inclusion of such information, grea

inclusion of such information, greater values are
positively related to venture funding. Following
the logic summarized in Table 1, this information
is learned elsewhere and not from the planning
documents.
DISCUSSION
Our results suggest that some information in business
planning solicitations and artifacts predicts
funding. However, the results also suggest that this
information is not learned from business planning
documents.
Our empirical strategy uses results based on
conditioned samples to distinguish between the
competing views (see Table 1). Our inferential
strategy relies heavily on the assumption that the
entrepreneur only includes information when it is
above a certain quality threshold. This strategy will
fail if VCs only care if a particular attribute of
the venture is ‘good enough’ and if entrepreneurs
accurately estimate how much is ‘good enough.’
In this case, conditional on the inclusion of information,
the values of the attribute will not predict
venture funding outcomes. However, our empirical
results are not consistent with this premise.
In fact, at times, entrepreneurs include information
that is not particularly helpful to their cause
as evidenced by the significant coefficients in the
conditional regressions. We suspect that in these
cases, entrepreneurs err in judging what it is VCs
wish to hear.
Nevertheless, to further explore this concern we
exploited one source of information that we systematically
observe for firms that did not submit
business planning documents. We thereby avoided
conditioning the sample on observability. In particular,
we estimated a model that predicted successful
attainment of second round VC financing for 93
firms that had received at least one round of venture
investment prior to their request of the FVC.
In this unreported regression, we distinguished
between VC funding amounts that were reported
in the planning documents and those that were
reported exclusively in public databases (Venture-
Xpert). Not surprisingly, both measures of previous
funding are positive, significant and large.
However, the coefficients of these variables are statistically
(and numerically almost) identical. This
suggests that information about previous funding
is collected and processed by venture capitalists