Sei sulla pagina 1di 33

Building Sustainable High-Growth Startup Companies: Management Systems as an Accelerator Author(s): Antonio Davila, George Foster, Ning Jia

Source: California Management Review, Vol. 52, No. 3 (Spring 2010), pp. 79-105 Published by: University of California Press Stable URL: http://www.jstor.org/stable/10.1525/cmr.2010.52.3.79 . Accessed: 03/10/2011 05:41
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

University of California Press is collaborating with JSTOR to digitize, preserve and extend access to California Management Review.

http://www.jstor.org

S p r i n g

2 0 1 0

V o l . 5 2 ,

N o . 3

R E P R I N T

S E R I E S

California Management
Review
Building Sustainable High-Growth Startup Companies: Management Systems as an Accelerator
Antonio Davila George Foster Ning Jia

2010 by The Regents of the University of California

Building Sustainable High-Growth Startup Companies:


MANAGEMENT SYSTEMS AS AN ACCELERATOR
Antonio Davila George Foster Ning Jia

any companies founded with enthusiasm and hope for high growth fail to meet their founders and investors expectations. Based on our research over the past decade (see Appendix A), one explanation for this underperformance is the entrepreneurs resistance to switching to a more structured management approach and adopting management systems and processes in a timely way. Management systems are technically dened in the literature as: formal, information-based routines and procedures managers use to maintain or alter patterns in organizational activities.1 A small company can be run on a personal basis in its inaugural days. The founder-CEO typically wears multiple hats and controls all aspects of her workplace.2 She can observe everything that goes on in the organization and makes all the important decisions. The corporate mission and chosen strategy can be effectively conveyed and reinforced through direct communication with each employee. As the company grows, however, this management style inevitably fails. The combination of growth and a personal management style can be deadly. The manager of one of the companies in our research provided a rule of thumb to the limits of a personal management style: We had management by personality and it became evident that that wouldnt scale. We gured our personalities can go through one oor and two walls. After that management by personality doesnt work anymore. Growth requires a drastic change in how the company is run. A manager in a telecom service startup company described this situation as follows:
Looking back at it, and comparing it to other companies Ive been working for, I think the toughest step for a company to go through is going from a small company where pretty much everyone knows the main goals, the main focus, and the main initiatives within the company to the point where you stop communicating the whole business and some people will know some and some people will

CALIFORNIA MANAGEMENT REVIEW VOL. 52, NO. 3

SPRING 2010

CMR.BERKELEY.EDU

79

Building Sustainable High-Growth Startup Companies: Management Systems as an Accelerator

know others, but not everyone sees the big picture. And I think its a very crucial threshold and its very easy to start having a lot of people running in many different directions. The difculty at that stage is to keep the whole energy focused. Its like moving from a two-person boat where its easy to make sure you paddle in the same rhythm, to a 10-person boat where suddenly you have to spend more time on making sure its the same rhythm for everyone.

Yet, alternative views argue that management systems are detrimental to startup companies and innovation more broadly.3 These systems are quickly equated to a bureaucracy4 that kills the entrepreneurial spirit that characterizes such companies. The argument is that if a startup company wants growth, it needs to relax these managerial tools. Otherwise, they stie the company. Management systems, if left on their own, grow into bureaucracy; and bureaucracy, as we commonly interpret it, is associated with wasting resources to comply with business processes that have little if any value. The question is: Are management systems positively associated with growth or do they constrain it? A classic business article identied the entrepreneurial crisis when the company reaches a certain scale and moves to a higher growth stage and the entrepreneur has to transition into becoming a manager:
Increased number of employees cannot be managed exclusively through informal communication. . . . The company founders nd themselves burdened with unwanted management responsibilities. . . . Founders often resist stepping aside, even though they are probably temperamentally unsuited to the job. So here is the rst critical choice in an organizations development: to locate and install a strong business manager.5

Ciscos early years (between 1984 and 1990) provide an excellent illustration of the tensions between a personal and a more-structured professional approach to management. In Ciscos case, the tensions arose when a new Antonio Davila is a Professor of entrepreneurship CEO was appointed by the board to at IESE Business School. <adavila@iese.edu> professionalize management systems George Foster is the Wattis Professor of and processes and encountered founder Management and Dhirubhai Ambani Faculty Fellow in Entrepreneurship at the Graduate pushback. Two CEOs faced this pushback School of Business, Stanford University. before the founders exited the com<gfoster@stanford.edu> pany. Cisco Systems, Inc. was founded Ning Jia is an Assistant Professor of accounting in December 1984 by Len Bosack (who at the School of Economics and Management, became the rst president) and his wife Tsinghua University. <jian@sem.tsinghua.edu.cn> Sandy Lerner (who became the rst CFO). The following quotes are taken from a Stanford Graduate School of Business case:6
The Early Years: 1984-1987 . . . The early company organization was unstructured informal . . . As Cisco evolved . . . Bosack was increasingly uncomfortable and dissatised with the day-to-day chores of managing a company. By January of 1987 Ciscos management team had reached the consensus opinion: the com-

80

UNIVERSITY OF CALIFORNIA, BERKELEY VOL. 52, NO. 3

SPRING 2010

CMR.BERKELEY.EDU

Building Sustainable High-Growth Startup Companies: Management Systems as an Accelerator

pany needed a more formalized, experienced management structure. As a result, Bosack stepped down as president of the company.7

In 1988, Bill Graves, an executive at Schlumberger, was brought in as President and CEO. Graves led the securing of the rst round of venture capital from Sequoia Capital in December 1987. Don Valentine of Sequoia became a director of Cisco.
Valentine Initiates ChangeGraves started to have interpersonal problems with the founders, Sandy Lerner in particular. Lerner had a very forceful, often difcult personality, with very strong ideas about how the company should grow and evolve. As Graves had assumed more operating control of the company, the two personalities had clashed, with Graves more and more frequently overruling Lerners opinions.8

In April 1988, Graves resigned from his position as CEO. Valentine led the search and settled on John Morgridge as the preferred CEO candidate:
As part of the screening process, Morgridge met with all of the senior managers and the two founders . . . With Sandy in particular, Morgridge met and interviewed four separate times. Morgridges goal was a modest one: to get Sandy to a point where she was at least neutral to Morgridges presence and role in the company. By October 1988, Morgridge had been marginally successful. He joined Cisco as the companys new president and CEO . . . [He] began by tackling the issue of organizational structure and professional management . . . Although manufacturing was Sandys area of responsibility, the group was suffering from a combination of Sandys inexperience and the lack of overall priority that it had been given within the company.9

From October 1988 onwards, Morgridge brought in a new disciplined professionally focused management team. By December 1989, it had 174 employees. Its revenue and protability growth from 1986 to 1989 (July 31 scal year) were: 1986 Net revenues
(in $millions) $0.129 $1.485 $5.450 $27.664

1987

1988

1989

Net Income
(in $millions) $(0.833) $0.083 $0.388 $4.178

Cisco went public on February 16, 1990. Tensions continued both between Lerner and Morgridge and between Lerner and the new management team. In August 1990, Lerner was red. Bosack resigned very quickly thereafter. Could Cisco have reached its current prominence without Morgridge professionalizing the management and instilling formal systems into the company in its startup days? Prior studies have pointed out two negative outcomes resulting from failure to deal with the entrepreneurial crisis. The rst one is capping the company at a size that can be managed through a personal style and

