Description
During this such a detailed outline about strategic choices in new venture development and the value of business planning.
DISCUSSION PAPER
SERIES IN
ECONOMICS AND
MANAGEMENT
Strategic choices in new venture development and
the value of business planning for nascent
entrepreneurs
C. Hopp
Discussion Paper No. 11-26
GERMAN ECONOMIC ASSOCIATION OF BUSINESS
ADMINISTRATION – GEABA
Strategic choices in new venture development and the value of business planning for nascent
entrepreneurs
Preliminary Version: Please do not cite without the authors permission
July 2011
Dr. Christian Hopp
University of Vienna
Chair for International Personnel Management
Brünner Straße 72, 3/241
Tel: +43-1- 4277-38166
Fax: +43-1- 4277-38164
E-Mail: [email protected]
Strategic choices in new venture development and the value of business planning for nascent
entrepreneurs
Abstract
Theory suggests a positive relationship of business planning with new venture formation. Yet,
empirical evidence is scarce, or ambiguous. In contrast to previous empirical studies, we treat the
decision to plan or not to plan as an endogenous strategic choice by entrepreneurs. Using data from the
PSED 2 study, we empirically correct for self-selection and account for endogeneity. Bearing in mind
that business plans serve a specific purpose, among others either for learning and guidance of the
entrepreneur or for third parties to attract outside capital, different forms of business planning are
beneficial for entrepreneurs. Business planning per se does not have a statistically significant impact
on the chances to found a new venture successfully. Rather, it is the process of planning and
subsequent learning through planning that matters for successful venture emergence. However, when
entrepreneurs are seeking outside funding, making financial projections is affecting the founding
probability positively, but not the ceremonial purpose of business plans. Moreover, market newness
might moderate the value of business planning for nascent entrepreneurs. When information becomes
outdated more quickly, the value of plans for learning purpose is reduced. We conclude that
understanding when business plans are needed and for which purposes, is crucial in empirically
disentangling performance impacts and providing practical advice.
1. Introduction
The value of business planning is widely debated in the academic literature. Some highlight the
advantage of bringing some structure amidst the chaos of uncertainty and ambiguity when founding a
new venture (Delmar and Shane, 2003; Gruber, 2007). Others suggest that nascent entrepreneurs
should spend their time on more relevant organizing activities, among others focusing on gaining
legitimacy and establishing contacts with customers (Bird, 1988; Bhidé, 2000). Generally, it is unclear
whether business planning is key to success or distracting time from other, more critical
entrepreneurial tasks.
In this work, we argue that business planning is self-selected and, therefore, endogenous. Subsequent
organizing activities can be traced back to the individuals pursuing these actions reflecting their
inherent resources or their preparatory actions to create the new venture. Different entrepreneurs may
face similar (or even the same) opportunities, yet organizing patterns differ (Bowman and Hurry,
1993; Baker and Nelson, 2005). Specifically, perceptions of entrepreneurs provide the basis for
decision-making. Hence, they explain why certain entrepreneurs rely on business planning and others
do not. Consequently, estimating strategic choices in empirical entrepreneurial research by neglecting
the endogenous nature of such decisions leads to erroneous results.
We document that endogenous strategic choices with respect to business planning based on pre-entry
experience can help to explain entrepreneurial success and/or failure. We extend the ideas put forward
in Hamilton and Nickerson (2003) by modeling the strategic choice and by accommodating for
imbalances caused by individuals self-selecting into planning activities. Thus, we can estimate how
performance changes conditional on the entrepreneur’s choices. In particular, we match the pre-entry
experience of entrepreneurs with the choice for (or against) business planning (and the extent of it).
Thus, we infer the counterfactual to estimate the performance differential caused by strategic decisions
made. Moreover, we emphasize the conditions under which certain opportunities lead to higher
performance. In sum, we demonstrate that empirical studies must appropriately account for the
strategic choices made by entrepreneurs.
Our results show that contrary to previous findings, starting to draw up a business plan does not affect
new venture emergence. Prescriptive advice that focuses on initiating business planning solely does
not constitute a valuable strategy for nascent entrepreneurs. In general, finishing the business plan
carries more value when we match nascent entrepreneurs that plan with those that did not plan.
Moreover, producing a formal business plan and making financial projections add significant value for
nascent entrepreneurs, whilst subsequently reworking the plan does not increase the chances to
perceive the venture as operational/emerged.
However, the purpose of the underlying planning activities significantly moderates the value created
by business planning. While some planning activities appear to be valuable for entrepreneurs that did
not seek outside financing, only financial projections are valuable for entrepreneurs that indicated that
they actively seek outside capital. Hence, when business plans are written for a mere ceremonial
purpose the value stemming from learning is reduced. This finding provides evidence that business
plans satisfy a ceremonial purpose or third party needs. Moreover, when business plans are written for
products that are essentially new to all potential customers, none of the business planning activities
creates value and affects the chances to perceive the venture as emerged. Hence, we conclude that
when information becomes outdated more quickly (or might simply be not available to be worked into
plans) learning from business planning activities cannot take place, and therefore business planning
does not add significant value for nascent entrepreneurs.
Accordingly, our results provide insights for research and practice alike. On the one hand, we
highlight the role of business planning as a dynamic and emergent tool that needs to be adapted
continuously in order to be successful. While starting to plan does not add value, it presents the entry
into the process of business planning. However, it can only add value if nascent entrepreneurs are
willing to accept business planning as a process comprising a series of activities, that they are willing
to undertake. On the other hand, we document that advice needs to be appropriately tailored to
entrepreneurial situations. The advice to plan should not be purely prescriptive and focused on the idea
of having a one-size-fits-all business plan. Rather it should appreciate the contingent nature of
planning itself, mirroring the ever-changing environment that eventually finds its way into the formal
business plan and more importantly, needs to work in the actual purpose of why nascent entrepreneurs
want to engage in planning activities in the first place.
The remainder of this article is structured as follows: section two outlines the theoretical background
and derives the hypotheses. Section three describes the data and presents the methodology. Results are
reported in section four. Section five discusses the findings, the studies potential limitations, and
provides concluding remarks.
2 Theoretical Background
2.1 The Role of Business Planning in Nascent Entrepreneurship
Very essentially, entrepreneurs pursue potentially profitable investment opportunities and the build of
organizations to actually reap the benefits of their effort (Haber and Reichel, 2007; Gartner, 1985).
While plagued by high levels of uncertainty and ambiguity, they have to structure a complex set of
interrelated tasks. Substantial discretion in entrepreneurial decisions is necessary to take the
appropriate organizing activity.
Drawing up a business plan is among the most recognized of such activities. It involves the collection
and analysis of relevant information to identify future tasks, risks, opportunities, and to derive viable
contingencies for future actions, likely in a written format (Delmar and Shane, 2003, Gruber, 2007,
Brinkmann et al., 2010). The value of a business plan for nascent entrepreneurs is widely debated in
the academic literature and the impact on future success, remains ambiguous.
Following Dencker et al. (2009) business planning is a learning tool, that comprises learning from
others and cognitive search. Business planning can facilitate learning by helping to identify and detect
patterns and to draw meaningful conclusions from them. In general, it comprises reachable goals and
situational analysis and decision making. In this vein, a pro-active business plan is necessary to
develop and interpret future information (Frese et al., 2007). Hence, business planning in the nascent
phase, that is prior to the launch of a new organization, involves the ex-ante collection and analysis of
information and subsequently helps to get an understanding of the requirements to be successful as an
organization. That is, business planning provides objectives and necessary steps to reach goals.
Therefore, rather than providing a clear plan to which entrepreneurs stick, business planning is a
process of learning about the business model, the customer requirements, and eventually the building
of an organization and hence, guides entrepreneurial actions. Consequently, business planning changes
the range of potential behaviors and therefore is an essential tool in the learning process of nascent
entrepreneurs (Dencker et al., 2009; Huber, 1991).
Business planning might reduce part of the uncertainty by deriving action plans and facilitating
decision-making and leads to a more effective organization of the founding process and, therefore to a
higher probability of success (Ansoff, 1991; Delmar and Shane, 2003). Planning could provide
entrepreneurs with a framework to structure tasks and get an overview of the overall structure of
entrepreneurial activities to reduce erratic behavior. The benefits of business planning stem from
predictions about future challenges and the corresponding implementation of milestones and allows
for more effective use of scarce resources. The quality of decisions and the allocation of resources
improves due to better information derived from previous organizing steps taken (Delmar and Shane,
2003; Brinkmann et al., 2010). Entrepreneurs become aware of options and strategic choices in
navigating through the founding process (Ansoff, 1991, Delmar and Shane, 2003). Moreover, business
planning provides a benchmark to test previously made assumptions against and therefore enables
entrepreneurs to overhaul strategic approaches if predicted outcomes do not materialize. Overall,
planning should be related to a more systematic way of organizing the foundation process and thus
foster goal achievement (Gruber, 2007).
2.2 Existing Knowledge and Experience as Antecedents of Business Planning
Complex tasks such as business planning need to rely on cognitive skills to allow for effective
planning given the interrelatedness of many facets. Chwolka and Raith (2011) refer to business
planning as a collection of skills that can be trained and that are inherently linked to the decision
maker’s prior knowledge and experience. Prior related knowledge is essential in scanning the external
environment to identify external knowledge and make do with the accessed new knowledge
(Lichtenthaler, 2009). Skills and knowledge provide the existence of absorptive capacity that enables
an entrepreneur to acquire and transfer new information collected in the process of business planning
into actions (Cohen and Levinthal, 1990; Zahra and George, 2002). Entrepreneurs that possess more
pre-entry experience will derive larger benefits from business planning. They already have experience
with general planning activities and hence, are equipped with “a structure for how to plan as well as
practice with the cognitive act of planning” (Dencker et al., 2009: 520). Skills and knowledge
therefore are related to proactive planning and equip people with skills to deal with new and complex
situations. Hence, they help in acquiring knowledge and facilitate decision making. Pre-entry
experience acts as an antecedent to elaborate and proactive planning (Frese et al., 2007).
