthcare portal,
where patients can virtually visit doctors
online.
In the case of TruClinic, WebRTC plays the
role of providing video calling stitched into
specific business processes ones that
need to be secure and private due to the
regulations
surrounding
healthcare
services.
Solution Type
Healthcare
Target Audience
Vertical
Business Model
Revenue sharing
Country
US
Twitter
@truclinic
Media Types
Voice, Video
Use of WebRTC
3rd party
Unlike other verticals, where friction of

using the service needs to be eliminated, healthcare requires a certain degree of friction
the ability to register a patient, record his history and health status, and be able to
authenticate him properly.
Things to Remember
78
In the domain of healthcare, a lot of attention is provided to regulatory standards.

WebRTC reduces barriers of entry to video conferencing domain, which in turn
enables savvy healthcare players to enter this market and offer their expertise in
the administrative components of the system
Tsahi Levent-Levi
Sponsored by
Gaming
One of the most untapped markets when it comes to WebRTC is gaming. There is a lot
of chatter about the potential uses of WebRTC in this space, but very little action.
This is why this section covers more demos than actual games. It is here to provide an
overview of what capabilities of WebRTC are currently being researched and played
around with by developers, and where that may lead us looking forward.
The games and concepts in this section are:
BananaBread a demo of a first person shooter game using WebRTC's data

channel to achieve low latency networking
CubeSlam Google's game concept, showing what can be achieved with
WebRTC in a 3D game environment
Jocly online multiplayer board games, with a touch of WebRTC
Snake a simple demo using the web camera via WebRTC to control the snake
It is interesting to note that the traditional VoIP people are more passionate about
WebRTC in gaming than the gaming vendors at this point.
79
BananaBread game project page [Mozilla]

The technology behind CubeSlam [CubeSlam]
Jocly's use of WebRTC for immersive gaming [Jocly]
Game of Snake that uses WebRTC [Nicolas Beauvais]
Tsahi Levent-Levi
Sponsored by
BananaBread
BananaBread is a traditional 3D first
person shooter game ported to run purely
on HTML5. This was done using Java
Script and WebGL, where most of the initial
code got automatic translation from C++.
WebRTC was introduced in 2013 to
BananaBread. The data channel was
integrated into BananaBread offering low
latency network support for the exchange
of information between gamers in a multiplayer gaming session.
Solution Type
Gaming
Status
Demo
Use of WebRTC
Data channel
For multiplayer games, where low latency

is important to provide better feedback to the gamers, the ability to offer P2P connections
directly between the gamers can reduce latencies and also the load on the game server.
But it hasn't been tested to see how much can this approach scale to larger groups of
players, where multiple data sessions need to be opened by each browser to the rest of
the players.
80
Tsahi Levent-Levi
Sponsored by
CubeSlam
CubeSlam is Google's attempt
showcasing two capabilities:
at
1. The ability of HTML5 to run

games, showing off all of its
capabilities
2. The usefulness of WebRTC for
games
The game itself is the classic pong game,
skinned in 3D, colors and the added
video chat.
Solution Type
Gaming
Status
Demo
Use of WebRTC
Voice, Video and Data

channel
The interaction and data passed between the two players of the game is sent via WebRTC
data channel. Added to the game is a video chat session that enables both players to see
each other in real time. The video display is tilted with the board game as necessary using
WebGL.
The video component of this game isn't used for the game itself, but rather as a way to
get the players closer to each other and enhance the experience of the game.
81
Tsahi Levent-Levi
Sponsored by
Jocly
Jocly Games is a bootstrap company from
France, who are focused at delivering
strategy and board games to their
community of players.
The games are all HTML5 games, and
recently, WebRTC was added to the
experience offering the ability of people
to have a more immersive experience,
where they see each other during the
game itself.
While not all the gamers are using the
video chat option that WebRTC offers
there, it does seem to be gaining traction.
Solution Type
Gaming
Target Audience
Consumers
Business Model
None
Country
France
Twitter
@JoclyGames
Media Types
Voice, Video
Use of WebRTC
Direct integration
At the moment, Jocly Games have no monetization going around, but this may change
as they grow in audience. The most obvious choice here would be ads.
82
Tsahi Levent-Levi
Sponsored by
Snake
The game of snake can be considered as
an HTML5 demo written by Nicolas
Beauvais for no other reason than to
show it is possible to do.
It is the classing Snake game, where a
small snake needs to be navigated to eat
eggs and grow in the process. The game
ends when the Stake steps on his own
body.
Solution Type
Gaming
The only difference here besides running

