1. BUSINESS IDEA SOURCES AND
ENTREPRENEURSHIP:
AN INSTITUTIONAL VIEW
Chuck Eesley (Stanford)
Xiao Hu (Tsinghua University)
Delin Yang (Tsinghua University)
2. Motivation
• Prior work shows that venture performance is shaped
by characteristics of:
– The founding team (Beckman, Burton, etc.)
– Growth Market (Eisenhardt and Schoonhoven, 1990)
– Technology (Shane)
– User-driven innovation (von Hippel, Sonali Shah, etc.)
• Gap: Prior work has rarely examined whether the
source of the startup idea exerts an influence on
performance
– Entrepreneurial Ideation and Organizational Performance:
Imprinting Effects (Eesley, Hsu, and Roberts, WP)
3. Policy-driven business idea source
• Government context from policies - contacting
with government (communicating with related
government officials) and following the policy
change closely.
– For example, entrepreneurs might start new
ventures based on property policy reform (Walder,
1999)
4. • Market-driven business idea source indicates
that entrepreneurs choose to obtain their
business ideas from market opportunities by
contacting other market participants
(suppliers, customers, competitors).
– When an entrepreneur chooses to obtain his
business idea from market-driven source, he has
many ways to identify better (new) potential
markets or better (new) ways to serve target
markets (Johne, 1999)
5. Hypothesis 1.
• In the transitional institutional environment
with high level of dynamics, entrepreneurs
who choose to obtain their business idea from
policy-driven source will create better
performing firms.
6. Hypothesis 2.
• In the stable institutional environment with
low dynamics and uncertainty, entrepreneurs
who choose to obtain their business idea from
market-driven source will create better
performing firms.
7. Hypothesis 3:
• Entrepreneurs who choose to obtain their
business idea from both policy-driven and
market-driven source will create better
performing firms.
8. Research Context: China
• transitional institutional from 1988 to 1999,
• stable institutional environment from 2000 to
2002,
• transitional institutional environment from
2003 to 2009,
9. 9
Tsinghua Survey
Unique data on firm origins, early
composition and multiple stages
of performance measures
• Alumni: ~26,700 mailed; 3,000
surveys, 11% response rate
–General biographic
information, education and career
histories, family background,
attitudes toward
entrepreneurship.
–718 individuals (24%) indicated
founding at least one company
• Additional questions for self-identified
entrepreneurs:
–Company origins, structure, and
performance.
• Advantages: Defined‘at risk’ set, first
abroad, detailed work history and founding
data, less biased by govt. concerns
Charles Zhang, founder
and CEO of Sohu.com
10.
11. Industries
11
INDUSTRY NUMBER OF FIRMS %
AEROSPACE 3 0.90
ARCHITECTURE 13 3.88
BIOTECH AND DRUGS 7 1.09
CHEMICALS 8 2.39
CONSUMER
PRODUCTS 17 5.07
ELECTRIC 12 3.58
ELECTRONICS 69 20.60
ENERGY 14 4.18
FINANCE 10 2.99
INTERNET 33 9.85
LAW, ACCOUNTING 22 6.57
MACHINERY 19 5.67
MANAGEMENT 21 6.27
MATERIALS 13 3.88
MED DEVICES 4 1.19
OTHER MFG 16 4.78
PUBLISHING 11 3.28
SOFTWARE 34 10.15
TELECOM 9 2.69
TOTAL 335 100
12. Sample
• Population:
– 62.5% in engineering
11.9% in sciences
– 12.9% in humanities
(architecture, medicine
and law comprise the
remainder)
– 25-30% women19.2%
doctorate degrees
– 53.4% graduate degrees
• Tsinghua survey sample:
– 62.2% engineering
– 10.6% sciences13.7%
humanities
– 28% women
– 19.3% doctorate degrees
– 53.9% graduate degrees
13. DV: ln(Revenue). We measured entrepreneurial performance as
the natural logarithm of revenue in the most recent fiscal year for
the entrepreneur’s firm.
IVs:
• Policy-driven Source measures, on a scale of 1 to 8, the degree
to which entrepreneurs believed that the government’s policy
was an important idea source for their ventures.
• Market-driven Source, as the measure of the extent that an
entrepreneur’s idea source depended on new market
opportunities at that time.
• Market-driven Source measures, on a scale of 1 to 8, the
degree to which entrepreneurs believed that market
opportunities were an important idea source for their ventures.13
Context - interviews
14. How are Tsinghua’s Entrepreneurs
Born?