Thus, we argue that pre-entry experience of nascent entrepreneurs serves as a basis for future learning.
During the process of business planning, pre-entry experience interacts with the organizational context
to create new knowledge and helps entrepreneurs to act accordingly. Gathering information through
business planning and processing them subsequently is greatly eased by the existence of pre-entry
industry and managerial experience. Thus, knowledge repositories enable entrepreneurs to plan more
effectively by knowing what questions to ask, how to analyze and most importantly, how to interpret
their findings to derive actions from them. Existing knowledge facilitates comprehending and
integrating new information. Consequently, initial endowments of nascent entrepreneurs influence
their adaptability and more experience subsequently gives more margins to maneuver (Huber, 1991;
Argote and Miron-Spektor, 2011; Dencker et al., 2009).
In essence, higher levels of pre-entry experience of nascent entrepreneurs increase the benefits from
subsequent learning activities. Entrepreneurs would choose to learn from business planning or forego
formal planning based on their attributes and experiences. Consequently, ceteris paribus, the higher the
existing knowledge and experience of a nascent entrepreneur is, the more will he benefit from business
planning and accordingly, the more will he be inclined to engage in business planning activities. Or as
Argote and Miron-Spektor (2011: 4) put it: “Learning begins with experience”.
2.3 Measuring the Impact of Business Planning
Gruber (2007: 801) suggests that “entrepreneurs need to be efficient planners, and need to know
exactly what to plan in new venture creation […].”, and points out the trade-off decisions
entrepreneurs face when carrying out the organizing activities within nascent entrepreneurial founding
processes. Eventually, when faced with the decision whether to carry out an organizing activity or to
invest time and effort elsewhere, entrepreneurs should explore the option that yields the highest
expected performance given their skills and previous experience. It appears that some entrepreneurs
perceive business planning as a valuable strategy while others do not. Different endowments and
circumstances may or may not render planning worthwhile. As such, the decision to plan is predicated
by this perceived benefit.
Yet, if the work by Gruber (2007) nicely depicts the value entrepreneurs create with effort and time
devoted systematically to planning activities, then why should the general decision to plan or not to
plan be exogenous? It appears that some entrepreneurs perceive business planning as a valuable
strategy while others do not. Thus, under the conditions for one nascent entrepreneur business
planning might be beneficial while under different circumstances learning by doing might be a
preferable alternative. The business planning decision is likely endogenous and thus, self-selected.
Based on these elucidations we bring forward the following hypothesis:
Hypothesis 1: Conditional on the pre-entry experience of the nascent entrepreneur(s) business
planning increases the chances of new venture emergence
2.4 Moderators of business plan effectiveness
2.4.1 Market newness as a moderator
Despite the notion that certain founding activities relate with venture emergence, recent work points
towards a moderating role of market newness for firm performance and correspondingly for the value
creation stemming from planning activities and correspondingly explorative learning (Lee and
Colarelli O´Connor, 2003; Salomo, Weise, and Gemünden, 2007; Kumar, J ones, Venkatesan, and
Leone, 2011). Innovations in newly founded firms generally target unknown and unexplored markets.
Given the uncertainty and ambiguity involved with founding a new venture, nascent entrepreneurs
might abstain from business planning either due to a lack of awareness or time to engage in these
activities (Song, DiBenedetto, and Parry, 2009). In fact, some entrepreneurs develop new and
previously non-existent products that cater emerging markets, while others occupy positions in well-
established markets with known products and obvious competitors. Hence, the availability of
information for making profound decisions is likely to differ.
The effectiveness of planning therefore depends crucially on the degree of market newness. With
greater degrees of newness come stronger uncertainties. Accordingly, information acquired becomes
outdated within a shorter time and hence, projects evolve quicker causing a faster obsolescence of
acquired knowledge through business planning. Hence, the value of business planning as a learning
device is reduced. Consequently, a more adaptive approach, with nascent entrepreneurs reacting to
changes in their environment and new opportunities, might be more appropriate (Wiltbank, Dew,
Read, and Sarasvathy, 2007; Bhidé, 2000). Creativity and adaptive problem solving allows counter-
acting unforeseen contingencies (Camillus, 1975; Mintzberg, 1978). Accordingly, the ability to learn
mitigates environmental threats and enhances the opportunity set (Bhide, 2000; Brinkmann et al.,
2010). Given the uncertainty and ambiguity involved with products that are essentially new to the
market, nascent entrepreneurs should be better off by relying on intuition rather than collecting and
interpreting information that have a short half-life (Sarasvathy, 2001; Bird, 1988). Accordingly, the
value of business planning might be lower. Therefore, we formulate the following hypothesis:
Hypothesis 2: Market newness negatively moderates the value of business planning
2.4.2 Financing Search as a Moderator
Empirical research indicates that an inadequate capitalization of young firms is one of the key reasons
underlying subsequent business failure (Bosma et al., 2004). Evans and J ovanovic (1989) put forward
the notion of liquidity constraints for new ventures. Start-ups require substantial capital to bring new
ventures to life. Accordingly, if personal financial resources of founders are inadequate for launching
the firm, capital markets provide an alternative to obtain financing. Therefore, finding suitable sources
of financing is one of the most critical tasks for nascent entrepreneurs.
Hence, business plans might not only serve the purpose for planning future activities but might also be
addressing needs of third parties, such as potential financiers (Eckhardt, Shane, and Delmar, 2006;
Honig and Karlson, 2004). Given the information asymmetry between entrepreneurs seeking capital
and financiers, business plans act as a signal to reveal firm or founder quality to potential financiers.
This way, the business plan acts as a costly to acquire signal that proxies for the unobservable quality
differences in comparison to other entrepreneurs (Spence, 1974). Following Kirsch et al. (2009)
business plans serve two purposes as signals. In a ceremonial way, business plans disclose information
that document an understanding of the norms of exchange and should therefore legitimize the venture.
Secondly, it might fulfill a communicative role such that it reveals information about the human or
organizational capital, or market and product features. Hence, this information signals the quality of
the venture.
When entrepreneurs write business plans for either purpose the learning effect mentioned previously,
plays only a negligible role. Business plans are mainly written for third party purposes. Moreover, in
their empirical analysis of business plans for venture funding, Kirsch et al. (2009 do not find evidence
that the presence of business plans enhances funding probabilities. They conclude that as business
planning becomes easier, and is heavier emphasized in business courses, textbooks and the media the
costs of the signal become weaker and hence, the value of business planning as a signal is eroded.
Accordingly, spending time on writing business plans for third party needs does not necessarily
enhance the chances of actually receiving financing and might take time away to use planning as a
strategic device for one´s own learning. Accordingly, the value of business planning is lower, ceteris
paribus, when it serves the purpose of acquiring outside financing, rather than acting as a learning tool.
This leads to hypothesis 3:
Hypothesis 3: The ceremonial purpose of business plans for acquiring outside financing
negatively moderates the value of business planning
2.5 Model Summary
Figure 1 presents a stylized depiction of our theoretical model linking pre-entry experience of nascent
entrepreneurs to the decision to draw up a business plan and subsequent performance.
[Insert Figure 1 about here]
Whether or not nascent entrepreneurs benefit from business planning is predicated on their pre-entry
experience. Pre-entry experience increases the ability to learn and therefore the benefits that nascent
entrepreneurs might derive from business planning. We argue that nascent entrepreneurs decide based
on their own attributes whether planning or no planning is beneficial and act accordingly (denoted by
the first linkage 1). We operationalize pre-entry experience by reference to the most widely cited
factors such as formal education (e.g., Evans and Leighton, 1989; Dickson, Solomon, and Weaver,
2008; Unger et al., 2011), labor market experience (e.g., Bosma et al, 2004), and entrepreneurial
experience (e.g., Robinson and Sexton, 1994; Bosma et al, 2004). Pre-entry experience of
entrepreneurs therefore provides the basis for decision-making and hence, explain why certain
entrepreneurs rely on business planning and others do not.
Moreover, different purposes of business planning and the environment in which the learning activity
takes place might moderate the relationship. Hence, we bring forward the notion that when business
plans are written in more fast-moving environments (denoted by 2) due to products that are essentially
new to the market, the information generated through business planning becomes outdated faster and
hence, the positive performance impact of business plans is negatively moderated by the market
newness of a product. Moreover, when business plans are written for the sake of acquiring financing
(denoted by 3 in our graphical depiction of the model), the ceremonial purpose of planning negatively
moderates the positive performance impacts.
Comparing outcomes of the business planning process is tedious since nascent entrepreneurs self-
select their strategies. Therefore, such comparisons cannot assume that strategies were randomly
chosen. Unconditional observations of performance do not provide evidence regarding the benefits of
business planning.
In the following, we argue that the choice to engage in business planning is drawn from a binary set of
options. Following Imbens (2004) and Rubin (1974) let the two options denote for business
planning and for no such activity. The performance, the outcome (perceived emergence or
intentional disbandment) of the nascent entrepreneurial process, is denoted Y. Hence, each strategy
(planning or no planning) results in if W had been chosen. Similarly, if the entrepreneur would
have chosen , the outcome would be observed. To understand the value added of business
planning, we are interested in estimating - , the performance of business planning vs. the
performance of the counterfactual (not planning). The problem with this model set-up is to find an
estimate for the counterfactual. Obviously, we only observe an entrepreneur either opting to plan or
not. Hence, we need to empirically derive the performance for for entrepreneurs that actually chose
. In essence,
might nor provide the counterfactual as entrepreneurs self-select based
on pre-entry experience, and hence the control and treatment are likely to differ on core-characteristics
that matter for new venture emergence.