Status
Demo
it on a web browser as opposed to a
Use of WebRTC
Get User Media only
monochrome monitor or an old Nokia
feature-phone is the ability to use the web
camera to control the movements of the Snake. This is achieved by using a Java Script
library called headtrackr which enables capturing the video with WebRTC
GetUserMedia and then process the images of the video to track the head.
While the end result is clunky and hard to use for this specific game, it might have
implications on user inputs in other WebRTC games especially considering the
capabilities introduced by Microsoft's Kinect camera and Leap Motion.
83
Tsahi Levent-Levi
Sponsored by
Social Networking and Consumer VoIP

I have decided to group consumer VoIP solutions (known also as communication OTT)
with social networks as both serve a similar purpose: enable people to communicate.
In this domain, vendors tend to provide free services with business models based on ads,
premium features or data monetization sometimes, they have no apparent business
model besides the need to grow to a certain size where they either get acquired or a new
business model presents itself.
I have selected two different vendors here, each focusing on a different target market:
MaxiAmigos a dating/social media service for Brazilians

Solaborate a business focused social platform
It is interesting to note, that most social network solutions prefer the route of direct
integration with WebRTC over the use of a 3rd party API vendor.
84
Social networks versus VoIP vendors

Comparing the size and interactions of Skype vs. Facebook
Monetization in the OTT VoIP world
Tsahi Levent-Levi
Sponsored by
MaxiAmigos
MaxiAmigos was the first dating/social
network to embrace WebRTC. They offer
an experience similar to Facebook, but
operate locally in Brazil, and in it, mostly in
So Paulo.
Solution Type
Social Network
Target Audience
Consumers
Freemium
MaxiAmigos offers an offline messaging Business Model
Brazil
service,
where
voice
and
video Country
communication comes secondary. They Twitter
n/a
first relied on a Flash based service, but Media Types
Voice, Video
once that service switched to a paid model,
Use of WebRTC
Direct integration
they opted for WebRTC. WebRTC in this
sense offered them more control and freedom over the solution.
MaxiAmigos started off integrating voice into the service, and recently they have added
video calling as well.
In terms of monetization, MaxiAmigos started off with an ads strategy and is planning to
move towards a different freemium model.
Things to Remember
85
While not a distinct dating service, this does show the potential. There are a couple
of dating services who are planning to introduce WebRTC in the near future
Switching from Flash to WebRTC offered MaxiAmigos the freedom they needed
and at the same time improved the quality and experience of the service they
provided to their users
Tsahi Levent-Levi
Sponsored by
Solaborate
Solaborate define themselves as the
missing link between Facebook,
LinkedIn and Yammer dedicated to
technology
professionals.
This
positions them as a niche social
network that caters the need of a
specific target audience.
Solaborate uses WebRTC to enable
voice and video calling both 1:1
calls as well as multipoint
conference calls.
Solution Type
Social Business
Target Audience
Soho
Business Model
None
Country
Macedonia
Twitter
@solaborate
Media Types
Voice, Video
Use of WebRTC
Direct integration
One of the interesting facts about Solaborate is that they opted to run in Microsoft Azure
cloud instead of other cloud providers. While this is different than how other vendors have
approached the backend challenge, it seems to work well for Solaborate.
Things to Remember
86
WebRTC enables embedding voice and video communication experiences into

other web services. In the case of Solaborate it enables enhancing social networks
There are different backend technologies that can work well with WebRTC. Pick
the one you are most comfortable with in terms of the engineers you have
Tsahi Levent-Levi
Sponsored by
Streaming and Content Delivery