CALIFORNIA MANAGEMENT REVIEW VOL. 52, NO. 3

SPRING 2010

CMR.BERKELEY.EDU

81

Building Sustainable High-Growth Startup Companies: Management Systems as an Accelerator

much below its business potential.10 The second one is management chaos.11 A high-growth startup in the medical devices space saw how its underdeveloped management systems led to problems with the FDA that sent them a letter with more than 150 quality issues reecting the chaos throughout the operations of the company. The management team quickly upgraded its systems to meet the increased complexity and kept its job. Serial entrepreneurs, who have experienced this entrepreneurial crisis before, prepare in advance to navigate smoothly through the transition. Salesforce.com, an enterprise software company, is a case in point. Early in its life, the company set up the management infrastructure to coordinate and integrate learning that would facilitate growth. This focus on building the infrastructure was not unrelated to its founders experience in both large companies such Oracle and their involvement in other startups.12 To examine the relevance of management systems for growth, we undertook a study of 78 high-growth startups in California termed the SEMAS project (Stanford Entrepreneurial MAnagement Systems).13 This project complemented our interactions with numerous startup managers through teaching, case writing, and additional research projects in this eld. Our purpose was to understand the transition point around the entrepreneurial crisis. In particular, we wanted to understand whether management systems were associated with the growth of these companies. High-growth startup companies may fail because the business model that was the basis of their strategy does not materialize. They may also fail because the market does not open up in the way and time anticipated. Such negative outcomes come with the territory of startup companies. High risk taking, often with radical innovations, is more often rewarded with failure than success. However, startups often fail or do not achieve their full potential because they get stuck in this entrepreneurial crisis.14 The equation is simple: as you grow bigger, its more efcient to use management tools. However, being simple does not mean that it is easy to do. Nor is it an activity that senior executives often relish. A manager with large company experience described the evolution of his company: As we were getting bigger, we had more professional managers. We could afford more professional managers and not do it all ourselves as we grew. In other words, growth requires a certain dose of discipline associated with formal systems and processes. These systems provide managers timely and accurate information for decision making as well as facilitate coordination, resource allocation, motivation, and performance measurement within an increasingly complex organization. Our ndings are consistent with Ciscos examplegrowth as well as delivering on the promises of an attractive business model needs management systems. These systems appear as necessary albeit not sufcient for growth.15 Governments that rightly believe in startup companies as a way to make their regional economies more dynamic often fail to take into account this entrepreneurial crisis and the need to transition into building the management infrastructure. They focus their resources on supporting entrepreneurs starting

82

UNIVERSITY OF CALIFORNIA, BERKELEY VOL. 52, NO. 3

SPRING 2010

CMR.BERKELEY.EDU

Building Sustainable High-Growth Startup Companies: Management Systems as an Accelerator

new companies but often ignore the need to support the transition through this stage. The outcome of such policies is lots of small companies that do not grow beyond a few dozen employees even if their business potential might be much higher; companies where the entrepreneur has stopped growth in order to keep the company under control. This outcome is one that government ofcials in several countries (such as Ireland) are seeking to reduce. It is possible that many startups could grow more if more attention was paid to making entrepreneurs aware of the need to implement systems that support such growth.

The Relevance of Management Systems for Growth


The evidence in our study supports the association between growth and the presence of management systems. We do not claim that these systems lead to growth. Growth depends on a host of other critical variables such as the viability of the business model and the industry the rm is in. However, management systems do provide the infrastructure that is associated with the likelihood of a venture safely transitioning through the entrepreneurial crisis. In many cases, management systems trail needs.16 Because of the demands on their limited resources, companies hardly ever develop systems ahead of their needs.17 Rather systems in place are either about right or, more often, are too primitive for the complexity that has to be managed. The learning point here is that in most startup companies adding more systems wont stie the entrepreneurial spirit but will sustain growth. Having no systems (chaos) is as damaging to a company as having too many (bureaucracy); and startup companies more often suffer from the former rather than the latter. Their managers tend to worry about avoiding bureaucracy but are blind to the danger of chaos. Figure 1 illustrates the build-up of management systems. In particular, we collected data from our sample companies on 46 different management systems clustered around eight different categories: nancial planning, nancial evaluation, human resource planning, human resources evaluation, strategic planning, product development, sales and marketing, and partnerships. Table 1 lists the 46 individual systems for the eight categories. For each one of these 46 systems, we asked when the company had adopted them. Figure 1 presents the mean adoption rate of the two most frequently adopted individual systems in each category in each year. There is marked evidence of management systems buildup in years 1 to 5 across each of the eight categories. Within our sample, there is also much variation across our companies in their speed of adopting different management systems. This variation enables us to probe the association between the intensiveness of management system adoption over time and company performance. Given our sample of companies, headcount is the most widely available and meaningful variable to assess performance. Most of the companies in the sample have undergone multiple rounds of venture funding where new and existing investors pass judgment on the viability of the company and its future potential to protably grow. For the subset of cases where both headcount and

CALIFORNIA MANAGEMENT REVIEW VOL. 52, NO. 3

SPRING 2010

CMR.BERKELEY.EDU

83

Building Sustainable High-Growth Startup Companies: Management Systems as an Accelerator

FIGURE 1. Rate of Adoption of Eight Types of Management Systems by Early-Stage Growth


Oriented Companies in Years 1 to 5

Financial Planning 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0%
Years

Financial Evaluation 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 2 3 4 5
Years 1

Human Resource Planning

Years 1

Human Resource Evaluation 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0%
Years

Strategic Planning 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 2 3 4 5

Product Development Management

Years 1

Years 1

Sales/Marketing Management 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0%
Years 1

Partnership Management 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0%

Years 1

Note: Graphs are formed by averaging the top two most frequently adopted systems within each category in each year.

valuation at each funding round is available, the correlation is both positive and signicant. Protability of venture capital-backed companies in their early years is typically not a good short-run success measure. The rationale for needing venture capital arises from cash outows exceeding cash inows in the early years as investments for growth are made. For many of these companies there is

84

UNIVERSITY OF CALIFORNIA, BERKELEY VOL. 52, NO. 3

SPRING 2010

CMR.BERKELEY.EDU

Building Sustainable High-Growth Startup Companies: Management Systems as an Accelerator

TABLE 1. Management Systems Examined in the SEMAS Research


Category of System
Financial Planning

Individual Systems Within Category


Operating budget Cash ow projections Sales projections

Financial Evaluation

Capital investment approval procedures Operating expenses approval procedures Routine analysis of nancial performance against target Product protability analysis Customer protability analysis Customer acquisition costs analysis

Human Resource Planning

Core values Mission statement Organizational chart Codes of conduct Written job descriptions Orientation program from new employees Company-wide newsletter

Human Resource Evaluation

Written performance objectives for managers Written performance evaluation reports Linking compensation to performance Individual incentive programs

Strategic Planning

Denition of strategic (non-nancial) milestones Customer development plan (plan to develop market) Headcount/human capital development plan Product portfolio plan (plan about future products) Investment budget continued on next page

an inverse relationship between protability and valuation at successive rounds of nancing.18 Figure 2 illustrates the differential growth rate of high-growth startup companies with different levels of management systems intensity.19 Systems intensity for each company over time is estimated by adding the number of systems adopted each year.20 We grouped our 78 companies into three groups according to systems intensity in year two and looked at their growth pattern as reported in Figure 2. The pattern is clear cut. Those companies with higher management systems intensity also grew much faster than their counterparts. By year ve, the average size of a company in the highest intensity group was about 135 people compared to barely 43 people for companies in the lowest

CALIFORNIA MANAGEMENT REVIEW VOL. 52, NO. 3

SPRING 2010

CMR.BERKELEY.EDU

85

Building Sustainable High-Growth Startup Companies: Management Systems as an Accelerator

TABLE 1. Management Systems Examined in the SEMAS Research (continued)


Category of System
Product Development Management

Individual Systems Within Category


Project milestones Product concept testing process Reports comparing actual progress to plan Project selection process Product portfolio roadmap Budget for development projects Project team composition guidelines

Sales/Marketing Management

Sales targets for salespeople Market research projects Sales force compensation system Sales force hiring and ring policies Reports on open sales Customer satisfaction feedback Sales process manual Sales force training program Marketing collaboration policies Customer relationship management system