1
W
0
W
0
W
E
1
Y
1
]
0
Y
]
1
Y
0
Y
0
Y
1
W [
0 0
W | Y
Hence, in our empirical analysis we are interested in estimating the differences in foundation
probabilities conditional on the strategic decision to plan or not. Rubin (1974) refers to this effect as
the “treatment effect” and Nickerson and Hamilton (2003) use the term “strategy effect”. This effect
captures how the performance of a nascent entrepreneur who chose to plan would have changed, had
he not planned . The question that we are interested in the following is not
whether business planning had a positive impact for the entrepreneur. Rather, we address how he
[ ] [
1 0 1 1
W | W | Y E Y E ?
would have performed had he not planned. Hence, we are interested in estimating the differences in
performance conditional on the strategic decision to plan or not. Thus, we can estimate how
performance changes conditional on the entrepreneur’s choices.
3 Methodology
3.1. Dataset
To test our hypotheses empirically, we draw on the Second Panel Study of Entrepreneurial Dynamics
(PSED II) dataset. The PSED II is a representative survey of entrepreneurial activities in the United
States that portrays individuals during their business creation process. The dataset describes the
characteristics of nascent entrepreneurs, documents the sequences of the organizing activities,
summarizes the types and volumes of resources committed, and characterizes the new ventures.
1
For PSED II individuals were identified in late 2005 (early 2006), with 4 recurrent follow up
interviews, each taking place after 12 months subsequently. The last wave was therefore completed in
J anuary 2010. The sample of active nascent entrepreneurs was sampled from an overall group of
31,845 individuals. Out of this probability sample, 1,214 active nascent entrepreneurs were identified
in an initial interview. Interviewees were identified through their answers on screening questions as to
whether they were intending to start a new firm, already carried out at least one start-up activity in past
years, were expected to own part of the firm, and did not have an already going business. Hence,
entrepreneurs eligible for the sample were involved with an ongoing, but not yet operational start-up.
These eligible cases are re-interviewed every 12 months over the course of five years. The early stage
screening was used to ensure representativeness of the data and more importantly, to reduce
distortions caused by potential survivorship biases. Throughout the data collection process,
respondents give affirmative answers concerning a package of start-up activities that reveal the
progress they make in terms of becoming operational. Given the re-interviewing over the course of
five years the resulting longitudinal structure with monthly indications of activities started and
1
Detailed descriptions for the methods and sampling used to generate PSED II and an overview on the data structure can be found in
Reynolds and Curtin (2009).
finished allows for inferences on the process of organizing activities and facilitates causal inferences
among dependent and independent variables.
Concerning the number of respondents, wave A identifies 1,214 nascent entrepreneurs that returned
the questionnaire, whilst the number drops subsequently due to non-respondents and disbandment to
972 for Wave B, and 746, 526, and 435 for the Waves C to E. Among the start-ups, 512 disbanded
their venture and 243 perceived their venture as operational. For the remainder of the other ventures,
180 reported ongoing activities as per wave 5, but did not perceive their venture as operational. Lastly,
279 start-ups out of the initial sample were omitting at least one wave or stopped to report their
progress entirely. Hence, for these entrepreneurs we do not know the whereabouts of their organizing
activities, and consequently, these observations are pruned. Hence, we start our empirical analysis
with a total of 935 observations.
Selection correction and covariate imbalance adjustment
Given the endogenous choice based on pre-entry experience, the entrepreneur who did not plan cannot
necessarily provide the counterfactual as a comparative difference based on pre-entry experience could
make business planning beneficial for the one, but not for the other. Moreover, if pre-entry experience
affects the likelihood of business planning and the chances for venture emergence simultaneously, the
error terms of the decision to plan correlate with the outcome regression (new venture formation)
(Dencker et al., 2009; Unger et al., 2011). Thus, coefficients from regressions that fail to accommodate
for endogenous choices are likely biased which eventually causes facile inferences (Hamilton and
Nickerson, 2003).
Simply regressing a dummy variable with the business planning decision on the outcome of the
process is only valid if entrepreneurs make mistakes (and hence, strategies are random) or if all factors
driving the outcome are observable (Shaver, 1998). Both conditions are unlikely to be satisfied for
non-experimental data. Even if cognitive heuristics are present in start-up environments (Busenitz and
Barney, 1997) and decision-making errors are common and cause severe biases (Franke et al., 2006)
decisions are not purely exogenous.
Hence, inferring the counterfactual from the control groups that are present in the data might lead to
erroneous results due to self-selection. Accordingly, we need to control for the imbalance in pre-entry
experience across the treatment (business planning) and control (no business planning) group. To
ensure that we can infer the counterfactual from the control group we employ coarsened exact
matching (CEM) (Iacus, King, and Porro, 2011; Singh and Agrawal, 2011). Matching helps to control
for some (or all) pretreatment confounding covariates and makes the empirical estimation less model
dependent, prunes observations that have no close match within both the control and treatment group,
and removes heterogeneity. Hence, it ensures a better balance between the treatment and control group
and consequently the covariates in both groups are more similar and thus, allow inferences despite
potential pretreatment endogeneity (caused by the self-selection into planning and no planning,
respectively).
As opposed to other matching methods, CEM reduces the covariate imbalance between the treatment
group (business planning) and the control group (no business planning) without tedious balance re-
checking (and likely model re-estimation). CEM specifies balance ex-ante and reduces model
dependence and the error of the treatment effect estimation (Iacus, King, and Porro, 2011; Ho et al.,
2007). Data is temporarily coarsened into meaningful groups, matched exactly on the coarsened data
(bins) and uses the actual (uncoarsened) data for the matched observations in the empirical analysis of
causal effects. Accordingly, the ex-ante specification ensures common support (all matched
observations within bins are by definition in the area of common support) and this need not be tested
ex-post. Moreover, CEM reduces model dependence and the balance need not be achieved through an
empirical model, i.e. linear regression or maximum likelihood estimator (Abadie et al., 2004). We rely
on the implementation in Stata, to choose the bins based on Scott´s rule (1992) rather than providing
chosen cut-off points (Iacus, King, and Porro, 2011; Marx, Singh, and Fleming, 2010).
Dependent Variable
Completion of entrepreneurial organizing activities
Researchers have discussed a wide variety of measures to determine the point at which a nascent
venture shifts from the end of the entrepreneurial organizing activities to an operational business:
ability of raising external money, legal establishment of the new venture, first sales, positive cash
flow, reaching the break-even point, etc. (Gartner and Carter, 2003; Davidsson and Gordon, 2011). In
the following we will discuss occurrence of a first positive cash flow combined with a self-reported
measure of being operational (or disbanding effort), the probability of becoming operational in the
presence of the option to terminate efforts, and making comparisons across the most appropriate group
level of analyses.
Bygrave (1989) asserts that the only way to know whether the new venture will generate a persistent
business is to wait until the new venture is generating positive cash flows. Other researchers claim that
there are limitations of cash flow as a measure for completing the entrepreneurial organizing activities.
Bhide (2000) argues that cash flow is not likely to be an early goal of most high-potential new
ventures. Katz and Cabezuelo (2004) make a case, that nascent entrepreneurs are not always
sophisticated enough to calculate positive cash flows exactly. Moreover, entrepreneurs might have
different objectives (Delmar and Shane, 2006; Gruber, MacMillan, and Thompson, 2008). As the
industry composition reveals, we mostly deal with “lifestyle”-ventures that are “low-tech”. Hence,
sustainability is important as it signals the ability to generate an income stream and a general level of
market acceptance for the entrepreneurial venture.
Given the debate on the applicability of researcher defined and self-reported outcome measures, we
employ a more prudent outcome measure by using information whether the entrepreneurs reached a
first positive cash flow (covering managerial salaries) and also perceived their venture as operational,
and compare these with entrepreneurs that indicated that they terminated their work on the venture.
Becoming operational is therefore a combination of question A35 and A41 asking whether monthly
revenues ever exceeded monthly expenses (including salaries for the managers) and whether based on
this achievement, the respondents would characterize their venture as being operational. If both
conditions are fulfilled the variable takes on the value of one. We label this outcome as “perceived
emergence”. Moreover, following Delmar and Shane (2003) we also investigate whether all involved
entrepreneurs report disengagement to allow comparison across these two reported statuses. If others
are still working on the venture, we did not treat these as disbanded as solely the key respondent
disengaged from the venture, which is distinct from disbandment. If all reported start-up members
disengaged from the venture the variable takes on the value of zero for the comparison category and if
others are still working we omit the data due the “still trying” nature of the observation. Hence, we fit
a logistic regression with CEM weights using the outcome perceived emergence against venture
disbandment as the dependent variable.
Event history model
In the following, we make also use of the longitudinal character of the dataset to estimate the impact of
business planning on the hazard to perceive the venture as emerged. We use the information provided
by nascent entrepreneurs on the months in which they carried out their organizing activities and the
month in which they either disbanded their organizing efforts or when their monthly revenues
exceeded expenses (including managerial salaries) for the first time and they perceived their venture as
emerged.
Venture disbandment takes place, when all parties identified to be working on the venture stop their
entrepreneurial activities. Hence, there is no person involved that follows through with the original
plan to bring this particular venture to life. Individual respondents were asked whether the disengaged
from the venture. Moreover, they indicated that in some cases other members of the team were still
continuing to work on the venture. Only if all members were disengaged from the venture, we treated
the venture as disbanded.
However, disbanding efforts and continuation might in part be related to the same underlying factors.
Business planning serves a two-fold purpose in this context. On the one hand, it helps to shape and
develop the actual business opportunity the entrepreneur wants to pursue, and on the other hand it
assists to evaluate the opportunity and to weed out good from bad alternatives. Information generated
from business planning effort can thus go both ways, providing positive feedback and negative
feedback alike. As such, disengagement of the venture might not always be a sign of bad performance,
but could also be an indication that some entrepreneurs question the feasibility of their aspired
business model and pursue alternative options in lieu of the initial venture. Consequently, the longer a
nascent entrepreneur is in the process of organizing a new venture, the higher will be the chances to
become operational (due to the effort put into organizing activities) and similarly the risk of
disbandment (if the effort does not show any material impact on venture emergence). In contrast to
Delmar and Shane (2003) we therefore treat the disbandment of ventures as a competing risk, rather
than a pure censoring event (Fine and Gray, 1999). In fact, disbanding the venture organizing efforts is
an endogenous decision that prevents further business planning and hence, prunes the number of
observations.