One of the most interesting components in WebRTC is the data channel, which enables
sending arbitrary data directly between browsers. As such, it provides capabilities that
weren't possible before.
There are several strong capabilities the data channel brings to the table capabilities
that are already being exploited by vendors:
1. Ability to reduce load on a central server when multiple consumers are getting at
the same piece of data, they can assist each other instead of hammering the same
central server with the same requests
2. Ability to communicate securely as anything else in WebRTC, the data channel
itself is encrypted. This enables sending data securely from the clients
The two basic building blocks above can be used in various ways to reach different sets
of results each time like Lego pieces. In this section, I'll introduce three different
vendors, who are using the data channel's capabilities to achieve different results:
87
Peer5 a vendor specializing in peer assisted data delivery, focused on CDN and
file transfer
PeerMesh a P2P GIS solution
Swarmify providing CDN augmentation via the data channel
Tsahi Levent-Levi
Sponsored by
Peer5
Peer5 is an Israeli startup that has focused
from the start on the data channel. The
entrepreneurs behind Peer5 came with
web experience and knowledge and no
background in VoIP.
What they found in WebRTC was the
ability to connect browsers directly, and
they have built on top of that basic function
the ability for multiple browsers to
cooperate directly amongst themselves
when trying to access similar content.
Solution Type
Distributed CDN
Target Audience
SMB, Enterprise
Business Model
Pay as you go
Country
Israel
Twitter
@peer5
Media Types
Data Channels
Use of WebRTC
Direct integration
That capability can be used to reduce the load from video streaming servers, which is
what Peer5 does; and it can be used to share files online, which is what Sharefest,
another service by Peer5 does.
The advantages of using such peer assisted technologies are:
88
Reduce load on the servers who hold the original content

Reduce bandwidth requirements from the server farm
The more people try to access the same content at the same time the better the
system is capable of reducing the load on it
Tsahi Levent-Levi
Sponsored by
PeerMesh
PeerMesh is
technology.
peer-to-peer
GIS
GIS, which stands for Geographic

Information Systems, is designed to
handle geographical data (aka Maps). One
of the challenges with GIS is the need to
send large amounts of data from the server
towards the clients of a GIS service.
PeerMesh comes to solve that by using
WebRTCs data channel similar to how
Peer5 does, but their main focus is GIS.
Solution Type
Distributed CDN
Target Audience
Vertical
Business Model
Unknown
Country
Turkey
Twitter
None
Media Types
Data Channels
Use of WebRTC
Direct integration
The service has been created by a developer and a UX designer living in Turkey and
working for a vendor in the defense industry. While they have no solid commercial plans
for the service at the moment, it shows how WebRTC can be used to solve challenges in
a large range of domains.
Things to Remember
89
A relatively small team can make use of WebRTC to solve a challenging problem
WebRTC isnt limited to voice and video communications. There are many areas
in which it can be suitable via its data channel
The use of WebRTC is worldwide phenomena. Developers the world over are
tinkering with it at any given time
Tsahi Levent-Levi
Sponsored by
Swarmify
Swarmify is a service that provides similar
capabilities to those offered by Peer5: it
enables multiple web browsers accessing
the same content to share pieces of the
content directly between each other,
similar to how BitTorrent works.
Solution Type
Distributed CDN
Target Audience
SMB, Enterprise
Business Model
Subscription
What makes Swarmify interesting is the Country

simplicity they provide: by adding a piece Twitter
of JavaScript code they provide to web
Media Types
pages, and from that moment and on, the
Use of WebRTC
website "takes care of it" by letting page
viewers share content pieces of the site automatically.
US
@SwarmCDN
Data Channels
Direct integration
On the business model front, Swarmify offers just as simple tiered price plan, with
customized plans for large enterprises. The idea behind it is the reduction of the CDN
costs paid by websites.
The advantages of using this scheme:
90
This can be used (and should be used) in parallel to running a CDN

When possible, Swarmify will kick in and just offload the work from the CDN. This
saves on cost but may also reduce latency
Even with Chrome and Firefox alone supporting WebRTC, this kind of a service
brings great value
Tsahi Levent-Levi
Sponsored by
Other
There are many more stories and use cases that didn't make it to this research paper.
Some because they bear similarities to other stories here, some because of breadth and
time while others still because of my knowledge on them.
In this section, I want to introduce a few more use cases ones that didn't fit elsewhere.
These are:
91
fresh tilled soil a web design house from Massachusetts