Partnership Management

Partnership development plan Policy for partnerships Partnership milestones Partner monitoring systems

intensity groupmore than three times bigger. We grouped companies in years one and three with very similar patterns. We also used statistical tools to further check the signicance of this nding. The pattern was repeatedly consistent with our argument: management systems are associated with growth,21 growth and management systems go together.22 This evidence does not mean that management systems lead to growth. This would be a very nave reading of our results. Growth depends on a number of factors such as the nature of the business opportunity, the depth of management talent, and the quality of the business model to a large extent. Yet, without the ability to execute as reected in the presence of management systems, the probability that growth will be elusive increases.23 The size of the lowest intensity group is interesting in that it reaches about 50 employees. There is some evidence suggesting that the number of employees that can be managed using the personal management style, this is without systems to support managers, is somewhere between 50 and 80 people depending on the complexity of the company and the skills and time commitment of the manager.24

86

UNIVERSITY OF CALIFORNIA, BERKELEY VOL. 52, NO. 3

SPRING 2010

CMR.BERKELEY.EDU

Building Sustainable High-Growth Startup Companies: Management Systems as an Accelerator

FIGURE 2. Company Growth Path in Years 1 to 5 According to Management System Adoption


Intensity

180 160 140

Highest Intensity of System Adoption

Headcount

120 100 80 60 40 20 0 1 2 3 4 5

Mid Intensity of System Adoption Lowest Intensity of System Adoption

Year

Note: Graph is formed by ranking the management system adoption intensity in year 2.

This evidence conrms the argument that the adoption of management systems is associated with growth.25 While this may be counter-intuitivewhy would fast-moving company need tools that appear to constrain creativity and slow down growth?it can be seen through the following metaphor. Think about a car: the faster it goes, the more sophisticated the technology required to keep it under control. At the very elite auto racing level, Formula 1 teams have highly complex and extensive systems infrastructure both on and off the track. The same logic applies to growth with startups. The faster they need to go, the more management systems infrastructure they need.

Venture Capital Investment as a Stimulus to Growth and Management Systems Adoption


Implementing management systems can be costly and entrepreneurs often need guidance in navigating through the entrepreneur crisis. The presence of venture capitalists is associated with faster adoption of management systems. Growth is an imperative for venture capital investors. The criteria many venture capitalists list for selecting their investments invariably includes the ability of the investee company to identify a large market opportunity they can quickly grow to exploit. Figure 3 (drawn from our prior research) highlights this growth

CALIFORNIA MANAGEMENT REVIEW VOL. 52, NO. 3

SPRING 2010

CMR.BERKELEY.EDU

87

Building Sustainable High-Growth Startup Companies: Management Systems as an Accelerator

FIGURE 3. Growth from Starting Month of Venture CapitalBacked vis--vis NonVenture


CapitalBacked Companies

Venture CapitalBacked 60 50 40 30 20 10 0 1 4 7 10 13 16 19

NonVenture CapitalBacked

Headcount

22

25

28

31

34

37

40

Month

Note: Graph is based on monthly payroll data for 194 venture capital-backed companies and 344 non-venture capital-backed companies from Trinet, a third party service provider.

imperative for a sample of 194 venture capital backed companies vis--vis a sample of 344 non-venture capital backed companies.26 In the rst 40 months of a companys existence, the venture capital backed companies grow on average twice as fast. Within our sample of 78 companies, we have 60 venture capital backed and 18 non-venture capital backed companies. Figure 4 presents relative systems adoption of these two groups for the eight management systems categories in Table 1.27 Venture capital backed companies not only adopt more management systems, but also adopt them earlier compared with non-venture capital backed companies.28 The most marked difference between the two groups is in the faster adoption by venture capital backed companies of nancial systems, product development systems, and sales/marketing systems. Successive rounds of venture nancing information are available for the 60 companies. For these companies, we observe a marked increase in the adoption of multiple individual systems within each category over successive rounds of venture nancing. For example, at the time of the series A funding round there is 32% adoption of the three individual systems in the nancial planning category; by series B, the percentage adoption increases to 74%; and by series C, it increases to 86%.

88

UNIVERSITY OF CALIFORNIA, BERKELEY VOL. 52, NO. 3

SPRING 2010

CMR.BERKELEY.EDU

Building Sustainable High-Growth Startup Companies: Management Systems as an Accelerator

FIGURE 4. Rate of Adoption of Eight Types of Management Systems in Years 1 to 5


Venture CapitalBacked vis--vis NonVenture CapitalBacked Companies

Venture CapitalBacked
Financial Planning 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0%
Years

NonVenture CapitalBacked
Financial Evaluation Human Resource Planning 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0%

100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 1 2 3 4 5
Years

Years

Human Resource Evaluation 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0%
Years

Strategic Planning 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 1 2 3 4 5

Product Development Management

Years

Years

Sales/Marketing Management 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0%
Years

Partnership Management 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0%

Years

Note: Graphs are formed by averaging the top two most frequently adopted systems within each category in each year.

CALIFORNIA MANAGEMENT REVIEW VOL. 52, NO. 3

SPRING 2010

CMR.BERKELEY.EDU

89

Building Sustainable High-Growth Startup Companies: Management Systems as an Accelerator

FIGURE 5. Percentage of Founders Replaced From CEO Position According to Management


System Adoption Intensity

80 70

Percentage of Founders Replaced from CEO Position

60 50 40 30 20 10 0 Group 1 Highest Intensity of System Adoption

Mid Intensity of System Adoption

Lowest Intensity of System Adoption

Group 2

Group 3

Note: Graph is formed by ranking the management system adoption intensity in year 2.

CEO Tenure and Management System Intensity


There are two negative outcomes resulting from failure to deal with the entrepreneurial crisis. The rst one is capping the company at a size that can be managed through a personal style. The second one is management chaos. There is an alternative way of highlighting the value of management systems and the negative effect on a CEOs personal career path of failing to invest in such systems as the company grows. Figure 5 provides evidence on this issue. We looked at the tenure of the founders over the life of their companies and related it to management systems intensityup until the day we collected the data for the company. Our measure of system intensity can also be interpreted as proxy for the level of professional tools that the CEO has put in place. We assigned each company in our sample into one of three groups according to systems intensity in the second year of the companys lifehigh, medium, and low. Then, we followed each founder over the life of the company to see whether they had been replaced.29 Finally, we looked at the percentage of founders that had been replaced within each group. Figure 5 depicts our ndings: 53% of founders in the low systems intensity group were replaced compared with only 31% for the highest systems

90

UNIVERSITY OF CALIFORNIA, BERKELEY VOL. 52, NO. 3

SPRING 2010

CMR.BERKELEY.EDU

Building Sustainable High-Growth Startup Companies: Management Systems as an Accelerator

TABLE 2. Alternative Management System Transition Paths at the CEO Position


CEO Path #1 Early Growth Days (1-50 Employees) Beyond Early Growth Days (51+ Employees)
Personal Management Style Personal Management Style

CEO Path #2
Personal Management Style Professional Management Style

CEO Path #3
Professional Management Style Professional Management Style

intensity.30 It appears that those founders that put in place the systems infrastructure to groweither because they knew it from prior experience, their advisers strongly suggested it, or they had the innate skillsare more likely to survive at the leadership position.31 CEOs from companies that have lower adoption of systems have a higher likelihood of not continuing as CEO.

