Start-ups that adopt big-data analytics face higher early failure rates because of greater costs and sales uncertainty, yet surviving adopters grow faster, earn more and are more likely to attract VC — suggesting big-data acts as a performance amplifier for firms that can manage the risk.
Big data-based management decisions and start-up performance
Using a large sample of German start-ups, the paper finds that BDA adoption is associated with lower survival (driven by higher costs and sales volatility) but, conditional on surviving, adopters enjoy higher sales, faster employee growth, and greater likelihood of attracting venture capital.
Abstract With big data (BD) becoming widely available, the question arises whether big data analytics (BDA) enhances managerial decision-making and firm performance not only for incumbents but also for start-ups. Given their liability of newness and limited resources, adopting this new technology poses significant risks for young ventures. Drawing on a large sample of start-ups in Germany, we examine BDA adoption and its economic impact across multiple performance measures, including survival, costs, sales, employee growth, and access to financing. Our findings show that BDA adoption is a risky strategy with potentially high rewards. Start-ups using BDA face lower survival rates, driven by two interrelated factors: higher operating costs and greater uncertainty in sales. At the same time, conditional on survival, adopters of BDA benefit from higher sales, stronger employee growth, and a greater likelihood of attracting venture capital. For high-performing BDA adopters, sales and employee growth are even more pronounced. Overall, BDA functions as a performance amplifier, yielding higher returns for ventures well-positioned to leverage its potential.
Summary
Main Finding
Using a large longitudinal sample of German start-ups (2010–2018), the authors find that adopting big data analytics (BDA) is a high-risk, high-reward strategy for early-stage ventures. BDA adopters have lower survival probabilities (driven by higher operating costs and more volatile sales), but conditional on surviving they show higher sales, faster employee growth, and a greater likelihood of receiving venture capital. Benefits are concentrated among top-performing adopters — BDA acts as a performance amplifier.
Citation: Rodepeter, E., Gschnaidtner, C., & Hottenrott, H. (2026). Big data-based management decisions and start-up performance. Small Business Economics. https://doi.org/10.1007/s11187-025-01164-6
Key Points
- Survival: Start-ups using BDA face lower survival rates than otherwise similar non-adopters (Proposition 1).
- Costs: BDA adoption is associated with higher operating costs, especially higher personnel costs, increasing financial strain on young ventures (Proposition 2).
- Sales & Volatility: Adopters exhibit greater sales conditional on survival, but also higher sales volatility—consistent with a riskier business profile.
- Employment growth: Conditional on survival, BDA adopters grow employment faster; this growth is particularly concentrated in the upper performance deciles (i.e., top performers benefit most).
- Financing: BDA adopters are more likely to attract venture capital, consistent with investor preference for high-risk/high-return opportunities.
- Overall pattern: The empirical evidence points to an asymmetric “amplifier” effect — BDA magnifies upside for well-positioned start-ups but increases the chance of failure for many others.
Data & Methods
- Data: A unique, extensive longitudinal dataset of German start-ups observed between 2010 and 2018. Start-ups are defined broadly as independent businesses aged up to 7 years. The dataset includes firm and founder characteristics, accounting variables (sales, costs), employment, and financing (VC links).
- Outcomes analyzed: firm survival, operating costs (and personnel costs), sales (and sales volatility), employee growth, and probability of receiving VC.
- Empirical approach (as reported): comparative longitudinal analysis of BDA adopters versus non-adopters controlling for observable firm and founder characteristics. The authors exploit panel variation to analyze:
- survival differences,
- cost and sales impacts,
- conditional (upon survival) performance,
- heterogeneity across performance deciles to detect amplifier effects,
- links between BDA adoption and VC financing.
- Caution on identification: the paper emphasizes conditional outcomes and heterogeneity to reveal the high-risk/high-reward pattern. The analysis controls for many observable covariates, but the paper documents selection and compositional effects (i.e., benefits are concentrated among survivors and top performers).
Implications for AI Economics
- Adoption trade-offs in young firms: BDA requires complementary assets (IT infrastructure, analytic talent) that are costly and can increase early failure risk. Economists should model BDA adoption as a technology choice with strong fixed and complementarity-driven costs that create selection into adoption and survivorship bias in observed returns.