OpenVRI a single person's pet project for the visually impaired
WebRTC School that offers certification programs for WebRTC
Tsahi Levent-Levi
Sponsored by
fresh tilled soil

fresh tilled soil is a UX design house. They
deal with web and mobile apps designing
and implementing them for customers.
One might as what does such a company
has to do with interactive video calling
capabilities, but the answer is quite
obvious WebRTC is an HTML5
capability, so using it and harnessing it for
customer projects is something that web
companies will probably start doing in the
near future.
Solution Type
Web development
Target Audience
Enterprise, SMB
Business Model
Project based
Country
US
Twitter
@freshtilledsoil
Media Types
Voice, Video
Use of WebRTC
Direct integration
fresh tilled soil started dabbling into WebRTC without a specific customer. In their case,
it meant being able to understand the technology and showcase it to potential future
customers that wanted something special for their next web project.
They ended up releasing their own WebRTC widget to the world as part of this pet project.
Things to Remember
92
fresh tilled soil is at the forefront of such development outfits. Others will follow if
customer demand for such capabilities start piling up
The challenge for such vendors will be in the deployment of such a service, which
requires a bit more handholding and maintenance than the usual web project. They
will probably end up leaning towards one of the tooling vendors that offer SaaS
models to deliver projects to their customers
Tsahi Levent-Levi
Sponsored by
OpenVRI
VRI stands for Video Remote Interpreting.
VRI enables two people who are located in
the same room to communicate when they
don't share the same language or when
one of them is hard of hearing this is
achieved by opening a video call to another
person, who knows languages of both of
the people or knows sign language and
can interpret a deaf person.
VRI solutions are usually expensive, as
they are provided by the same devices and
services of enterprise video conferencing
solutions.
Solution Type
Distributed CDN
Target Audience
Vertical (VRI)
Business Model
None
Country
US
Twitter
None
Media Types
Voice, Video
Use of WebRTC
Direct integration
OpenVRI is the result of a single person, hard of hearing himself, who decided to offer a
similar solution based on WeRTC and make it freely available to anyone both as a
service and in source code form.
OpenVRI is different in many aspects from both other VRI solutions as well as WebRTC
video calling solutions:
93
It has been built by a single person, hard of hearing himself this isn't an easy
feat, especially considering the fact that it includes a voice channel
It is free in every sense of the word. The service is hosted online and is accessible
and usable freely, while the source code of the service can be forked and modified
via github
It has been developed for the personal use of the developer he needed such a
service, so he just created it
The service is based around the concept of ad-hoc meetings. It generates a
hashed, one-time ink that can be emailed to the relevant parties that need to be
on the same session
There is a text chat window for each participant, as opposed to a single timeline
for all participants. This makes more sense to this type of a service, where people
would like to see the text as soon as it is typed, and not necessarily at its entirety
at times, this is the main channel of conversation
The developer of OpenVRI has since then joined a VRI company, where he
continues to develop WebRTC based solution for the hard of hearing in an
enterprise scale
Tsahi Levent-Levi
Sponsored by
WebRTC School
WebRTC School isn't a WebRTC vendor in
the "classic" sense. It doesn't develop
anything that is WebRTC-based, but rather
provides
training
services
around
WebRTC.
Solution Type
Learning
Target Audience
Developers
WebRTC School is the work of Vocale, the

Business Model
Subscription
vendor who also runs the SIP School. They
have decided to tackle WebRTC and provide much needed training and certification
programs.
The WebRTC School provides two separate courses, one targeted at developers and the
other at IT people who need to integrate and maintain WebRTC services.
WebRTC School holds a supportive role in the ecosystem. There is room for more
vendors and services to take similar roles and provide services not directly related to
WebRTC, but ones that assist vendors in reaching their goals.
94
Tsahi Levent-Levi
Sponsored by
The Missing Verticals