Alternative Management System Transition Paths at the CEO Position


Our interviews with many early stage CEOs show that one of the following three transition paths for a CEO typically occurs as regards to attitudes about the adoption of management systems. We use 50 employees as the dividing line between the Early Growth Days Era (1-50 employees) and the Beyond Early Growth Days Era (51+ employees). A personal management style in the current context is one where the CEO relies on personal contact for communicating and directing the organization. A professional management style is one where respect for analyzing and investing in management systems architecture is part of the CEO leadership. Table 2 summarizes the three different paths that we observed. Entrepreneurs who follow CEO Path #1 continue to manage Beyond Early Growth Days Era as if their personality and interaction were sufcient to provide the growth infrastructure. One marketing manager in our research who reported to a CEO following Path #1 lamented on their many lost opportunities. The CEO (Sam) was the founder and the pivotal technology innovator:
After several lost years, the Board had to crowbar Sam out of the CEO job. It was only after many quarters of missing agreed upon and then imposed upon milestones that we were freed from his chaotic management style.

A classic and at times ugly scenario with CEO Path #1 is that the founder hires a senior executive from the outside and then that external hire subsequently is associated with the removing of the founder. One current CEO in our study described with much personal anguish such a situation. The founder and then rst CEO knew he needed help in professionalizing the management team and its systems but simply could not do it:

CALIFORNIA MANAGEMENT REVIEW VOL. 52, NO. 3

SPRING 2010

CMR.BERKELEY.EDU

91

Building Sustainable High-Growth Startup Companies: Management Systems as an Accelerator

I guess I made the assumption that they were using best practices up front because the person who brought me into the company was the guy who runs all of this. I ultimately had to remove him. Not because he didnt believe in the processes, he just couldnt get it executed.

Ciscos example presented in the introduction illustrates such tension when two successive external CEOs struggled with founders in the transition from a personal management style to a professional management style. It is not uncommon that struggles over the introduction of management systems are part of a broader clash of very different interpersonal styles between a founder and a newly appointed CEO. Companies can follow CEO Path #2 (Personal Management Style in Early Growth Days followed by Professional Management Style) in several ways. One approach is that the CEO retains that role and both changes her management style as well as builds up a professional management team. A second approach is for the board of directors to appoint a new CEO who has a professional management style and that CEO then leads the management team buildup. In both approaches, the import-in concept we encountered in many companies plays a key role. Where a new CEO is appointed in CEO Path #2, the prior CEO may or may not voluntarily resign. Most cases that attract media attention are forced resignations. However, we encountered situations where the founder/CEO proactively planned (and often led) the transition. In one case, the founder (Norman Godinho of NetLogic Microsystems) remains an active and highly valued board member eight years after he planned his exit from the founder/CEO role. Companies following CEO Path #3 (Professional Management Style from the outset and also in the Beyond Early Growth Era) are typically started by serial entrepreneurs. These CEOs often bring along a small cadre of tagalong senior executives from their prior ventures. CEO Path #3 does not require a heavy adoption of management systems from the outset. Rather, the CEO uses her prior experience to bring in pre-identied management systems in a timely way as key milestones are achieved either in the market place or internally within the company.

The Import-in Concept


Even if the entrepreneur adopts a managers mindset and is aware of the need for management systems, she may not have the capability to design and implement them on her own. Another consistent nding in our study is that the design and implementation of management systems requires specic knowledge about the systems themselves as well as experience in using them. In other words, designing these systems is not something that can be done simply by following the instructions in a bookit is not like a do-it-yourself bookshelf that you take home and build yourself. Experience is an important ingredient. This idea translates into the import in concept. A sizeable number of the systems required to grow come with hiring a manager who brings this knowledge with

92

UNIVERSITY OF CALIFORNIA, BERKELEY VOL. 52, NO. 3

SPRING 2010

CMR.BERKELEY.EDU

Building Sustainable High-Growth Startup Companies: Management Systems as an Accelerator

her. As a CEO described the hiring of a VP of engineering: I brought him in because I needed someone with the discipline to make all that happen. The CFO position illustrates this idea. Financial management embeds signicant specic knowledge that is hard to communicatenancial accounting, managerial accounting, relationship with the banks, cash management, valuation, nancial plans, and economic impact of decisions just to name a few. Rather than training a person from the initial bookkeeping all the way to preparing an IPO and managing the nances of a public corporation, startups may hire a person at each stage that comes with the knowledge, implements the systems required at that stage, and uses them effectively from the rst day they become operational to move the company through that particular stage in its growth path. For instance, nancial systems such as budgets or protability reports may appear an easy tool to design. Actually, they require a lot of specic knowledge that only experience can provide. A do-it-yourself approachusing a person within the company with no experience or the entrepreneur herself going about designing itis inefcient. It distracts attention from other activities that may be better suited to that person. The designer without the appropriate background has to go through a costly learning-by-doing process. Not only because of the mistakes that the company has to endure, but also the delay that it imposes on a growth environment. The usual shortcut is to import in the knowledge with a person. In the case of budgets and protability reports, this knowledge is imported into the company when a nancial manager (or CFO) is hired. For instance, a CFO in one of the companies described his hiring as:
I brought a lot of things from a nancial control, nancial reporting, and nancial planning perspective. For example, there was no three-month rolling cash forecast until I arrived. Now, every week, we produce a rolling three-month cash forecast. There wasnt any multiple-year forecast until I got there. There wasnt any acknowledgement that we had large loss carry forwards that might be a benet at some point in the future.

A manager illustrates this import in concept in describing the transition into a professional company:
Ultimately as entrepreneurs, we have to understand when systems are needed and then we hire the people who are the more systems-oriented people and delegate to them. So its not like we are initiating or driving the use of systems as much as through experience understanding when systems are necessary and building a team that can build these systems.

The import in concept does not mean that every CEO has to be replaced or that every system requires a new hire. We saw entrepreneurs that transitioned into being professional managers (although a good portion of those had larger company experience before founding their company) and companies developing their own systems. However, importing this knowledge can accelerate growth.32

CALIFORNIA MANAGEMENT REVIEW VOL. 52, NO. 3

SPRING 2010

CMR.BERKELEY.EDU

93

Building Sustainable High-Growth Startup Companies: Management Systems as an Accelerator

Why Companies Adopt and How They Use Management Systems


The previous ndings raise two related questions. The rst one is: What drives startup companies to realize that they need to make this transition and adopt management systems and processes to facilitate growth? The second question is: What role do these systems play? If these systems are associated with growth, what do they bring to the company and why some companies are more efcient in adopting these systems. These two questions are distinct. While the rst one looks at the reasons that managers described as leading to the adoption of the systems, the second reects the roles that managers identied with these roles once they were on board. The rst is associated with events and settings prior to the adoption, the second looks at these systems after adoption. A fair number of entrepreneurs buy the argument that the bureaucratic nature of management systems kills the entrepreneurial spirit (possibly because it ts their preferences for a personal management style). The previous ndings suggest that their relevance to company growth is relatively obvious. Yet, the adoption of these systems varies a great deal across companies in our sample. Three different researchers analyzed the transcription of 234 interviews (3 interviews for each of the 78 companies) with the purpose of identifying reasons that managers identied as leading to the adoption of the systems. Each researcher identied sentences within the transcripts related to both questions: the prior-to-adoption events and circumstances that managers associated with the adoption and the post-adoption roles that they identied. Next, each researcher grouped these sentences into categories based on their commonalities.33 Finally, the researchers got together, compared the classications that emerged from their coding, discussed any differences in interpretation, and converged into the categories identied. While, pre- and post-adoption categories (reasons for adoption and use, respectively) are distinct, they share some common categories where the reason for adoption was not a particular event or circumstance but to fulll a particular role. Table 3 summarizes the various reasons for adoption, their frequency in our sample, and quotes illustrating them. The six reasons-for-adoption are descriptive of the experiences of the managers in this study: Proactive: Manager BackgroundSenior managers at startup companies with a larger company background often adopted systems and processes because they were used to them. Their large company experience had exposed them to the various systems and processes that are the backbone of execution in large companies; they became second nature to those managers. Systems were adopted because they were an integral part of these managers working style. A different manager might have chosen to use a personal management style that was still feasible given the complexity of the company. Yet these managers chose to use a more professional management style and to adopt management systems.