The disbandment event is distinct from the classical analysis of survival using exit as a censoring
event. Censored events generally are treated as being “at risk”, whilst the researcher cannot observe
what happens to the individual under investigation. However, for the case of intentional venture
disbandment we know with certainty, that the venture will not become operational as the entrepreneur
gave up working on the venture. Hence, venture disbandment is permanent and prevents the venture
from becoming operational. Censoring obstructs the researcher from observing the event, whilst the
competing event disbandment prevents the event “perceived emergence” from occurring.
Consequently, we treat the month in which the respondent disengaged as a competing risk to the
perception of emergence measure in the event history analysis. The use of competing risks controls for
inherent differences among entrepreneur characteristics that link to both, the chances to become
operational and the risk of disbandment, such as pre-entry experience and the quality of the underlying
opportunity pursued.
We take into consideration the month in which the ventures undertook their first activity and the
months, in which they either perceived their venture as emerged or disbanded their venture organizing
efforts, the hazard therefore has a monthly spell. A positive coefficient indicates that an increase in the
covariates for the business planning variables increases the hazard of becoming operational. Given the
combination of emerged, terminated and in-process ventures, this analysis also allows to distinguish
among different gradations of venture emergence that overcome limitations of right censoring (Dimov,
2010). Moreover, we employ coarsened exact matching first to reduce covariate imbalance across the
treatment and control group.
Sample Comparison and Complementary Analyses
Censoring: To measure successful completion of the founding process, we also accommodate
for potential censoring problems. Some entrepreneurs might have started a long time prior to the
interview and, consequently, have higher chances to reach certain milestones before others.
Specifically, firms that started prior to 2005 had more time to prepare the venture. Hence, the 5 years
covered in the data set might not be sufficient to compare the different groups and truncation periods.
Gartner and Carter (2003) and Lichtenstein et al. (2007) therefore suggest including only
entrepreneurs who underwent their first activity within 24 month prior to interview time. The approach
is similar to Delmar and Shane (2003) that included only start-ups that underwent activities within the
year of the interview to accommodate potential censoring and selection biases. Accordingly, we
reduce the sample further by excluding 128 entrepreneurs that did not pursue their first activity within
the time span covering the 24 month prior to the first interview to wave E.
Moreover, Davidsson and Gordon (2011) report the existence of “dilettante dreamers” or hobbyists in
nascent entrepreneurial sample. Accordingly, to rule out entrepreneurs that are not serious about their
start-up intentions and to make the nascent phase comparable across start-ups Davidsson and Gordon
(2011) suggest to compare start-ups without the “still trying”-category to avoid distortions by less
serious entrepreneurs (Parker and Belghitar, 2006). We report our first set of results using a
comparison of those entrepreneurs that perceived their venture as emerged and those that indicated
that they disbanded their effort. Hence, this analysis excludes further 154 entrepreneurs that report a
“still trying” status as per wave E. Moreover, we report the survival analysis making use of the time-
stamp information concerning all activities that nascent entrepreneurs undertook to measure the
probability of becoming operational in the presence of the option to disband. Here we explicitly make
use of the three different categories.
Predictor Variables
We use several measures to construct the corresponding treatment and control groups. First, we
include all nascent entrepreneurs who indicate that they started to work on a business plan; secondly,
we use the group of nascent entrepreneurs that indicated that they successfully finished a version of the
business plan. Third, PSED II reports whether the finished version of the business plan is formal or
informal. Hence, we also define a treatment group of nascent entrepreneurs that finished a formal
version of the business plan. Fourth, to capture the formal nature of the business plan in more detail,
we also construct a treatment group for the entrepreneurs that made financial projections. Fifth, during
every phase, entrepreneurs indicated whether they made changes to the business plan. To capture the
dynamic nature of business planning, we define yet another treatment group comprising of nascent
entrepreneurs that made changes to their business plan. Hence, we have five different proxies to
capture endogenous business planning strategies as a treatment effect. We follow the extant literature
(Delmar and Shane, 2003; Gruber, 2007; Dimov, 2010) and administer the business planning variables
in dichotomous form.
Conditioning Variables
We match entrepreneurs who decided to plan and those who decided not plan, based on their pre-entry
experiences. We operationalize pre-entry experience of nascent entrepreneurs by reference to
the most widely cited factors such as formal education (e.g., Evans and Leighton, 1989;
Dickson, Solomon, and Weaver, 2008; Unger et al., 2011), labor market experience (e.g.,
Bosma et al, 2004), and entrepreneurial experience (e.g., Robinson and Sexton, 1994; Bosma
et al, 2004). A principal factor analysis (with varimax rotation) yields a clear factor structure
(with eigenvalues for each factor greater than one) for these three measures. The variables are
calculated for solo-entrepreneurs and team-foundations. In the latter cases, we include the
average levels of these factors.
Formal Education: To collect PSED II data, respondents were asked to indicate the highest
level of education all members of the entrepreneurial team had completed. We recoded this
variable, ranging from elementary school to PhD, into number of years of education (see e.g.
Davidsson and Honig, 2003; Iacus, King, and Porro, 2011).
Labor Market Experience: PSED II provides information about the years of work
experience in the industry a new venture is active in, years of full-time paid work experience,
and years of managerial, supervisory, or administrative responsibilities of the nascent
entrepreneurs. Cronbachs alpha is 0.72 for the three factors.
Entrepreneurial Experience: We use information on the number of other businesses
previously helped to start as an owner, and the number of other businesses they own(ed). The
correlation among these two variables is 0.6 (which would result in Cronbach’s alpha of
0.65).
Moderators:
Seek Financing: As Business Plans might not only serve the purpose of planning future activities
but might also be addressing needs of third parties, such as potential financiers, the analysis accounts
for the impact of entrepreneurs writing business plans to seek outside funding. We account for whether
or not the entrepreneurs did actively seek financial capital. The variable is a dichotomous measure that
equals “1” if entrepreneurs seek outside financing, and zero otherwise.
Market Newness: We measure the perception of market newness by using the answer to the
question whether the product is unfamiliar to all, some or none of the potential customers. The
variable uses a three-point scale. The scale is: “3” equals “all customers will be unfamiliar with this
new product or service”, “2” equals “some customers will be unfamiliar with this new product or
service”, and “1” equals “none of the customers will be unfamiliar with this new product or service”.
The variables are also in part included in the index developed in Dahlqvist and Wiklund (2011) and
represent newness of a product to customers. To ensure the appropriateness of the index to measure
whether a product is new to customers, we regress the measure on the probability to engage in the
development of a proprietary technology or process or whether entrepreneurs apply for a patent or get
a patent granted. All coefficients for the scale are positive and highly significant at the 0.1% level.
Hence, we are confident that this measures in fact proxies for market newness of a product.
Control Variables
Industry: Since the role of business planning can differ across industries we parce out these
effects by including industry dummy variables. We control for retail (86 firms), consumer services
(214), health (50), consulting (43) manufacturing and construction (73), real estate and finance (61)
and other industries (100). We omit “other industries” as the reference group in each regression to
avoid perfect collinearity. Inclusion of industry dummies is indicated in the corresponding table.
Age: We control for the age of the entrepreneurs using the average over all team
members as indicated in wave A.
Team Size: We control for team size using all team members as indicated by the respondent
to the questionnaire in wave A.
Competition: We measure the perception of competition using a three-point scale. The scale is:
“3” equals “there are many other businesses offering the same product or service”, “2” equals “there
are few other businesses offering the same product or service”, “1” equals “there are no other
businesses offering the same product or service”.
Motivation: As the motivation to start a business might vary among entrepreneurs, we follow
Dimov (2010) and include questions from PSED II on entrepreneurial motivations into our empirical
analysis. The measure of start-up motivation comprises the answer to the questions “There is no limit
as to how long I would give maximum effort to establish my business” and “My personal philosophy
is to do whatever it takes to establish my own business”. Combining both variables would result in
Cronbach´s alpha of 0.71 (comparable to Dimov (2010).
Self-Efficacy: We measure perception of self-efficacy using the five questions identified in
Dimov (2010). The questions employ a five-point Likert-type scale. A confirmatory factor
analysis reveals that three of the five items load on one factor and result in satisfactory
Cronbach´s alpha of 0.68 (corresponds to questions AY4 to AY8 from PSED 2). One could
improve the fit by using only the questions AY6 to AY8 to 0.72. Nevertheless, we followed
the recent approach in Dimov (2010) to make our results comparable to previously published
studies and employ a measure comprising all five items. The original PSED 2 scale is inverted
so that higher values indicate higher levels of self-efficacy.
4. Results
Table 1 summarizes the details for the underlying sample of this study including 627 nascent
entrepreneurial ventures. Among these, 23% considered their venture to be operational while some
56% disbanded their organizing efforts. These performances within PSED II are similar to other
studies analyzing new firm creation in the US (Spletzer et al., 2004; Reynolds, 2009). Concerning the
prevalence of business planning around three quarter of the ventures engaged in business planning,
while half of them indicated that they finished their planning activities subsequently. Additionally,
only 20% finished a first formal version of the business plan whilst around 50% produced
accompanying financial projections. Lastly, around every third respondent indicated that changes were
made to the first version of the business plan.
[Insert table 1 about here]
Table 2 reports logistic regressions with the five different dimensions of business planning as
explanatory variables. The outcomes variable is equal to 1 when entrepreneurs perceive their venture
as emerged and zero when all people involved disbanded their effort. Four out of the five dimensions
of business planning positively affect the perceived emergence of a venture. Starting to plan has no
significant impact on the emergence of a nascent venture (ß =0.072). Also, those entrepreneurs that
finish the process of business planning (ß =0.121, p
During this such a detailed outline about strategic choices in new venture development and the value of business planning.