The list of use cases and verticals is by no means exhaustive. There are many more
verticals that weren't covered here. Some of them traditional, such as telecom vendors,
who provide SBC, gateways and other equipment to carriers and enterprises; and others,
are markets of their own, such as education, banking, etc.
The use cases were selected because each tells a story usually different than the ones
that existed prior to WebRTC.
The ecosystem of WebRTC today covers around 500 vendors and services, many with
no clear business model. In the coming years, the ecosystem will go through transition
where some vendors will disappear, new ones will appear, others will grow and others
still will merge and consolidate.
In some ways, WebRTC has been relegated into the background already, with services
using it without announcing that fact at all. This shift indicates a more focused view on
delivering the use case than on the technology.
95
Tsahi Levent-Levi
Sponsored by
Recommendations
This chapter can be seen as a shortlist of recommendations of the various insights found
throughout this research paper. They are general guidelines to those who wish to build
services that use WebRTC.
1
2
Start from a business case

Identify a challenge or a problem you wish to tackle, preferably with a
pain-point applicable to a specific market niche
Understand who will be willing to pay for what to solve this problem
Validate that WebRTC can cover your needs

Make sure it is available on the web browsers that your target
audience uses
Decide what your fallback strategy to browsers that don't support
WebRTC is (options are: none, Flash, plugin)
Dont forget to decide on how you wish your service to run on mobile
devices: which ones, in an app or on the web, etc.
Analyze the ability to use existing tooling solutions

Are you comfortable with outsourcing the work for external vendors
that specialize in it?
Can you use an API vendor to deal with the whole WebRTC part?
Should you self-host everything or use PaaS offerings for things like
NAT traversal or signaling?
Are there open source modules available that you can integrate into
your solution?
96
Research the competition

There are many WebRTC vendors out there providing solutions. Make
sure you know who provides similar capabilities to what you plan
today
Google to find who they are, check on WebRTC blogs and news
outlets to see who is out there
Play a bit with WebRTC services online see how the interactions
feel like, what have people done with it
Tsahi Levent-Levi
97
Sponsored by
Select technologies based on existing skillsets

What kind of backend do you need for your service?
Which frontend capabilities are require?
Do you have the resources in your company to handle this?
Be sure to have a web developer on board
Depending on the use case, you might want to consider having a
veteran VoIP engineer or architect on board
Tsahi Levent-Levi
Sponsored by
Appendix A: Online Resources

This appendix provides some links to other online resources out there, where the main
focus isn't a technical one. If you are looking for thought leadership, business insights or
market related information about WebRTC, then this can serve as an initial reference list.
Blogs
There are many blogs who are now covering WebRTC; some of them are large news
outlets while others bring deep thinking as well as user stories.
Below is a hand-picked list of such blogs. You can consider them mandatory reading list
when it comes to understanding WebRTC.
Direct links to the blogs outlined here
BlogGeek.me
My very own blog, where I write a lot about WebRTC and try also to give room from
experts by way of guest posts.
WebRTC Weekly
A weekly newsletter providing roundup of posts and articles from around the web about
WebRTC. Curated by Chris Koehncke and me.
WebRTC Glossary
A glossary site for all terms related to WebRTC.
webrtcHacks
This is a blog site operated by Chad Hart, Reid Stidolph, Victor Pascual vila and Tsahi
Levent-Levi. While dealing more with the technical aspects of WebRTC, it does offer
business insights as well as providing an events directory for WebRTC.
98
Tsahi Levent-Levi
Sponsored by
Chris Kranky
Chris Koehncke's personal blog is a great source of information about WebRTC. He is
opinionated, to the point and blunt. He also writes with a unique voice that is enjoyable to
read every time.
Alan Quayle
Alan Quayle, an independent consultant in the telecommunication space. His focus
around Telco APIs and enablers to these have brought him to look closely at WebRTC
as well.
Disruptive Analysis
Disruptive Analysis is Dean Bubley's blog. He is a consultant in the telecommunication
space that is looking at WebRTC as well.
He has his own research paper on WebRTC.
WebRTC World
WebRTC World is a dedicated portal for WebRTC created and maintained by TMCnet. It
is the place to go for up to date information about vendors and their plans and
announcements around WebRTC.
99
Tsahi Levent-Levi
Sponsored by
Communities
There are several online communities where conversations around WebRTC happen.
The ones here are those that I find the most informative and engaging.
Direct links to the communities outlined here
WebRTC on Google+
Google+ has two distinct groups for WebRTC. There is no difference in terms of the type
of content between them. People in this group are very responsive to questions asked.
WebRTC Forum on Facebook