94

UNIVERSITY OF CALIFORNIA, BERKELEY VOL. 52, NO. 3

SPRING 2010

CMR.BERKELEY.EDU

Building Sustainable High-Growth Startup Companies: Management Systems as an Accelerator

TABLE 3. Reasons for Adopting Management Systems


Frequency of Observation

Reasons

Illustrative Quotes Internal Reasons-for-Adoption

ProactiveManager Background

24

Both of us had come from big company backgrounds where we had used project tracking systems, it seemed natural [], even though it was a very small amount that we were tracking, we were just used to it. That was the normal thing to do. Id have to say that weve had project tracking systems, red-yellow-green status reports, and so forth, from the very beginning. Now were getting to the place where the return is lower compared to the investment, right? Because weve already taken up the low hanging fruit. So now were having to use stepladders to go get fruit. And now it makes more sense to be measuring exactly how many grapes I am getting for how much effort when moving the stepladder. The original engineering team would give me dates when they were going to have certain things done and never make the dates. They didnt communicate amongst themselves, so even when they released something, it didnt do what they said it was going to do. There was no QA process. Often when they did release it, they would introduce more bugs than they would x. Things became less ad hoc because of experience and documentation of process, retaining history of what worked and didnt work, nally came to place. [] By that time, wed had enough experience with enough projects to say, Okay, now we get what works and doesnt work. We can estimate accurately. And it just took doing it a number of times in this new industry and both succeeding and failing to understand okay now we can generalize, we can systematize, we can come up with methodologies that really work. With the advancement of the ofces, we started having 7, 8, 10 different ofces around the nation, so we wanted to standardize the process so that an ofce in Los Angeles was pretty much doing the same type of project management as our ofce in Chicago was.

ProactiveNeed to Focus

ReactiveChaos

11

ReactiveLearning

External Reasons-for-Adoption
Legitimize 2 [Our customer] said I want to see what processes have you instituted in your system. We are going to buy version 5.0 of your product, I want to see what you did from version 3.0, 4.0, 5.0. What were the milestones? What bugs did you x? What was testing that went through? We had done all those things but we didnt have the documentation. So we had to show them and to be candid [recreate it]. Were also nding that the customers get a feeling of control when you give them more data [...]. By seeing data that theyve never seen before, we look good compared to the internal IT, which is one of our strong competitors. Contract 7 When we won the Motorola contract, Motorola forced us to get our act together and so that was a forcing function for all these [product development] processes. Motorola insisted on proper program management, proper change control, proper project reporting, monthly business reviews, monthly project reviews, etc. and they sent audit teams in to audit where we were and make specic recommendations.

CALIFORNIA MANAGEMENT REVIEW VOL. 52, NO. 3

SPRING 2010

CMR.BERKELEY.EDU

95

Building Sustainable High-Growth Startup Companies: Management Systems as an Accelerator

Proactive: Need to FocusManagers in this group had managed relying on the personal style but sensed the need to adopt these systems to enhance execution and shifted to using these management tools.34 Rather than implementing these systems because of their management preferences, managers in these companies waited until it was clear to them that having these systems in place would facilitate continued/new growth. The particular needs that led to the adoption include coordination of a geographically dispersed workforce, improve communication and coordination, or increase organizational efciency. Reactive: ChaosSystem adoption was also associated with management teams experiencing unexpected events, negative outcomes, or recurring problems.35 A manager described it as: Generally you nd that a big change occurs when something goes wrong. Something did with us, and so big change occurred. In most cases, the chaos is unintended and managers were ill prepared to respond to it. Typical unwelcome surprises include badly missing product development or nancial milestones, unexpected liquidity problems, or failing to meet deadlines given to customers. Reactive: LearningSome of the companies in the sample adopted management systems to codify processes that up to that point had been run informally. Once a process has been done repeatedly, managers realize that it is much more efcient to formalize the knowledge in these informal practices through management systems. These systems do not come from managers experience but rather from learning by doing. Setting up a system can reduce the level of reliance on the one person with knowhow. Having a manual provides a roadmap for others in the company to do it. It helps in avoiding mistakes every time that the process has to be executed; but most importantly it liberates precious management attention from ordinary routines. Formalized processes can be delegated most of the time and be managed by exceptionmanagement needs to devote attention only when unexpected events happen. A related benet of establishing a set of protocol is that it helps to minimize the level of impact on the operation if that key personnel leaves the company. Systems can be updated to incorporate new learning. Over time, they become better than any single person in the company (of course, if they are badly updated they can be as dangerous as having no systems). The need to code learning not only goes across time, but also across geography. Coding a process allows this process to travel around the geography without having to move people around.36 External: LegitimizeCertain management systems are adopted to build up the credibility of the company and signal to external parties that they are reliable and well managed; the role of these systems is to legitimize the company.37 In the same way that employee number ve might be a receptionist that answers the phone as if the company was a blue chip or asks visitors to wait in a room before the vice-president comes, certain systems give an image of the company as a professional, well-run, reli-

96

UNIVERSITY OF CALIFORNIA, BERKELEY VOL. 52, NO. 3

SPRING 2010

CMR.BERKELEY.EDU

Building Sustainable High-Growth Startup Companies: Management Systems as an Accelerator

able company. The presence of these systems signals to potential investors, customers, and partners that management is professional, reliable, and knows what it is doing. These systems range from nancial planning when dealing with nancial investors, product development when dealing with large partners, or sales when working with large customers. External: ContractingFinally, some companies adopt management infrastructure because a partner, customer, or government regulation imposes the systems. Biotechnology startups are good examples because they often get early revenues (cash infusions) from contracts with large pharmaceutical companies. For instance, one of the companies in the sample put in place well-structured systems in product development because Pzer, as part of the agreement, put in place milestones that we had to achieve. These contracts require a disciplined management of the research efforts. Biotechnology startups also work in a tighter regulatory environment that requires careful documentation of their development efforts. Having these systems imposed from outside the organization is not exclusive to biotechnology companies. Companies such as Cisco and Hewlett-Packard are known to require early stage companies to have a systems infrastructure before becoming certied vendors. We further examined the association between speed of adoption38 and the categories described. Managers background was the category related to fastest adoption. Thus, when managers are aware of the need to start using management systems, these systems are adopted earlier. This observation is consistent with these systems growing behind the actual needs of the startup company; those managers that are not sensitive to this transition challenge are driven into these systems as a reaction to chaos or routines or as a demand from external parties. The previous classication lists reasons-for-adoption that managers identied during the interviews. The interviews also allowed us to understand how managers used these systems. Certain reasons for adopting management systems t into a particular role that these systems fulll such as legitimizing the company to external constituencies and contracting. Yet other reasons are associated with particular events that map into important roles that these systems have. Table 4 provides illustrative quotes for these additional uses identied: Making Goals Explicit and StableManagement systems in uncertain environments such as high-growth startups make goals explicit and stable.39 Sharing goals has several benets in these companies. It has everybody on the same page, facilitates discussion within a shared view of the future, and provides the stability required to execute. When goals are not discussed or explicit, they may mean different things to different managers. Vivid examples often come from the software industry. Customers continuously identify new needs. In a company with a few dozens of customers, the amount of weekly requests on software developers can be very large. Without a system that freezes the specications to those

CALIFORNIA MANAGEMENT REVIEW VOL. 52, NO. 3

SPRING 2010

CMR.BERKELEY.EDU

97

Building Sustainable High-Growth Startup Companies: Management Systems as an Accelerator