The entrepreneurs surveyed reported that their initial business ideas came from:
Class at
Tsinghua
Research at
Tsinghua
(internal
funding)
Research at
Tsinghua
(outside
funding)
Thesis at
Tsinghua
Discussion with
Tsinghua
classmates or
teachers
A visiting
lecturer at
Tsinghua
14.6% 4.1% 4.8% 9.9% 27.4% 12.1%
An internship
while at
Tsinghua
Reading
professional
literature at
Tsinghua
Industry
experience
Army or
government
experience
An academic
conference
Discussion with
fellow
professionals
12.1% 11.1% 61.8% 7.3% 11.1% 43.9%
14
15. Where did the primary idea for the start-up come from?
Idea sources MIT Data Tsinghua data
percentage percentage
In school-doing outside-funded research 2.40 1.66
In school- graduate thesis 4.64 3.96
In school- in class 1.98 5.88
In school-informal discussion w/ students 3.41 11.00
In school-other research 2.28 1.92
In school-professional literature 1.73 4.48
In school- visiting scientists, engineers 1.77 4.86
In school-working w/ outside company 3.20 4.86
Other sources-discussions in
social/professional conferences
21.54 17.65
Other sources-research conference 2.66 4.48
Other sources-working in the industry 41.44 24.81
Other sources- working in the military
(government experience)
4.01 2.94
Doing outside-funded research 2.07 0.77
Total 100 100
Number of observations 1284 110
16. Variable ln(Revenue)
(1-4)
88-99
(2-3)
00-02
(3-4)
03-07
Policy-driven Source 0.752** 0.118 0.374***
(3.91) (0.90) (3.79)
Market-driven Source 0.0233 0.379** 0.0261
(0.14) (2.98) (0.29)
Interaction between
Sources
-0.00328 -0.0432 -0.0341
(-0.05) (-0.69) (-0.87)
Controls
Age 0.00179 -0.0252 0.00183
(0.02) (-0.37) (0.05)
Gender -0.372 0 -0.737
(-0.28) (.) (-0.75)
Overseas 1.087 0.799 -0.0856
(0.90) (1.35) (-0.19)
Length of Work -0.0198 0.0329 0.0146
(-0.28) (0.55) (0.37)
Position Diversity 0.600 -0.115 -0.149
(1.42) (-0.59) (-1.01)
Market Experience 0.698 -0.340 0.402*
(1.84) (-1.59) (2.26)
Entrepreneur Ties -0.570* 0.109 0.0473
(-2.41) (0.70) (0.37)
Communist Party -0.928 -0.0672 0.153
(-1.02) (-0.14) (0.38)
Parent Official 0.378 -0.223 0.0163
(0.93) (-1.02) (0.08)
Work for Government -0.0460 -0.552** -0.362*
(-0.07) (-2.99) (-2.06)
Government Tie 0.361 0.483* 0.0227
17. Variable In(Revenue) In(Revenue)
(4-3) (4-4)
Policy-driven Source 0.365*** 0.367***
(5.06) (5.11)
Market-driven Source 0.0814 0.0864
(1.22) (1.30)
Interaction between Sources -0.0390
(-1.43)
Control
Age -0.00680 -0.00450
(-0.23) (-0.15)
Gender -0.821 -0.883
(-1.12) (-1.20)
Overseas 0.509 0.536
(1.55) (1.63)
Length of Work -0.00251 -0.00466
(-0.09) (-0.16)
Position Diversity -0.0558 -0.0422
(-0.51) (-0.39)
Market Experience 0.232 0.221
(1.84) (1.75)
Entrepreneur Ties -0.0936 -0.100
(-1.04) (-1.11)
Communist Party -0.0788 -0.0815
(-0.27) (-0.28)
Parent Official 0.148 0.132
(1.08) (0.97)
Work for Government -0.439*** -0.438***
(-3.51) (-3.50)
Government Ties 0.128 0.120
(1.19) (1.13)
18. Contributions
• Ideas for customers/users may not always be
the best
• Source of the idea appears to play an
important role in venture quality – deserving
of greater attention in future research
• Varies with the institutional environment
Editor's Notes
Wide range of work backgrounds, but not formally randomized.
The Tsinghua Alumni Association set up interviews, specifically asked to talk with some who were not successful.
Probably weighted towards more successful entrepreneurs
For these numbers I took the responses divided by 319 (the number of founder + privatized responders).