There is a lively forum around WebRTC that has formed in Facebook.
The main news items related to WebRTC are raised and discussed within this group,
which has just passed 300 members.
Following WebRTC on Quora

Quora has a WebRTC tag, where many questions and answers can be found.
It can be a good place to ask your own business related questions.
Meetup Events
At the time of writing, there were 18 different WebRTC related meetup groups around the
world.
These are localized groups that conduct local events about WebRTC. The target
audience as well as the content of each meeting vary widely, but if there is a meetup
located close by to where you live, join it to gauge the local community that is hacking its
way around WebRTC.
100
Tsahi Levent-Levi
Sponsored by
Appendix B: Choosing a WebRTC API Platform report

If you are interested in developing using WebRTC then you might consider my other
report: Choosing a WebRTC API Platform
This report is intended in assisting you in making an informed decision on how to develop
your next WebRTC-based use case. With it you will be making a more educated decision
about which WebRTC API platform to use at the end of the day.
Download the reports introduction to learn more.
Spanning 131 pages, this paper covers in depth what WebRTC is, the challenges of using
WebRTC, the different ways developers tackle WebRTC, and the vendor selection
process that needs to be taken when choosing a WebRTC API Platform.
The report also details information about 16 different WebRTC API Platforms, to help you
get started faster in your selection process.
Who is this report for?
This report is useful if you are:
An entrepreneur, seeking to start a real-time communication related initiative
A product manager, trying to understand the options in front of you
A developer, needing to decide which route to take with WebRTC
Key questions addressed

The research paper addresses the following key questions:
What is WebRTC?
What are the challenges in developing with WebRTC?
What are the available options to WebRTC development?
What are the KPIs when selecting an API platform for WebRTC?
What API platforms are there and how do they fit your needs?
101
Tsahi Levent-Levi
Sponsored by
Appendix C: Finding out More

Other than this report, the following services are available:
BlogGeek.Me Blog
My blog is the place where the bulk of my writing exists. It is available freely and can be
subscribed via both RSS and email. See http://bloggeek.me
WebRTCweekly Weekly Newsletter

The WebRTCweekly is a weekly newsletter. It lists articles from around the web about
WebRTC from the past week. This service is offered freely and is curated by me and
Chris Koehncke. See http://webrtcweekly.com/
BlogGeek.Me Monthly Newsletter

I am publishing a short monthly newsletter. It includes an answer to a question I was
asked during the month, statistics out of the research on WebRTC that I am doing, an
interesting project and a few other tidbits. See http://bloggeek.me/newsletters/
Reports
This is my second report about WebRTC. To find out about other paid reports I have
written, see http://bloggeek.me/reports/
Consulting
I provide consulting services to vendors, especially around VoIP, video conferencing
and WebRTC. See http://bloggeek.me/consulting/
102
Tsahi Levent-Levi
Sponsored by
Appendix D: About Kandy

Kandy is GENBANDs Real Time Communications Platform as a Service (PaaS). Kandy
enables service providers, integrators, partners and developers to enrich their
applications and services with real time contextual communications. By providing
developer-friendly access to the entire suite of GENBAND real-time communication
toolbox in the cloud. Kandy allows anyone from IT to business owners to developers to
embed video, voice, text, presence and collaboration into existing and new Web and
Mobile applications in an aggressive time to market and within a cost effective
subscription business model.
Real Time Communications enables businesses to interact with their customers, and
employees to interact with one another, by seamlessly integrating live and immediate
communication, improving efficiency and user satisfaction.
Kandy offers developers a sandbox and community, app showcase and innovative
QuickStarts, APIs and SDKs for Web and mobile.
Kandy is available at kandy.io. Come build with Kandy today.
About GENBAND
GENBAND is a global leader in smart networking solutions for service providers and
enterprises in more than 80 countries. From the Core to the Edge to the Experience,
the companys technology seamlessly evolves IP networks to new levels in scalability,
security, profitability and efficiency. GENBANDs market-leading technology facilitates
multimedia voice, data and video sessions and anywhere and any device services that
scale on public and private networks. GENBAND is headquartered in Frisco, Texas, and
has R&D, sales and support resources in more than 50 countries.
103
Tsahi Levent-Levi
Sponsored by
Appendix E: Kandy Usage Scenarios