TABLE 4. Using Management Systems


Roles
Making Goals Explicit and Stable

Illustrative Quotes
As a result of commercial development, myself and a couple of the other folks thought wed better put a structure in place where at some point we freeze that and then we develop against that specication. We also have an approval process that goes through and decides who clears that to be released. Its a challenge because you have a moving target most of the time, but the reality is if that specication changes on too dynamic a basis, theres no way really a technology team can execute against it. [Informal communication] worked okay when we were 20. Now were 160 and the same thing doesnt happen. Before, we could communicate in a certain way. We almost did not have to communicate. Everyone always knew everything. We didnt have to worry about communication. As a bigger company we need to worry about communication. Does everyone know? Everyone is not as well integrated as well as we were when we were 20. When it was 20 people we didnt have a lot of emails necessary. Now we think about whether we should send an email to everyone to communicate this issue. Or we have weekly company meetings. Now were more diligent about using those as means of communication. Now, there are weekly company meetings, and people are required to come. Financial planning forces the company, the whole executive team, and the board to get on the same pagehere is what were trying to do. If you didnt have that, there could be some ambiguity on what the goals are. The benets of [product development processes] formalization is that you get more accountability. I think it is very easy from an engineering perspective to say I can do that. [] What you get is (a) you can hit that expectation in a timely manner, and (b) develop features that meet the market on time. [] As the company matures, people get more accountable.

Help with Coordination and Plan the Sequence of Steps

Facilitate Decision Making and Resource Allocation Promote Accountability and Facilitate Control

that these engineers are working towards, specications would change too fast for them to be able to deliver to an ever-changing set of requests. Help with Coordination and Plan the Sequence of StepsManagement systems facilitate coordination within the company. A plan provides a sequence of actions to achieve a goal. A sales person can promise a certain feature to a customer because she knows that the product roadmap plan indicates that the feature will be available at an indicated date. Smaller companies can coordinate through direct interaction. But the number of channels of communication increases in factorial proportion with the number of people.40 Coordination without any management infrastructure quickly becomes unfeasible as a company quickly grows. Facilitate Decision Making and Resource AllocationA signicant challenge facing managers of high-growth startup companies is to obtain timely and accurate information generated within the company for decision making. Management systems can help compile and categorize detailed operating information into an aggregated level. This systematic treatment of information greatly enhances the timeliness and efciency of communication between business functions and with top management as everyone is on the same page. It also increases the quality of information available to

98

UNIVERSITY OF CALIFORNIA, BERKELEY VOL. 52, NO. 3

SPRING 2010

CMR.BERKELEY.EDU

Building Sustainable High-Growth Startup Companies: Management Systems as an Accelerator

managers for decision making. Management systems also allow managers to see the big picture and identify easily where the bottleneck is, where the improvements can be made, and/or where the resources are most needed. Promote Accountability and Facilitate ControlManagement systems liberate managers attention from routine activities. Once formal procedures are in place to execute a certain process, managers do not need to constantly monitor it as long as it behaves within expectations. They can devote their energies to pursue new opportunities. Only when there are deviationsand the system will alert management if it happensis attention required. Without these warning signals, control by exception is not feasible. Furthermore, employee motivation requires aligning incentives between employees and the company. This in turn requires performance measures where everyone is accountable for his or her actions and their consequences, and is rewarded or punished accordingly (the performance-pay link). When the company is small, performance evaluation is relatively easy as the entrepreneur works closely with every employee and is well aware of each individual contribution. However, as the company scales up in terms of headcount and product lines, it becomes much harder to fairly evaluate and compensate employees without using some formal systems that document each individuals scope of responsibility and outcome levels.

Which Systems to Adopt First? Sequencing Management Systems Adoption


A nal question that emerges from the analysis of the eight categories and 46 individual management systems is whether there is a more effective sequencing of adoption associated with their relevance to future performance. To address this question, we traced the adoption pattern of the various systems across our sample companies since their founding. The thrust in the data is that there is no one size ts all solution and the sequencing of adoption varies with the needs of the company. Nevertheless, various patterns do emerge. First, as illustrated in Figure 4, venture capita backed companies give higher priority to nancial planning and evaluation categories. Interview data indicates that these companies need to formalize this information to manage their cash burn rate and negotiate a new round of nancing with enough time for the negotiation process. They also facilitate communication with venture capitalists that closely monitor rm performance. Startup companies in the investment phase where cash ows are negative benet from adopting nancial planning and evaluation systems early on. These systems facilitate resource allocation and allow the entrepreneur to set priorities and focus on what is important so she can plan her subsequent activities accordingly. This is particularly crucial in the startup phase when the entrepreneur still wears multiple hats and manages many aspects of her workplace by herself. These companies

CALIFORNIA MANAGEMENT REVIEW VOL. 52, NO. 3

SPRING 2010

CMR.BERKELEY.EDU

99

Building Sustainable High-Growth Startup Companies: Management Systems as an Accelerator

can effectively structure their strategic thinking within the company, with board members, and with investors through these systems. The dialogue around the nancing needs and expectations brings together ideas around the business model and the cost structure of the company. These systems often behave as substitutes of strategic and human resource planning systems because the information exchange structured around these latter systems can happen through the nancial plans. Statistical analyses conrmed this observation. Evaluation systems are implemented after planning systems are in place, as common sense predicts. Interestingly, however, within the three planning systems categoriesnancing, human resources, and strategicthe statistical analysis identied a substitution effect between nancing system and the combination of human resources and strategic systems. Specically, companies that adopt nancial planning tend to delay the adoption of human resources and strategic planning, while those adopting human resource planning systems also adopt strategic planning systems and delay nancial systems. Second, the adoption of human resource planning and evaluation systems was not linked as much to events or circumstances of the company as to the management model of the founding and management team.41 Certain managers give a strong relevance to managing the soft side of the company. For instance a CEO described his view as: You have to have a goal, a purpose. Before we even opened the door, there was a mission statement and core values in place. Core values are extremely important in terms of client relationships and how youre going to present and hold yourself. When should a company dene its recruiting policies, its values, or its evaluation systems? While the need for nancial systems is heavily associated with negative cash ows and the presence of external investors, human resources systems are often not driven by the business structure but by the founders management philosophy. Third, our analyses also suggest that rms with longer R&D cycles, such as biotechnology and hardware rms, adopt new product development systems sooner while delaying marketing and sales systems. Our interviews with biotechnology companies indicate that these systems are put in place early because of the large number of scientists that need to be coordinated, the high regulation requirements coming from the FDA, the large number of contracts with established pharmaceutical companies, and the high visibility that investors and partners demand. Fourth, the go-to-market stage has a signicant effect on marketing and sales systems. Their adoption is delayed until this turning point. Companies that adopt a direct sales strategy are more likely to bring in sales and marketing systems earlier than those using an indirect sales strategy. The go-to-market stage happens at different time frames across industries. In biotechnology rms, this stage happens late, if at all (these companies are often sold before they try to sell their products in the market). Information technology companies experience this stage on a time horizon between one and ve years. Only then, these systems start to have a signicant role. Other industries see this go to market very

100

UNIVERSITY OF CALIFORNIA, BERKELEY VOL. 52, NO. 3

SPRING 2010

CMR.BERKELEY.EDU

Building Sustainable High-Growth Startup Companies: Management Systems as an Accelerator

early in their lives. For instance, a consulting company in our sample quickly adopted and used these sales and marketing systems intensively form very early on in its life. The company was selling from day one. Customer relationship and sales pipeline management were crucial to their business model. In contrast, nancial planning and evaluation were much less relevant as the company was cash ow positive and cash was not a signicant constraint to growth. Finally, the data indicate that there is a high positive correlation between time-to-adoption of a new system and the number of systems already in place. This is consistent with the repetitive momentum hypothesis in the organizational change literature. That is, change is considered to be a self-reinforcing process; prior changes increase the likelihood of a subsequent change. In the context of this study, prior implementation of management systems likely reduces organization inertia and the marginal cost of setting up another system.42 This observation suggests a consistent view of managers in each company regarding the relevance of management systems to growth as if these systems reected a particular management philosophy rather than a preference for certain systems at the expense of others.