- Winner-take-most dynamics: The amplifier effect implies that BDA may exacerbate heterogeneity across start-ups — a subset of well-resourced or better-managed entrants capture outsized returns while many adopters fail. This can increase concentration among successful firms and shape market structure in digital-intensive sectors.
- Finance and risk preferences: Venture capitalists prefer BDA adopters despite higher failure risk, highlighting the role of financing in enabling high-risk/high-return technological strategies. Models of entrepreneurial finance should incorporate how investors fund adoption of complementarity-heavy AI/BDA technologies.
- Labor-market effects: Increased personnel costs for data talent imply rising demand and wage pressure for analytic skills in the start-up sector, potentially reallocating scarce human capital and raising entry costs.
- Policy relevance: To unlock broader societal gains from BDA in entrepreneurship, targeted policies (training, infrastructure, subsidized access to analytics talent or platforms, early-stage financial support) could mitigate adoption costs and failure externalities, improving the social return on diffusion of data-driven management.
- Research directions: Future work should identify causal mechanisms (e.g., instrumenting for adoption or exploiting exogenous shocks), distinguish BDA functions (prediction automation vs. decision-support), examine sectoral heterogeneity, and quantify long-run reallocation effects from failures versus superstar growth.
Assessment
Paper Typecorrelational
Evidence Strengthmedium — The paper uses a large, observational sample of start-ups and documents robust associations across multiple outcomes (survival, costs, sales, employment, VC funding), including conditional analyses (e.g., conditional on survival). However, there is no clear quasi-experimental identification (no random assignment, instrument, or natural experiment reported in the abstract), so selection, reverse causality, and survivorship bias can plausibly drive some or all of the observed relationships.
Methods Rigormedium — Strengths include a large sample of start-ups and multiple economic outcomes, which allows for complementary tests (e.g., heterogeneity by performance). But without details on strategies to address endogeneity (matching, panel fixed effects, IVs, regression discontinuity) and measurement of BDA adoption, the study likely relies on covariate adjustment and conditional comparisons, limiting causal interpretation.
SampleA large sample of German start-ups (firm-level data) analyzed across outcomes including survival, operating costs, sales, employee growth, and venture-capital attraction; the abstract does not specify years, data sources (survey vs administrative), industry coverage, or sample size.
Themesadoption productivity innovation
GeneralizabilityGeographic: limited to start-ups in Germany; institutional and financing environments differ elsewhere, Population: start-ups only; findings may not generalize to established incumbents or SMEs, Technology definition: 'big data analytics' is a broad category—heterogeneity in tools and implementation may limit applicability, Selection: adopters may differ systematically (capabilities, risk tolerance, sector) from non-adopters, limiting causal extrapolation, Time period unspecified: results may depend on the specific technology maturity and market conditions when data were collected
Claims (10)
| Claim | Direction | Confidence | Outcome | Details |
|---|---|---|---|---|
| Big data analytics (BDA) adoption is a risky strategy with potentially high rewards for start-ups. Firm Productivity | mixed | high | overall performance/risk–reward tradeoff |
0.3
|
| Start-ups using BDA face lower survival rates. Firm Productivity | negative | high | survival (firm exit / failure) |
0.3
|
| Lower survival rates among BDA adopters are driven by higher operating costs. Organizational Efficiency | negative | high | operating costs |
0.3
|
| Lower survival rates among BDA adopters are driven by greater uncertainty in sales. Firm Revenue | negative | high | uncertainty in sales (sales volatility/variance) |
0.3
|
| Conditional on survival, BDA adopters have higher sales. Firm Revenue | positive | high | sales (revenue levels) |
0.3
|
| Conditional on survival, BDA adopters show stronger employee growth. Hiring | positive | high | employee growth (headcount growth) |
0.3
|
| Conditional on survival, BDA adopters are more likely to attract venture capital financing. Adoption Rate | positive | high | access to financing (venture capital uptake) |
0.3
|
| For high-performing BDA adopters, increases in sales are even more pronounced. Firm Revenue | positive | high | sales (revenue levels) for high-performing adopters |
0.3
|
| For high-performing BDA adopters, employee growth is even more pronounced. Hiring | positive | high | employee growth (headcount growth) for high-performing adopters |
0.3
|
| Overall, BDA functions as a performance amplifier, yielding higher returns for ventures well-positioned to leverage its potential. Firm Productivity | positive | high | performance/returns to adopters |
0.3
|