The persistent gap

Female-founded and mixed-gender founding teams remain heavily underfunded compared to all-male teams across every VC funding stage in Europe. The gap is widening, not narrowing.

What the numbers show

Over the past twenty years, approximately 6% of European tech companies have been founded by all-women teams. This percentage has remained virtually unchanged since 2016.

These teams receive a disproportionately small share of VC capital at every investment stage. Even mixed-gender teams, which historically captured 32% of Series B funding in 2016, saw their share drop to 22% by 2025.

Out of the ten largest VC rounds in Europe in 2025, only one went to a mixed-gender founding team. The largest round for an all-female team did not make the top fifty rounds of the year.

Capital allocation by stage (2025)

Source: Atomico State of European Tech 2025

All-male teams continue to capture the overwhelming majority of VC funding. Mixed-gender and all-women teams remain systematically underrepresented at every stage.

The structural nature of the problem

This is not a pipeline issue. The data shows consistent patterns across stages, and academic research reveals why these disparities persist despite equal or better performance.

Performance parity: Research demonstrates no significant performance differences based on the gender of entrepreneurs. Female-led ventures perform comparably to male-led ventures on objective metrics, yet female founders receive significantly less venture capital funding even after controlling for industry, geography, and business characteristics.

Evaluation bias: Early-stage investors consistently show preferences for pitches presented by male entrepreneurs compared with those presented by female entrepreneurs, even when the presented information is identical. This bias manifests in concrete ways during the evaluation process.

Question framing: When venture capitalists interact with entrepreneurs, they ask fundamentally different types of questions based on gender. Male entrepreneurs receive primarily promotion-focused questions about their ventures’ potential gains and opportunities. Female entrepreneurs receive mainly prevention-focused questions about potential losses and risks. This asymmetry in questioning reflects and reinforces perceived risk differences that have no basis in actual venture performance.

Language of evaluation: The discourse used to evaluate entrepreneurs varies systematically by gender. When male entrepreneurs lack experience, this is often framed as “young and promising.” Similar gaps in female entrepreneurs’ experience are framed as “young and inexperienced.” When female founders signal entrepreneurial traits such as assertiveness and risk-taking, they face disadvantage resulting in revised funding decisions due to higher perceived risk, creating a double bind where conforming to expected gender roles signals weakness while violating them signals risk.

Systemic coping mechanisms: The persistence and recognition of these biases has led to concerning practices, including ‘venture bearding,’ where women employ men as front persons in an attempt to lower perceived risk for male investors. The fact that such strategies exist and are discussed openly indicates both the severity and intractability of the problem.

Why the patterns persist

Multiple theoretical and empirical mechanisms explain the persistence of these biases:

Homophily effects: Venture capitalists tend to invest in entrepreneurs who share their demographic characteristics. Given that 93 percent of VCs are male, this creates systematic disadvantages in a male-dominated industry.

Role congruity: Leadership and scaling potential are associated with agentic traits such as assertiveness and risk-taking, which are culturally coded as masculine. This creates perception that female-led ventures carry higher risk, regardless of objective venture characteristics.

Pattern reinforcement: VCs develop mental models of successful entrepreneurs based on past experiences. Given that successful exits have historically involved predominantly male founders, these historical patterns influence current decision-making processes. The venture assessment process perpetuates existing patterns, as prior investment experiences reinforce the tendency to stick to established strategies.

Institutional variation: The gender gap is lower in larger and older venture capital firms with more formal evaluation systems in place. This suggests the problem is not inevitable but rather reflects specific organizational and decision-making structures that can be modified.

Investor-level dynamics

The disparity extends beyond entrepreneur evaluation to investor success patterns. Female venture capitalists do not benefit from the track record of their male colleagues in the same way that male investors show improved investment success in the presence of more successful colleagues. This differential is particularly significant in smaller firms.

Diverse teams, with women on board, lead to improved performance through more balanced decision-making, not only in ventures but also for venture capital firms themselves.

Market implications

The Atomico report describes this as a “vast set of companies remain critically underfunded.” This language is precise. We are not discussing a small cohort or niche segment. We are discussing roughly one-quarter of the European tech founding population (combining all-women and mixed-gender teams) receiving single-digit to low double-digit percentages of available capital.

For investors focused on fundamental analysis rather than pattern-matching, this represents a systematic market inefficiency. When large segments of the market receive capital allocations potentially misaligned with business quality, pricing opportunities may emerge for disciplined buyers.