Conclusions
This article reports ndings consistent with the following main ideas. At some point in the growth of a company, somewhere between 50 and 100 employees, the management style need to change from a personal management style typical of many early-stage entrepreneurial companies into a more professional style. Some companies and their CEOs are unable to make these transitions and fail in what is called the entrepreneurial crisis. Putting together the management infrastructure is associated with higher growth and lower CEO rotation. The presence of professional investors (such as venture capitalists) is associated with higher growthmost likely because these investors have a vested interest in growth and will force the transition at the CEO position when the entrepreneur is unable to transition into becoming a manager. These systems are adopted proactively when the management team has the relevant management knowledge or reactively when bad outcomes or mistakes happen that highlight the limitations of the entrepreneurial management style. These systems have multiple roles that they can fulll in providing the adequate management infrastructure. Finally, the sequencing of adoption is contingent on the particular needs of the company.

APPENDIX A Research Sources


The insights that we report in this article are based on various sources that we have developed over the last decade:

CALIFORNIA MANAGEMENT REVIEW VOL. 52, NO. 3

SPRING 2010

CMR.BERKELEY.EDU

101

Building Sustainable High-Growth Startup Companies: Management Systems as an Accelerator

TABLE A. Characteristics of the SEMAS Research Project Companies


Mean Number of CEOs Years of total experience of CEOa Age of the company Number of employeesb R&D intensityc (%) Revenues (000)d Income (000)d Number of rounds of VC funding Systems infrastructure International rms CEO MAS model: planning (%)
a. Estimated for all CEOs in the sample. b. Number of employees is calculated at the peak of each companys size. c. R&D intensity are estimated as a percentage of total employees dened as the sum of R&D employees for each of the years reported divided by the sum of total employees for those same years. Only companies that reported R&D employees are included. d. Revenues and prots are for the last year of data available. Source: A. Davila and G. Foster, Management Accounting systems Adoption Decisions: Evidence and Performance Implications from Early-Stage/ Startup Companies, The Accounting Review 80/4 (2005): 1039-1068.

Std. Dev.
0.78 8.48

Q1
1 12

Median
1 20

Q3
2 25

1.65 18.42

5.47 118 39.57 10,923 9,455 3.43

2.44 62 26.08 11,853 15,668 1.78

3 71 19.66 2,437 12,469 2

5 113 37.78 7,300 -4,700 3

7 155 60.56 15,000 29 5

-0.05 0.60 17

0.88 0.49

-0.78 0

-0.26 1

0.68 1

The SEMAS (Stanford Entrepreneurial Management Systems) research project. This study looked at 78 startup companies located for the most part in Silicon Valley. We focused on startups that have grown to more than 50 employees and were less than 10 years old. Table A provides more detailed information about this study. The Stanford EPGC (Executive Program for Growing Companies) and its business challenges component. EPCG has been a successful executive program offered at Stanford for many years. Over the last 15 years, the school has developed a rich database on the challenges that startup companies face as they move forward in their lives. Business cases. Over the years, we have been fortunate to work together with a lot of different companies, writing case studies with them or simply sharing their successes and failures. These cases contain the experi-

102

UNIVERSITY OF CALIFORNIA, BERKELEY VOL. 52, NO. 3

SPRING 2010

CMR.BERKELEY.EDU

Building Sustainable High-Growth Startup Companies: Management Systems as an Accelerator

ences of these companies and the learning that can be extracted from these experiences. The pre-IPO valuation research project. This study used various sources (including companies SEC lings and the Venture One database) to understand how to value pre-IPO companies. The employee growth study looked at the different growth path of venture capitalbacked versus non-venture capital-backed startup companies. Figure 3 illustrates the very different growth path of these companies; it plots the mean number of employees over time. Enterprise Ireland Stanford program. This year-long program funded by Enterprise Ireland brings together 25 to 30 CEOs of Irish growth-oriented companies. The aim in carrying these various projects was to study companies that had growth aspirations and had achieved at least a minimum size. These types of startups contribute a signicant percentage of the economic value that entrepreneurial endeavors bring to society. Not surprisingly, the most frequent source of funding for most of the companies that we worked with was venture capital and the companies clustered around technologytelecommunications, information technology, and biotech. Notes
1. This denition is widely used in the literature and comes from R. Simons, Levers of Control: How Managers Use Innovative Control Systems to Drive Strategic Renewal (Boston, MA: Harvard Business School Press, 1995), p. 5. 2. We use founder as a generic term. In some cases it is a single person. In other cases, it is two or more people. Our conclusion relates to the collective effect of decisions made by the founder(s). 3. Innovation is often associated with entrepreneurial companies and the blocks to innovation are often extended to entrepreneurial companies, see for instance J. Freeman and J.S. Engel, Models of Innovation: Startups and Mature Corporations, California Management Review, 50/1 (Fall 2007): 94-119. 4. Early empirical evidence is consistent with this argument, see for instance, F. Damanpour, Organizational Innovation: A Meta-Analysis of Effects of Determinants and Moderators, Academy of Management Journal, 34/3 (September 1991): 555-590. 5. The classic reference to the entrepreneurial crisis is L.E. Greiner, Evolution and Revolution as Organizations Grow, Harvard Business Review, 50/4 (July/August 1972): 37-46. 6. Cisco Systems, Stanford Graduate School of Business Case S-SB-124, March 3, 1992. 7. Cisco Systems, op. cit., pp. 5-6. 8. Cisco Systems, op. cit., pp. 7-8. 9. Cisco Systems, op. cit., p. 9. 10. Research studies that have illustrated this alternative are L.B. Cardinal, S.B. Sitkin and C.P. Long, Balancing and Rebalancing in the Creation and Evolution of Organizational Control, Organization Science, 15/4 (July 2004): 411-431; P.M. Collier, Entrepreneurial Control and the Construction of Relevant Accounting, Management Accounting Research, 16/3 (September 2005): 321-339. 11. The limitations of entrepreneurs as managers has often been referred to in the entrepreneurship literature and has been associated with the characteristics of the entrepreneur, see G.E. Willard, D.A. Krueger, and H.E. Feeser, In Order to Grow, Must the Founder Go: A Comparison of Performance Between Founder and Non-Founder Managed High-Growth Manufacturing Firms, Journal of Business Venturing, 7/3 (May 1992): 181-194. 12. A more detailed description of Salesforce.com early days can be found in Salesfoce.com: The Evolution of Marketing Systems, Stanford University case number E-145.