DISCUSSION PAPER
SERIES IN
ECONOMICS AND
MANAGEMENT
Strategic choices in new venture development and
the value of business planning for nascent
entrepreneurs
C. Hopp
Discussion Paper No. 11-26
GERMAN ECONOMIC ASSOCIATION OF BUSINESS
ADMINISTRATION – GEABA
Strategic choices in new venture development and the value of business planning for nascent
entrepreneurs
Preliminary Version: Please do not cite without the authors permission
July 2011
Dr. Christian Hopp
University of Vienna
Chair for International Personnel Management
Brünner Straße 72, 3/241
Tel: +43-1- 4277-38166
Fax: +43-1- 4277-38164
E-Mail: [email protected]
Strategic choices in new venture development and the value of business planning for nascent
entrepreneurs
Abstract
Theory suggests a positive relationship of business planning with new venture formation. Yet,
empirical evidence is scarce, or ambiguous. In contrast to previous empirical studies, we treat the
decision to plan or not to plan as an endogenous strategic choice by entrepreneurs. Using data from the
PSED 2 study, we empirically correct for self-selection and account for endogeneity. Bearing in mind
that business plans serve a specific purpose, among others either for learning and guidance of the
entrepreneur or for third parties to attract outside capital, different forms of business planning are
beneficial for entrepreneurs. Business planning per se does not have a statistically significant impact
on the chances to found a new venture successfully. Rather, it is the process of planning and
subsequent learning through planning that matters for successful venture emergence. However, when
entrepreneurs are seeking outside funding, making financial projections is affecting the founding
probability positively, but not the ceremonial purpose of business plans. Moreover, market newness
might moderate the value of business planning for nascent entrepreneurs. When information becomes
outdated more quickly, the value of plans for learning purpose is reduced. We conclude that
understanding when business plans are needed and for which purposes, is crucial in empirically
disentangling performance impacts and providing practical advice.
1. Introduction
The value of business planning is widely debated in the academic literature. Some highlight the
advantage of bringing some structure amidst the chaos of uncertainty and ambiguity when founding a
new venture (Delmar and Shane, 2003; Gruber, 2007). Others suggest that nascent entrepreneurs
should spend their time on more relevant organizing activities, among others focusing on gaining
legitimacy and establishing contacts with customers (Bird, 1988; Bhidé, 2000). Generally, it is unclear
whether business planning is key to success or distracting time from other, more critical
entrepreneurial tasks.
In this work, we argue that business planning is self-selected and, therefore, endogenous. Subsequent
organizing activities can be traced back to the individuals pursuing these actions reflecting their
inherent resources or their preparatory actions to create the new venture. Different entrepreneurs may
face similar (or even the same) opportunities, yet organizing patterns differ (Bowman and Hurry,
1993; Baker and Nelson, 2005). Specifically, perceptions of entrepreneurs provide the basis for
decision-making. Hence, they explain why certain entrepreneurs rely on business planning and others
do not. Consequently, estimating strategic choices in empirical entrepreneurial research by neglecting
the endogenous nature of such decisions leads to erroneous results.
We document that endogenous strategic choices with respect to business planning based on pre-entry
experience can help to explain entrepreneurial success and/or failure. We extend the ideas put forward
in Hamilton and Nickerson (2003) by modeling the strategic choice and by accommodating for
imbalances caused by individuals self-selecting into planning activities. Thus, we can estimate how
performance changes conditional on the entrepreneur’s choices. In particular, we match the pre-entry
experience of entrepreneurs with the choice for (or against) business planning (and the extent of it).
Thus, we infer the counterfactual to estimate the performance differential caused by strategic decisions
made. Moreover, we emphasize the conditions under which certain opportunities lead to higher
performance. In sum, we demonstrate that empirical studies must appropriately account for the
strategic choices made by entrepreneurs.
Our results show that contrary to previous findings, starting to draw up a business plan does not affect
new venture emergence. Prescriptive advice that focuses on initiating business planning solely does
not constitute a valuable strategy for nascent entrepreneurs. In general, finishing the business plan
carries more value when we match nascent entrepreneurs that plan with those that did not plan.
Moreover, producing a formal business plan and making financial projections add significant value for
nascent entrepreneurs, whilst subsequently reworking the plan does not increase the chances to
perceive the venture as operational/emerged.
However, the purpose of the underlying planning activities significantly moderates the value created
by business planning. While some planning activities appear to be valuable for entrepreneurs that did
not seek outside financing, only financial projections are valuable for entrepreneurs that indicated that
they actively seek outside capital. Hence, when business plans are written for a mere ceremonial
purpose the value stemming from learning is reduced. This finding provides evidence that business
plans satisfy a ceremonial purpose or third party needs. Moreover, when business plans are written for
products that are essentially new to all potential customers, none of the business planning activities
creates value and affects the chances to perceive the venture as emerged. Hence, we conclude that
when information becomes outdated more quickly (or might simply be not available to be worked into
plans) learning from business planning activities cannot take place, and therefore business planning
does not add significant value for nascent entrepreneurs.
Accordingly, our results provide insights for research and practice alike. On the one hand, we
highlight the role of business planning as a dynamic and emergent tool that needs to be adapted
continuously in order to be successful. While starting to plan does not add value, it presents the entry
into the process of business planning. However, it can only add value if nascent entrepreneurs are
willing to accept business planning as a process comprising a series of activities, that they are willing
to undertake. On the other hand, we document that advice needs to be appropriately tailored to
entrepreneurial situations. The advice to plan should not be purely prescriptive and focused on the idea
of having a one-size-fits-all business plan. Rather it should appreciate the contingent nature of
planning itself, mirroring the ever-changing environment that eventually finds its way into the formal
business plan and more importantly, needs to work in the actual purpose of why nascent entrepreneurs
want to engage in planning activities in the first place.
The remainder of this article is structured as follows: section two outlines the theoretical background
and derives the hypotheses. Section three describes the data and presents the methodology. Results are
reported in section four. Section five discusses the findings, the studies potential limitations, and
provides concluding remarks.
2 Theoretical Background
2.1 The Role of Business Planning in Nascent Entrepreneurship
Very essentially, entrepreneurs pursue potentially profitable investment opportunities and the build of
organizations to actually reap the benefits of their effort (Haber and Reichel, 2007; Gartner, 1985).
While plagued by high levels of uncertainty and ambiguity, they have to structure a complex set of
interrelated tasks. Substantial discretion in entrepreneurial decisions is necessary to take the
appropriate organizing activity.
Drawing up a business plan is among the most recognized of such activities. It involves the collection
and analysis of relevant information to identify future tasks, risks, opportunities, and to derive viable
contingencies for future actions, likely in a written format (Delmar and Shane, 2003, Gruber, 2007,
Brinkmann et al., 2010). The value of a business plan for nascent entrepreneurs is widely debated in
the academic literature and the impact on future success, remains ambiguous.
Following Dencker et al. (2009) business planning is a learning tool, that comprises learning from
others and cognitive search. Business planning can facilitate learning by helping to identify and detect
patterns and to draw meaningful conclusions from them. In general, it comprises reachable goals and
situational analysis and decision making. In this vein, a pro-active business plan is necessary to
develop and interpret future information (Frese et al., 2007). Hence, business planning in the nascent
phase, that is prior to the launch of a new organization, involves the ex-ante collection and analysis of
information and subsequently helps to get an understanding of the requirements to be successful as an
organization. That is, business planning provides objectives and necessary steps to reach goals.
Therefore, rather than providing a clear plan to which entrepreneurs stick, business planning is a
process of learning about the business model, the customer requirements, and eventually the building
of an organization and hence, guides entrepreneurial actions. Consequently, business planning changes
the range of potential behaviors and therefore is an essential tool in the learning process of nascent
entrepreneurs (Dencker et al., 2009; Huber, 1991).
Business planning might reduce part of the uncertainty by deriving action plans and facilitating
decision-making and leads to a more effective organization of the founding process and, therefore to a
higher probability of success (Ansoff, 1991; Delmar and Shane, 2003). Planning could provide
entrepreneurs with a framework to structure tasks and get an overview of the overall structure of
entrepreneurial activities to reduce erratic behavior. The benefits of business planning stem from
predictions about future challenges and the corresponding implementation of milestones and allows
for more effective use of scarce resources. The quality of decisions and the allocation of resources
improves due to better information derived from previous organizing steps taken (Delmar and Shane,
2003; Brinkmann et al., 2010). Entrepreneurs become aware of options and strategic choices in
navigating through the founding process (Ansoff, 1991, Delmar and Shane, 2003). Moreover, business
planning provides a benchmark to test previously made assumptions against and therefore enables
entrepreneurs to overhaul strategic approaches if predicted outcomes do not materialize. Overall,
planning should be related to a more systematic way of organizing the foundation process and thus
foster goal achievement (Gruber, 2007).
2.2 Existing Knowledge and Experience as Antecedents of Business Planning
Complex tasks such as business planning need to rely on cognitive skills to allow for effective
planning given the interrelatedness of many facets. Chwolka and Raith (2011) refer to business
planning as a collection of skills that can be trained and that are inherently linked to the decision
maker’s prior knowledge and experience. Prior related knowledge is essential in scanning the external
environment to identify external knowledge and make do with the accessed new knowledge
(Lichtenthaler, 2009). Skills and knowledge provide the existence of absorptive capacity that enables
an entrepreneur to acquire and transfer new information collected in the process of business planning
into actions (Cohen and Levinthal, 1990; Zahra and George, 2002). Entrepreneurs that possess more
pre-entry experience will derive larger benefits from business planning. They already have experience
with general planning activities and hence, are equipped with “a structure for how to plan as well as
practice with the cognitive act of planning” (Dencker et al., 2009: 520). Skills and knowledge
therefore are related to proactive planning and equip people with skills to deal with new and complex
situations. Hence, they help in acquiring knowledge and facilitate decision making. Pre-entry
experience acts as an antecedent to elaborate and proactive planning (Frese et al., 2007).
