This section defines the variables used in the empirical analysis and introduces the hypotheses that will be tested using the database compiled from Crunchbase, in the next chapter. The goal is to operationalise and figure out the central research question—Are corporate accelerators springboards for startups or merely sand traps?—by linking measurable performance indicators to accelerator participation. These variables form the analytical foundation of the regression and ANOVA models.
This post represents a series of articles related to a research and dissertation called “Are corporate accelerators springboards for startups: a performance analysis of the Microsoft’s and Google’s accelerated.
Variables
INDEPENDENT VARIABLE: ACCELERATOR PARTICIPATION
The primary independent variable in this study is the startup’s accelerator affiliation. Each startup in the database is associated with only one corporate accelerator, either Google or Microsoft and allow for both individual performance assessments (e.g., outcomes of all Google-accelerated startups) and comparative analysis between the two programs.
DEPENDENT VARIABLES: STARTUP OUTCOMES
Startup performance in this study is evaluated through several measurable outcomes. One key variable is if it survived 3 years+, a binary indicator where 1 represents that the startup remained active for at least three years after acceleration, and 0 indicates closure. This variable reflects the short-term resilience of a startup post-program.
Another variable, IPO Timing, captures whether the startup reached an initial public offering within 1, 2, or 3 years after acceleration—or if it never did. This discrete outcome provides a useful signal for investor confidence and perceived scalability.
The Total Amount Funded is treated as a continuous variable that records the full sum of capital raised by the startup across all investment rounds. It offers insight into the financial traction achieved during and after acceleration. The quantity will be in USD.
The Number of Funding Rounds is a count variable that tracks how many distinct rounds of financing a startup completed. It indicates sustained investor interest.
It includes a variable labeled Closed, a binary measure showing whether the startup ceased operations (1) or remained active (0).
Together, these variables, among many others, provide multiple angles from which to assess the outcomes of acceleration, offering a multidimensional view of what “success” or “failure” may look like in a post-accelerator context.
CONTROL VARIABLES (DESCRIPTIVE)
Although no formal control variables are used in the comparative models, the database includes:
- Country and City of headquarters (used to identify geographic concentration)
- Industry SIC Code (useful for future segmentation or industry-specific filtering)
- Founded Date (used to infer company age at the time of acceleration)
HYPOTHESES
Based on the variables above, the following five hypotheses will be tested:
HYPOTHESE 1
Startups accelerated by Google are more likely to survive at least three years post-acceleration than those accelerated by Microsoft.
HYPOTHESE 2
Google-accelerated startups are more likely to go public (IPO) within three years than Microsoft-accelerated startups.
HYPOTHESE 3
The total amount of funding raised is higher among Google-accelerated startups than among those accelerated by Microsoft.
HYPOTHESE 4
There is a statistically significant difference in the average number of funding rounds between startups accelerated by Google and those accelerated by Microsoft (tested via ANOVA).
HYPOTHESE 5
Microsoft-accelerated startups are more likely to be closed (defunct) three years after acceleration compared to those accelerated by Google.
These hypotheses will be examined through the quantitative techniques described in Section 2.7 and applied in Chapter 4.
CONCLUSION
This section has defined the key variables and articulated five testable hypotheses based on accelerator participation. By translating the research question into statistical language, the study now has a framework for analyzing startup performance across both corporate programs. In the next chapter, each hypothesis will be tested using the Crunchbase database and the analytical methods already introduced.
References
Crunchbase. (n.d.). Crunchbase: Discover innovative companies and the people behind them. Retrieved from https://www.crunchbase.com/
Seitz, N., Krieger, B., Mauer, R., & Brettel, M. (2023). Corporate accelerators: Design and startup performance. Small Business Economics. https://doi.org/10.1007/s11187-023-00732-y