CALIFORNIA MANAGEMENT REVIEW VOL. 52, NO. 3

SPRING 2010

CMR.BERKELEY.EDU

103

Building Sustainable High-Growth Startup Companies: Management Systems as an Accelerator

13. The sample was built as follows. The selection criteria were for companies to have between 50 and 150 employees at the sampling date (to be going or have recently gone through the transition that a group of CFOs for startup companies identied to be on this range), be less than 10 years old (to be able to reconstruct the history of the company), and be independent (to avoid having them adopt systems from their parent company). We used different databases such as CorpTech Internet Directory of Technology Companies, Richs High-Tech Business Guide to Silicon Valley and Northern California, BioScan, or U.S. Business Browser to identify companies. To each company we sent a letter with a description of the project and called them to arrange their participation. Those companies that did not answer the phone were dropped from the sample. From an initial sample of 624 companies, 213 were eligible and 78 participated. The data was collected through questionnaires and interviews to the CEO, CFO, and business development manager. 14. Additional references to the growth of startup companies include I. Adizes, Corporate Life Cycles: How and Why Corporations Grow and Die, and What to Do about It (Englewood Cliffs, NJ: Prentice Hall, 1989); E.G. Flamholtz and Y. Randle, Growing Pains: Transitioning from an Entrepreneurship to a Professionally Managed Firm (San Francisco, CA: Jossey-Bass, 1990). 15. Because we are interested in the transition around the entrepreneurial crisis, we sampled companies that had reached a certain complexity as measured by headcount. Thus, the ndings are conditional on companies reaching a size where the entrepreneurial crisis becomes relevant. Our ndings do not address whether early adoption of management systems are associated with increased failure at smaller sizes. However, these systems are typically not present (or present at a very simple scale) in smaller rms [see G. Cassar, Financial Statement and Projection Preparation in Startup Ventures, Accounting Review, 84/1 (2009): 2751] and thus failure is not likely to be associated with them. 16. Recent theoretical developments have provided the tools to understand when the previous arguments may hold. See, for instance, R. Simons, Levers of Control: How Managers Use Innovative Control Systems to Drive Strategic Renewal (Boston, MA: Harvard Business School Press, 1995); P.S. Adler and B. Borys, Two Types of Bureaucracy: Enabling and Coercive, Administrative Science Quarterly, 41/1 (March 1996): 61-89. Empirical evidence has started to emerge consistent with these arguments, such as L.B. Cardinal, S.B. Sitkin, and C.P. Long, Balancing and Rebalancing in the Creation and Evolution of Organizational Control, Organization Science, 15/4 (July 2004): 411-431; M. Granlund and J. Taipaleenmaki, Management Control and Controllership in New Economy Firmsa Life Cycle Perspective, Management Accounting Research, 16/1 (March 2005): 21-57. 17. From our interviews with three managers in each of the 78 companies, we only identied one company where systems were implemented ahead of the needs. In all other companies, managers indicated a management infrastructure that was not as strong as they would like it to be. 18. For further evidence supporting these arguments over a large number of venture capitalbacked companies, see C. Armstrong, A. Davila, and G. Foster, Venture-Backed Private Equity Valuation and Financial statement Information, Review of Accounting Studies, 17/1 (March 2006): 119-154. 19. Figure 2 illustrates the argument. More advanced statistical testing can be found in A. Davila and G. Foster, Management Control Systems in Early-Stage Startup Companies, The Accounting Review, 82/4 (July 2007): 907-937. 20. This variable does not capture the complexity of the system. However, most of the systems were simple and comparable. If the bureaucracy argument affected our sample, we would see the highest intensity not growing as fast (being dragged by the bureaucracy). As we have argued, most of the companies had systems that trailed their needs. 21. We appreciate the specic comment of one of the reviewers to clarify this point. 22. It is important to reinforce that systems intensity is a necessary but not sufcient condition. Thus, adopting systems will not lead to growth per se, but sustained growth without systems is much less likely. 23. The growth prole of startup companies varies across industries. We focused our research on information technology, biotechnology, and non-technology companies with venturebacked funding. In all our statistical tests, we controlled for the potential effect of industry. 24. For instance, see A. Davila, An Exploratory Study on the Emergence of Management Control Systems: Formalizing Human Resources in Small Growing Firms, Accounting, Organizations and Society, 30/3 (April 2005): 223-248.

104

UNIVERSITY OF CALIFORNIA, BERKELEY VOL. 52, NO. 3

SPRING 2010

CMR.BERKELEY.EDU

Building Sustainable High-Growth Startup Companies: Management Systems as an Accelerator

25. The argument does not imply causality. Moreover, growth and the adoption of management systems are endogenous variables simultaneously determined. In our statistical tests, we use a simultaneous equation model to address endogeneity to the best extent possible. The argument coming out of the test is that the association between growth and management systems goes both ways. 26. See Davila, Foster, and Gupta (2003), op. cit. The monthly headcount information in Figure 3 is based on payroll data from Trinet, a third-party payroll service provider. 27. For the effect of venture capital funding on management systems adoption see Davila and Foster (2007), op. cit. 28. For additional evidence on the impact of venture capital funding on the adoption of professional characteristics see T. Hellmann, and M. Puri, Venture Capital and the Professionalization of Start-Up Firms: Empirical Evidence, Journal of Finance, 57/1 (February 2002): 169-198. 29. Statistical tools support the argument and include alternative variables to make sure that the pattern is not explained by other variables such as industry, CEO startup experience, and age of rm; the association was there. See Davila and Foster (2007), op. cit. 30. The association between the presence of management systems and CEO replacement does not imply that a CEO is replaced because she failed in trying to implement the systems (we do not suggest causality). It might be the CEO is replaced because she is perceived to be unable or unwilling to do so or any other reason correlated with the presence of management systems. 31. To make sure that the nding was not an artifact of some unique event in year two of a companys life, we ran the same test grouping companies by systems intensity in year one and again in year three. We found the same pattern. 32. A process known in the literature as grafting. Background is closely associated with congenital learning, where individuals . . . have knowledge about . . . the processes the organization can use to carry out its creators intentions. G.P. Huber, Organizational Learning: The Contributing Processes and the Literatures, Organization Science, 2/1 (February 1991): 88115, at p. 91. 33. This process is often referred as grounded theory as categories emerge from the data. 34. This reason for adoption is typical of entrepreneurs who are able to transition into becoming managers. These entrepreneurs sense the need to change their style and implement these systems through importing in through new hires or because they have large company background. 35. See R. Simons, Levers of Control: How Managers Use Innovative Control Systems to Drive Strategic Renewal (Boston, MA: Harvard Business School Press, 1995). 36. On the topic of procedures helping innovation by coding learning from past experience see C.C. Lundberg, Learning in and by Organizations: Three Conceptual Issues, International Journal of Organizational Analysis, 3/1 (January 1995): 10-24; B. Levitt and J.G. March, Organizational Learning, Annual Review of Sociology, 14 (August 1988): 319-340. Coded routines facilitate the diffusion across the organization and over time of organizational capabilities, see R.R. Nelson and S.G. Winter, An Evolutionary Theory of Economic Change (Cambridge, MA: Harvard University Press, 1982). 37. See for example W.W. Powell, Learning from Collaboration: Knowledge and Networks in the Biotechnology and Pharmaceutical Industries, California Management Review, 40/3 (Spring 1998): 228-240. For a traditional description of the relevance of the external context in explaining how rms are organized, see J. Pfeffer and G.R. Salancik, The External Control of Organizations: A Resource Dependence Perspective (New York, NY: Harper & Row, 1978). 38. We ran this test on a particular product development system, project milestones, to examine innovation-related systems. We found similar patterns for planning and evaluation systems. 39. For the importance of stable goals in the broader innovation literature see T.M. Amabile, How to Kill Creativity, Harvard Business Review, 76/5 (September/October 1998): 76-81. 40. A related point in new product development is made by A. Ditillo, Dealing with Uncertainty in Knowledge-Intensive Firms: The Role of Management Control Systems as Knowledge Integration Mechanisms, Accounting, Organizations and Society, 29/3-4 (April/May 2004): 401-422. 41. The idea of the inuence of management and founders models on how the company is organized is documented in J.N. Baron and M.T. Hannan, Organizational Blueprints for Success in High-Tech Start-Ups: Lessons from the Stanford Project on Emerging Companies, California Management Review, 44/3 (Spring 2002): 8-36.

CALIFORNIA MANAGEMENT REVIEW VOL. 52, NO. 3

SPRING 2010

CMR.BERKELEY.EDU

105

Subscribe, renew, and order reprints online at cmr.berkeley.edu


California Management Review
University of California F501 Haas School of Business #1900 Berkeley, CA 94720-1900 (510) 642-7159 fax: (510) 642-1318 e-mail: cmr@haas.berkeley.edu web site: cmr.berkeley.edu