Thus, we argue that pre-entry experience of nascent entrepreneurs serves as a basis for future learning.
During the process of business planning, pre-entry experience interacts with the organizational context
to create new knowledge and helps entrepreneurs to act accordingly. Gathering information through
business planning and processing them subsequently is greatly eased by the existence of pre-entry
industry and managerial experience. Thus, knowledge repositories enable entrepreneurs to plan more
effectively by knowing what questions to ask, how to analyze and most importantly, how to interpret
their findings to derive actions from them. Existing knowledge facilitates comprehending and
integrating new information. Consequently, initial endowments of nascent entrepreneurs influence
their adaptability and more experience subsequently gives more margins to maneuver (Huber, 1991;
Argote and Miron-Spektor, 2011; Dencker et al., 2009).
In essence, higher levels of pre-entry experience of nascent entrepreneurs increase the benefits from
subsequent learning activities. Entrepreneurs would choose to learn from business planning or forego
formal planning based on their attributes and experiences. Consequently, ceteris paribus, the higher the
existing knowledge and experience of a nascent entrepreneur is, the more will he benefit from business
planning and accordingly, the more will he be inclined to engage in business planning activities. Or as
Argote and Miron-Spektor (2011: 4) put it: “Learning begins with experience”.
2.3 Measuring the Impact of Business Planning
Gruber (2007: 801) suggests that “entrepreneurs need to be efficient planners, and need to know
exactly what to plan in new venture creation […].”, and points out the trade-off decisions
entrepreneurs face when carrying out the organizing activities within nascent entrepreneurial founding
processes. Eventually, when faced with the decision whether to carry out an organizing activity or to
invest time and effort elsewhere, entrepreneurs should explore the option that yields the highest
expected performance given their skills and previous experience. It appears that some entrepreneurs
perceive business planning as a valuable strategy while others do not. Different endowments and
circumstances may or may not render planning worthwhile. As such, the decision to plan is predicated
by this perceived benefit.
Yet, if the work by Gruber (2007) nicely depicts the value entrepreneurs create with effort and time
devoted systematically to planning activities, then why should the general decision to plan or not to
plan be exogenous? It appears that some entrepreneurs perceive business planning as a valuable
strategy while others do not. Thus, under the conditions for one nascent entrepreneur business
planning might be beneficial while under different circumstances learning by doing might be a
preferable alternative. The business planning decision is likely endogenous and thus, self-selected.
Based on these elucidations we bring forward the following hypothesis:
Hypothesis 1: Conditional on the pre-entry experience of the nascent entrepreneur(s) business
planning increases the chances of new venture emergence
2.4 Moderators of business plan effectiveness
2.4.1 Market newness as a moderator
Despite the notion that certain founding activities relate with venture emergence, recent work points
towards a moderating role of market newness for firm performance and correspondingly for the value
creation stemming from planning activities and correspondingly explorative learning (Lee and
Colarelli O´Connor, 2003; Salomo, Weise, and Gemünden, 2007; Kumar, J ones, Venkatesan, and
Leone, 2011). Innovations in newly founded firms generally target unknown and unexplored markets.
Given the uncertainty and ambiguity involved with founding a new venture, nascent entrepreneurs
might abstain from business planning either due to a lack of awareness or time to engage in these
activities (Song, DiBenedetto, and Parry, 2009). In fact, some entrepreneurs develop new and
previously non-existent products that cater emerging markets, while others occupy positions in well-
established markets with known products and obvious competitors. Hence, the availability of
information for making profound decisions is likely to differ.
The effectiveness of planning therefore depends crucially on the degree of market newness. With
greater degrees of newness come stronger uncertainties. Accordingly, information acquired becomes
outdated within a shorter time and hence, projects evolve quicker causing a faster obsolescence of
acquired knowledge through business planning. Hence, the value of business planning as a learning
device is reduced. Consequently, a more adaptive approach, with nascent entrepreneurs reacting to
changes in their environment and new opportunities, might be more appropriate (Wiltbank, Dew,
Read, and Sarasvathy, 2007; Bhidé, 2000). Creativity and adaptive problem solving allows counter-
acting unforeseen contingencies (Camillus, 1975; Mintzberg, 1978). Accordingly, the ability to learn
mitigates environmental threats and enhances the opportunity set (Bhide, 2000; Brinkmann et al.,
2010). Given the uncertainty and ambiguity involved with products that are essentially new to the
market, nascent entrepreneurs should be better off by relying on intuition rather than collecting and
interpreting information that have a short half-life (Sarasvathy, 2001; Bird, 1988). Accordingly, the
value of business planning might be lower. Therefore, we formulate the following hypothesis:
Hypothesis 2: Market newness negatively moderates the value of business planning
2.4.2 Financing Search as a Moderator
Empirical research indicates that an inadequate capitalization of young firms is one of the key reasons
underlying subsequent business failure (Bosma et al., 2004). Evans and J ovanovic (1989) put forward
the notion of liquidity constraints for new ventures. Start-ups require substantial capital to bring new
ventures to life. Accordingly, if personal financial resources of founders are inadequate for launching
the firm, capital markets provide an alternative to obtain financing. Therefore, finding suitable sources
of financing is one of the most critical tasks for nascent entrepreneurs.
Hence, business plans might not only serve the purpose for planning future activities but might also be
addressing needs of third parties, such as potential financiers (Eckhardt, Shane, and Delmar, 2006;
Honig and Karlson, 2004). Given the information asymmetry between entrepreneurs seeking capital
and financiers, business plans act as a signal to reveal firm or founder quality to potential financiers.
This way, the business plan acts as a costly to acquire signal that proxies for the unobservable quality
differences in comparison to other entrepreneurs (Spence, 1974). Following Kirsch et al. (2009)
business plans serve two purposes as signals. In a ceremonial way, business plans disclose information
that document an understanding of the norms of exchange and should therefore legitimize the venture.
Secondly, it might fulfill a communicative role such that it reveals information about the human or
organizational capital, or market and product features. Hence, this information signals the quality of
the venture.
When entrepreneurs write business plans for either purpose the learning effect mentioned previously,
plays only a negligible role. Business plans are mainly written for third party purposes. Moreover, in
their empirical analysis of business plans for venture funding, Kirsch et al. (2009 do not find evidence
that the presence of business plans enhances funding probabilities. They conclude that as business
planning becomes easier, and is heavier emphasized in business courses, textbooks and the media the
costs of the signal become weaker and hence, the value of business planning as a signal is eroded.
Accordingly, spending time on writing business plans for third party needs does not necessarily
enhance the chances of actually receiving financing and might take time away to use planning as a
strategic device for one´s own learning. Accordingly, the value of business planning is lower, ceteris
paribus, when it serves the purpose of acquiring outside financing, rather than acting as a learning tool.
This leads to hypothesis 3:
Hypothesis 3: The ceremonial purpose of business plans for acquiring outside financing
negatively moderates the value of business planning
2.5 Model Summary
Figure 1 presents a stylized depiction of our theoretical model linking pre-entry experience of nascent
entrepreneurs to the decision to draw up a business plan and subsequent performance.
[Insert Figure 1 about here]
Whether or not nascent entrepreneurs benefit from business planning is predicated on their pre-entry
experience. Pre-entry experience increases the ability to learn and therefore the benefits that nascent
entrepreneurs might derive from business planning. We argue that nascent entrepreneurs decide based
on their own attributes whether planning or no planning is beneficial and act accordingly (denoted by
the first linkage 1). We operationalize pre-entry experience by reference to the most widely cited
factors such as formal education (e.g., Evans and Leighton, 1989; Dickson, Solomon, and Weaver,
2008; Unger et al., 2011), labor market experience (e.g., Bosma et al, 2004), and entrepreneurial
experience (e.g., Robinson and Sexton, 1994; Bosma et al, 2004). Pre-entry experience of
entrepreneurs therefore provides the basis for decision-making and hence, explain why certain
entrepreneurs rely on business planning and others do not.
Moreover, different purposes of business planning and the environment in which the learning activity
takes place might moderate the relationship. Hence, we bring forward the notion that when business
plans are written in more fast-moving environments (denoted by 2) due to products that are essentially
new to the market, the information generated through business planning becomes outdated faster and
hence, the positive performance impact of business plans is negatively moderated by the market
newness of a product. Moreover, when business plans are written for the sake of acquiring financing
(denoted by 3 in our graphical depiction of the model), the ceremonial purpose of planning negatively
moderates the positive performance impacts.
Comparing outcomes of the business planning process is tedious since nascent entrepreneurs self-
select their strategies. Therefore, such comparisons cannot assume that strategies were randomly
chosen. Unconditional observations of performance do not provide evidence regarding the benefits of
business planning.
In the following, we argue that the choice to engage in business planning is drawn from a binary set of
options. Following Imbens (2004) and Rubin (1974) let the two options denote for business
planning and for no such activity. The performance, the outcome (perceived emergence or
intentional disbandment) of the nascent entrepreneurial process, is denoted Y. Hence, each strategy
(planning or no planning) results in if W had been chosen. Similarly, if the entrepreneur would
have chosen , the outcome would be observed. To understand the value added of business
planning, we are interested in estimating - , the performance of business planning vs. the
performance of the counterfactual (not planning). The problem with this model set-up is to find an
estimate for the counterfactual. Obviously, we only observe an entrepreneur either opting to plan or
not. Hence, we need to empirically derive the performance for for entrepreneurs that actually chose
. In essence,
might nor provide the counterfactual as entrepreneurs self-select based
on pre-entry experience, and hence the control and treatment are likely to differ on core-characteristics
that matter for new venture emergence.