With Kandy you can quickly integrate communications into your existing applications.
Concierge provide a premium service to your customers with the ability to video chat,
share screens, recommend products, and more. See http://concierge.kandy.io/
Collaboration - improve the efficiency of your communications by providing real time

sharing of documents, images, videos, and more. See http://collaboration.kandy.io/
104
Tsahi Levent-Levi
Sponsored by
Business Assistant - Keep work and home life separate with your own business
number you can use for voice, texting, and voicemail.
See http://assistant.kandy.io/
105

WP WebRTCForBusinessPeople Kandy 1114

Caricato da

Informazioni sul documento

Copyright

Formati disponibili

Condividi questo documento

Condividi o incorpora il documento

Opzioni di condivisione

Hai trovato utile questo documento?

Questo contenuto è inappropriato?

Copyright:

Formati disponibili

WP WebRTCForBusinessPeople Kandy 1114

Caricato da

Copyright:

Formati disponibili

WebRTC

for Business People

WebRTC for Business People

WebRTC for Business People

Firewall and NAT Traversal .....................................................................................................................................28

WebRTC for Business People

WebRTC for Business People

WebRTC for Business People

Why WebRTC for Business People?

WebRTC for Business People

Why is this Report Open?

Id like to thank the Kandy team for sponsoring this report.

WebRTC for Business People

Further Reading: 001 bloggeek.me/ref/biz001/

The history of WebRTC inside Google [Quora]

Two important aspects of WebRTC:

WebRTC for Business People

1. WebRTC the Specification

2. WebRTC the Open Source Project

WebRTC for Business People

Further Reading: 002 bloggeek.me/ref/biz002/

WebRTC's Job to be done [UC Strategies]

Porting it to operating systems and environments where similar media processing

WebRTC for Business People

Further Reading: 003 bloggeek.me/ref/biz003/

WebRTC is for web developers and not VoIP developers [NoJitter]

WebRTC for Business People

ORTC and WebRTC

WebRTC for Business People

WebRTCs Job to be Done

Barriers of Entry for New Vendors

WebRTC for Business People

Reducing End User Friction

The Innovators Dilemma and WebRTC

WebRTC for Business People

WebRTC and the Innovators Dilemma

In our case, WebRTC is the disruptive technology.

It targets a different type and set of developers

Further Reading: 004 bloggeek.me/ref/biz004/

A WebRTC JTBD (Job to be Done) Weekend

WebRTC for Business People

May 2011: WebRTC Announced

Jan 2013: Firefox 20 adds WebRTC

Nov 2011: Chrome 23 adds WebRTC

Jun 2013: Firefox 22 released

Jul 2013: Chrome for Android Beta with WebRTC

WebRTC for Business People

Oct 2013: Opera 18 Beta intros WebRTC

Mar 2014: Opera for Android with WebRTC

May 2014: Microsoft promises to support GUM

Figure 2: Major milestones of WebRTC support in browsers

Get User Media

Table 1: Browsers support for WebRTC

The table indicates two important aspects of WebRTC adoption:

WebRTC for Business People

Further Reading: 005 bloggeek.me/ref/biz005/

Up to date browser support of WebRTC [&yet]

Interoperability between Chrome and Firefox

WebRTC for Business People

Handling Safari and IE

Further Reading: 006 bloggeek.me/ref/biz006/

How to get support into old browsers? [Disruptive Analysis]

WebRTC for Business People

Dealing with Mobile