1
W
0
W
0
W
E
1
Y
1
]
0
Y
]
1
Y
0
Y
0
Y
1
W [
0 0
W | Y
Hence, in our empirical analysis we are interested in estimating the differences in foundation
probabilities conditional on the strategic decision to plan or not. Rubin (1974) refers to this effect as
the “treatment effect” and Nickerson and Hamilton (2003) use the term “strategy effect”. This effect
captures how the performance of a nascent entrepreneur who chose to plan would have changed, had
he not planned . The question that we are interested in the following is not
whether business planning had a positive impact for the entrepreneur. Rather, we address how he
[ ] [
1 0 1 1
W | W | Y E Y E ?
would have performed had he not planned. Hence, we are interested in estimating the differences in
performance conditional on the strategic decision to plan or not. Thus, we can estimate how
performance changes conditional on the entrepreneur’s choices.
3 Methodology
3.1. Dataset
To test our hypotheses empirically, we draw on the Second Panel Study of Entrepreneurial Dynamics
(PSED II) dataset. The PSED II is a representative survey of entrepreneurial activities in the United
States that portrays individuals during their business creation process. The dataset describes the
characteristics of nascent entrepreneurs, documents the sequences of the organizing activities,
summarizes the types and volumes of resources committed, and characterizes the new ventures.
1
For PSED II individuals were identified in late 2005 (early 2006), with 4 recurrent follow up
interviews, each taking place after 12 months subsequently. The last wave was therefore completed in
J anuary 2010. The sample of active nascent entrepreneurs was sampled from an overall group of
31,845 individuals. Out of this probability sample, 1,214 active nascent entrepreneurs were identified
in an initial interview. Interviewees were identified through their answers on screening questions as to
whether they were intending to start a new firm, already carried out at least one start-up activity in past
years, were expected to own part of the firm, and did not have an already going business. Hence,
entrepreneurs eligible for the sample were involved with an ongoing, but not yet operational start-up.
These eligible cases are re-interviewed every 12 months over the course of five years. The early stage
screening was used to ensure representativeness of the data and more importantly, to reduce
distortions caused by potential survivorship biases. Throughout the data collection process,
respondents give affirmative answers concerning a package of start-up activities that reveal the
progress they make in terms of becoming operational. Given the re-interviewing over the course of
five years the resulting longitudinal structure with monthly indications of activities started and
1
Detailed descriptions for the methods and sampling used to generate PSED II and an overview on the data structure can be found in
Reynolds and Curtin (2009).
finished allows for inferences on the process of organizing activities and facilitates causal inferences
among dependent and independent variables.
Concerning the number of respondents, wave A identifies 1,214 nascent entrepreneurs that returned
the questionnaire, whilst the number drops subsequently due to non-respondents and disbandment to
972 for Wave B, and 746, 526, and 435 for the Waves C to E. Among the start-ups, 512 disbanded
their venture and 243 perceived their venture as operational. For the remainder of the other ventures,
180 reported ongoing activities as per wave 5, but did not perceive their venture as operational. Lastly,
279 start-ups out of the initial sample were omitting at least one wave or stopped to report their
progress entirely. Hence, for these entrepreneurs we do not know the whereabouts of their organizing
activities, and consequently, these observations are pruned. Hence, we start our empirical analysis
with a total of 935 observations.
Selection correction and covariate imbalance adjustment
Given the endogenous choice based on pre-entry experience, the entrepreneur who did not plan cannot
necessarily provide the counterfactual as a comparative difference based on pre-entry experience could
make business planning beneficial for the one, but not for the other. Moreover, if pre-entry experience
affects the likelihood of business planning and the chances for venture emergence simultaneously, the
error terms of the decision to plan correlate with the outcome regression (new venture formation)
(Dencker et al., 2009; Unger et al., 2011). Thus, coefficients from regressions that fail to accommodate
for endogenous choices are likely biased which eventually causes facile inferences (Hamilton and
Nickerson, 2003).
Simply regressing a dummy variable with the business planning decision on the outcome of the
process is only valid if entrepreneurs make mistakes (and hence, strategies are random) or if all factors
driving the outcome are observable (Shaver, 1998). Both conditions are unlikely to be satisfied for
non-experimental data. Even if cognitive heuristics are present in start-up environments (Busenitz and
Barney, 1997) and decision-making errors are common and cause severe biases (Franke et al., 2006)
decisions are not purely exogenous.
Hence, inferring the counterfactual from the control groups that are present in the data might lead to
erroneous results due to self-selection. Accordingly, we need to control for the imbalance in pre-entry
experience across the treatment (business planning) and control (no business planning) group. To
ensure that we can infer the counterfactual from the control group we employ coarsened exact
matching (CEM) (Iacus, King, and Porro, 2011; Singh and Agrawal, 2011). Matching helps to control
for some (or all) pretreatment confounding covariates and makes the empirical estimation less model
dependent, prunes observations that have no close match within both the control and treatment group,
and removes heterogeneity. Hence, it ensures a better balance between the treatment and control group
and consequently the covariates in both groups are more similar and thus, allow inferences despite
potential pretreatment endogeneity (caused by the self-selection into planning and no planning,
respectively).
As opposed to other matching methods, CEM reduces the covariate imbalance between the treatment
group (business planning) and the control group (no business planning) without tedious balance re-
checking (and likely model re-estimation). CEM specifies balance ex-ante and reduces model
dependence and the error of the treatment effect estimation (Iacus, King, and Porro, 2011; Ho et al.,
2007). Data is temporarily coarsened into meaningful groups, matched exactly on the coarsened data
(bins) and uses the actual (uncoarsened) data for the matched observations in the empirical analysis of
causal effects. Accordingly, the ex-ante specification ensures common support (all matched
observations within bins are by definition in the area of common support) and this need not be tested
ex-post. Moreover, CEM reduces model dependence and the balance need not be achieved through an
empirical model, i.e. linear regression or maximum likelihood estimator (Abadie et al., 2004). We rely
on the implementation in Stata, to choose the bins based on Scott´s rule (1992) rather than providing
chosen cut-off points (Iacus, King, and Porro, 2011; Marx, Singh, and Fleming, 2010).
Dependent Variable
Completion of entrepreneurial organizing activities
Researchers have discussed a wide variety of measures to determine the point at which a nascent
venture shifts from the end of the entrepreneurial organizing activities to an operational business:
ability of raising external money, legal establishment of the new venture, first sales, positive cash
flow, reaching the break-even point, etc. (Gartner and Carter, 2003; Davidsson and Gordon, 2011). In
the following we will discuss occurrence of a first positive cash flow combined with a self-reported
measure of being operational (or disbanding effort), the probability of becoming operational in the
presence of the option to terminate efforts, and making comparisons across the most appropriate group
level of analyses.
Bygrave (1989) asserts that the only way to know whether the new venture will generate a persistent
business is to wait until the new venture is generating positive cash flows. Other researchers claim that
there are limitations of cash flow as a measure for completing the entrepreneurial organizing activities.
Bhide (2000) argues that cash flow is not likely to be an early goal of most high-potential new
ventures. Katz and Cabezuelo (2004) make a case, that nascent entrepreneurs are not always
sophisticated enough to calculate positive cash flows exactly. Moreover, entrepreneurs might have
different objectives (Delmar and Shane, 2006; Gruber, MacMillan, and Thompson, 2008). As the
industry composition reveals, we mostly deal with “lifestyle”-ventures that are “low-tech”. Hence,
sustainability is important as it signals the ability to generate an income stream and a general level of
market acceptance for the entrepreneurial venture.
Given the debate on the applicability of researcher defined and self-reported outcome measures, we
employ a more prudent outcome measure by using information whether the entrepreneurs reached a
first positive cash flow (covering managerial salaries) and also perceived their venture as operational,
and compare these with entrepreneurs that indicated that they terminated their work on the venture.
Becoming operational is therefore a combination of question A35 and A41 asking whether monthly
revenues ever exceeded monthly expenses (including salaries for the managers) and whether based on
this achievement, the respondents would characterize their venture as being operational. If both
conditions are fulfilled the variable takes on the value of one. We label this outcome as “perceived
emergence”. Moreover, following Delmar and Shane (2003) we also investigate whether all involved
entrepreneurs report disengagement to allow comparison across these two reported statuses. If others
are still working on the venture, we did not treat these as disbanded as solely the key respondent
disengaged from the venture, which is distinct from disbandment. If all reported start-up members
disengaged from the venture the variable takes on the value of zero for the comparison category and if
others are still working we omit the data due the “still trying” nature of the observation. Hence, we fit
a logistic regression with CEM weights using the outcome perceived emergence against venture
disbandment as the dependent variable.
Event history model
In the following, we make also use of the longitudinal character of the dataset to estimate the impact of
business planning on the hazard to perceive the venture as emerged. We use the information provided
by nascent entrepreneurs on the months in which they carried out their organizing activities and the
month in which they either disbanded their organizing efforts or when their monthly revenues
exceeded expenses (including managerial salaries) for the first time and they perceived their venture as
emerged.
Venture disbandment takes place, when all parties identified to be working on the venture stop their
entrepreneurial activities. Hence, there is no person involved that follows through with the original
plan to bring this particular venture to life. Individual respondents were asked whether the disengaged
from the venture. Moreover, they indicated that in some cases other members of the team were still
continuing to work on the venture. Only if all members were disengaged from the venture, we treated
the venture as disbanded.
However, disbanding efforts and continuation might in part be related to the same underlying factors.
Business planning serves a two-fold purpose in this context. On the one hand, it helps to shape and
develop the actual business opportunity the entrepreneur wants to pursue, and on the other hand it
assists to evaluate the opportunity and to weed out good from bad alternatives. Information generated
from business planning effort can thus go both ways, providing positive feedback and negative
feedback alike. As such, disengagement of the venture might not always be a sign of bad performance,
but could also be an indication that some entrepreneurs question the feasibility of their aspired
business model and pursue alternative options in lieu of the initial venture. Consequently, the longer a
nascent entrepreneur is in the process of organizing a new venture, the higher will be the chances to
become operational (due to the effort put into organizing activities) and similarly the risk of
disbandment (if the effort does not show any material impact on venture emergence). In contrast to
Delmar and Shane (2003) we therefore treat the disbandment of ventures as a competing risk, rather
than a pure censoring event (Fine and Gray, 1999). In fact, disbanding the venture organizing efforts is
an endogenous decision that prevents further business planning and hence, prunes the number of
observations.
The disbandment event is distinct from the classical analysis of survival using exit as a censoring
event. Censored events generally are treated as being “at risk”, whilst the researcher cannot observe
what happens to the individual under investigation. However, for the case of intentional venture
disbandment we know with certainty, that the venture will not become operational as the entrepreneur
gave up working on the venture. Hence, venture disbandment is permanent and prevents the venture
from becoming operational. Censoring obstructs the researcher from observing the event, whilst the
competing event disbandment prevents the event “perceived emergence” from occurring.
Consequently, we treat the month in which the respondent disengaged as a competing risk to the
perception of emergence measure in the event history analysis. The use of competing risks controls for
inherent differences among entrepreneur characteristics that link to both, the chances to become
operational and the risk of disbandment, such as pre-entry experience and the quality of the underlying
opportunity pursued.
We take into consideration the month in which the ventures undertook their first activity and the
months, in which they either perceived their venture as emerged or disbanded their venture organizing
efforts, the hazard therefore has a monthly spell. A positive coefficient indicates that an increase in the
covariates for the business planning variables increases the hazard of becoming operational. Given the
combination of emerged, terminated and in-process ventures, this analysis also allows to distinguish
among different gradations of venture emergence that overcome limitations of right censoring (Dimov,
2010). Moreover, we employ coarsened exact matching first to reduce covariate imbalance across the
treatment and control group.
Sample Comparison and Complementary Analyses
Censoring: To measure successful completion of the founding process, we also accommodate
for potential censoring problems. Some entrepreneurs might have started a long time prior to the
interview and, consequently, have higher chances to reach certain milestones before others.
Specifically, firms that started prior to 2005 had more time to prepare the venture. Hence, the 5 years
covered in the data set might not be sufficient to compare the different groups and truncation periods.
Gartner and Carter (2003) and Lichtenstein et al. (2007) therefore suggest including only
entrepreneurs who underwent their first activity within 24 month prior to interview time. The approach
is similar to Delmar and Shane (2003) that included only start-ups that underwent activities within the
year of the interview to accommodate potential censoring and selection biases. Accordingly, we
reduce the sample further by excluding 128 entrepreneurs that did not pursue their first activity within
the time span covering the 24 month prior to the first interview to wave E.
Moreover, Davidsson and Gordon (2011) report the existence of “dilettante dreamers” or hobbyists in
nascent entrepreneurial sample. Accordingly, to rule out entrepreneurs that are not serious about their
start-up intentions and to make the nascent phase comparable across start-ups Davidsson and Gordon
(2011) suggest to compare start-ups without the “still trying”-category to avoid distortions by less
serious entrepreneurs (Parker and Belghitar, 2006). We report our first set of results using a
comparison of those entrepreneurs that perceived their venture as emerged and those that indicated
that they disbanded their effort. Hence, this analysis excludes further 154 entrepreneurs that report a
“still trying” status as per wave E. Moreover, we report the survival analysis making use of the time-
stamp information concerning all activities that nascent entrepreneurs undertook to measure the
probability of becoming operational in the presence of the option to disband. Here we explicitly make
use of the three different categories.
Predictor Variables
We use several measures to construct the corresponding treatment and control groups. First, we
include all nascent entrepreneurs who indicate that they started to work on a business plan; secondly,
we use the group of nascent entrepreneurs that indicated that they successfully finished a version of the
business plan. Third, PSED II reports whether the finished version of the business plan is formal or
informal. Hence, we also define a treatment group of nascent entrepreneurs that finished a formal
version of the business plan. Fourth, to capture the formal nature of the business plan in more detail,
we also construct a treatment group for the entrepreneurs that made financial projections. Fifth, during
every phase, entrepreneurs indicated whether they made changes to the business plan. To capture the
dynamic nature of business planning, we define yet another treatment group comprising of nascent
entrepreneurs that made changes to their business plan. Hence, we have five different proxies to
capture endogenous business planning strategies as a treatment effect. We follow the extant literature
(Delmar and Shane, 2003; Gruber, 2007; Dimov, 2010) and administer the business planning variables
in dichotomous form.
Conditioning Variables
We match entrepreneurs who decided to plan and those who decided not plan, based on their pre-entry
experiences. We operationalize pre-entry experience of nascent entrepreneurs by reference to
the most widely cited factors such as formal education (e.g., Evans and Leighton, 1989;
Dickson, Solomon, and Weaver, 2008; Unger et al., 2011), labor market experience (e.g.,
Bosma et al, 2004), and entrepreneurial experience (e.g., Robinson and Sexton, 1994; Bosma
et al, 2004). A principal factor analysis (with varimax rotation) yields a clear factor structure
(with eigenvalues for each factor greater than one) for these three measures. The variables are
calculated for solo-entrepreneurs and team-foundations. In the latter cases, we include the
average levels of these factors.
Formal Education: To collect PSED II data, respondents were asked to indicate the highest
level of education all members of the entrepreneurial team had completed. We recoded this
variable, ranging from elementary school to PhD, into number of years of education (see e.g.
Davidsson and Honig, 2003; Iacus, King, and Porro, 2011).
Labor Market Experience: PSED II provides information about the years of work
experience in the industry a new venture is active in, years of full-time paid work experience,
and years of managerial, supervisory, or administrative responsibilities of the nascent
entrepreneurs. Cronbachs alpha is 0.72 for the three factors.
Entrepreneurial Experience: We use information on the number of other businesses
previously helped to start as an owner, and the number of other businesses they own(ed). The
correlation among these two variables is 0.6 (which would result in Cronbach’s alpha of
0.65).
Moderators:
Seek Financing: As Business Plans might not only serve the purpose of planning future activities
but might also be addressing needs of third parties, such as potential financiers, the analysis accounts
for the impact of entrepreneurs writing business plans to seek outside funding. We account for whether
or not the entrepreneurs did actively seek financial capital. The variable is a dichotomous measure that
equals “1” if entrepreneurs seek outside financing, and zero otherwise.
Market Newness: We measure the perception of market newness by using the answer to the
question whether the product is unfamiliar to all, some or none of the potential customers. The
variable uses a three-point scale. The scale is: “3” equals “all customers will be unfamiliar with this
new product or service”, “2” equals “some customers will be unfamiliar with this new product or
service”, and “1” equals “none of the customers will be unfamiliar with this new product or service”.
The variables are also in part included in the index developed in Dahlqvist and Wiklund (2011) and
represent newness of a product to customers. To ensure the appropriateness of the index to measure
whether a product is new to customers, we regress the measure on the probability to engage in the
development of a proprietary technology or process or whether entrepreneurs apply for a patent or get
a patent granted. All coefficients for the scale are positive and highly significant at the 0.1% level.
Hence, we are confident that this measures in fact proxies for market newness of a product.
Control Variables
Industry: Since the role of business planning can differ across industries we parce out these
effects by including industry dummy variables. We control for retail (86 firms), consumer services
(214), health (50), consulting (43) manufacturing and construction (73), real estate and finance (61)
and other industries (100). We omit “other industries” as the reference group in each regression to
avoid perfect collinearity. Inclusion of industry dummies is indicated in the corresponding table.
Age: We control for the age of the entrepreneurs using the average over all team
members as indicated in wave A.
Team Size: We control for team size using all team members as indicated by the respondent
to the questionnaire in wave A.
Competition: We measure the perception of competition using a three-point scale. The scale is:
“3” equals “there are many other businesses offering the same product or service”, “2” equals “there
are few other businesses offering the same product or service”, “1” equals “there are no other
businesses offering the same product or service”.
Motivation: As the motivation to start a business might vary among entrepreneurs, we follow
Dimov (2010) and include questions from PSED II on entrepreneurial motivations into our empirical
analysis. The measure of start-up motivation comprises the answer to the questions “There is no limit
as to how long I would give maximum effort to establish my business” and “My personal philosophy
is to do whatever it takes to establish my own business”. Combining both variables would result in
Cronbach´s alpha of 0.71 (comparable to Dimov (2010).
Self-Efficacy: We measure perception of self-efficacy using the five questions identified in
Dimov (2010). The questions employ a five-point Likert-type scale. A confirmatory factor
analysis reveals that three of the five items load on one factor and result in satisfactory
Cronbach´s alpha of 0.68 (corresponds to questions AY4 to AY8 from PSED 2). One could
improve the fit by using only the questions AY6 to AY8 to 0.72. Nevertheless, we followed
the recent approach in Dimov (2010) to make our results comparable to previously published
studies and employ a measure comprising all five items. The original PSED 2 scale is inverted
so that higher values indicate higher levels of self-efficacy.
4. Results
Table 1 summarizes the details for the underlying sample of this study including 627 nascent
entrepreneurial ventures. Among these, 23% considered their venture to be operational while some
56% disbanded their organizing efforts. These performances within PSED II are similar to other
studies analyzing new firm creation in the US (Spletzer et al., 2004; Reynolds, 2009). Concerning the
prevalence of business planning around three quarter of the ventures engaged in business planning,
while half of them indicated that they finished their planning activities subsequently. Additionally,
only 20% finished a first formal version of the business plan whilst around 50% produced
accompanying financial projections. Lastly, around every third respondent indicated that changes were
made to the first version of the business plan.
[Insert table 1 about here]
Table 2 reports logistic regressions with the five different dimensions of business planning as
explanatory variables. The outcomes variable is equal to 1 when entrepreneurs perceive their venture
as emerged and zero when all people involved disbanded their effort. Four out of the five dimensions
of business planning positively affect the perceived emergence of a venture. Starting to plan has no
significant impact on the emergence of a nascent venture (ß =0.072). Also, those entrepreneurs that
finish the process of business planning (ß =0.121, p