What has not changed

Policy discussions and ecosystem rhetoric have increased substantially over the past decade. Implementation and capital allocation have not kept pace.

The structural barriers remain intact. Female founders and gender-diverse teams continue to face significantly higher bars for capital access across identical ventures and comparable performance.

Bottom line

After twenty years of data, the pattern is clear and persistent. Female-founded and mixed-gender teams receive a fraction of VC funding relative to their numbers and performance.

This is not improving. Mixed-gender teams captured 32% of Series B funding in 2016. By 2025, that figure dropped to 22%. The trajectory is moving in the wrong direction.

For the European tech ecosystem, this represents a structural inefficiency in capital allocation that has persisted for two decades despite documented performance parity.

Sources: Atomico State of European Tech 2025; Brooks et al. (2014); Guzman & Kacperczyk (2019); Kanze et al. (2018); Malmström et al. (2020); Gompers et al. (2022); Calder-Wang & Gompers (2021); Ewens & Townsend (2020); Edwards & McGinley (2019); Franke et al. (2006); Eagly & Karau (2002)

When examining a new deal, venture capitalists rely heavily on pattern recognition, connecting dots between past experiences and present opportunities. This ability is widely considered essential to investment success.

But cognitive science research reveals a critical distinction we often miss: the difference between genuine pattern recognition and what might better be termed “pattern projection.”

The science of expert judgment

Let’s start with what research actually tells us about expert decision-making.

Psychologist Gary Klein’s groundbreaking research on naturalistic decision-making showed that experts don’t actually analyze situations objectively. Instead, they engage in rapid pattern matching, comparing new scenarios against a mental library of previous experiences.

This Recognition-Primed Decision model explains why experienced investors can often assess opportunities quickly. But it also reveals a vulnerability, i.e. our tendency to project familiar patterns onto new situations regardless of fit.

The empirical evidence on expert decision-making comes from multiple fields. Philip Tetlock’s 20-year study of expert predictions, published in his book “Superforecasting,” demonstrated that specialists often perform worse than generalists when facing changing environments, precisely because their deeply ingrained patterns become liabilities when contexts shift.

The representativeness heuristic in venture decisions

The psychological mechanism underlying this pattern projection is what Nobel Prize winners Kahneman and Tversky identified as the representativeness heuristic. More specific, our tendency to judge probability based on similarity to mental prototypes rather than actual statistical likelihood.

For venture investors, this manifests in specific, documentable ways:

1. Founder prototype bias: Research published in the Journal of Business Venturing shows investors systematically favor founders who match their mental prototype of “successful entrepreneurs,” often based on superficial similarities to previous winners.

2. Business model familiarity: Studies from the Strategic Management Journal demonstrate how investors overweight similarities to previous successful investments while underweighting critical differences in timing, technology, or market structure.

3. Historical pattern extrapolation: Research on investment decision-making reveals how investors frequently assume similar companies will follow similar trajectories, despite different market conditions.

The paradigm shift problem

The pattern projection problem becomes most damaging during paradigm shifts. Academic literature on disruptive innovation shows this consistently across technology waves.

Clayton Christensen’s research documented how established patterns of evaluation become actively misleading when technologies follow a disruptive rather than sustaining path. The mental models that served investors well in one era become obstacles in the next.

Consider three well-documented historical examples from the academic literature:

1. The SaaS transition: Early SaaS companies showed different unit economics than traditional software firms. Research from Harvard Business School documents how investors who applied conventional software metrics systematically undervalued these opportunities.

2. Mobile app economics: When mobile applications emerged, many investors evaluated them using web application metrics. Research on valuation errors shows how this pattern mismatch led to systematic undervaluation of mobile-first opportunities.

3. Platform business models: Studies on two-sided markets demonstrate how investors accustomed to linear business models often misunderstood the network effects and different growth patterns of platform businesses.

Building better pattern recognition

If our investment patterns can become intellectual liabilities, how do we improve?

Research on expert performance suggests several evidence-based approaches:

1. Deliberate variance testing: Studies of expert development show that deliberately seeking out exceptions to patterns enhances nuanced understanding. For each investment thesis, actively search for environments where the pattern might not hold.

2. Multiple mental models: Research on forecasting accuracy demonstrates that maintaining several competing explanatory frameworks produces better predictions than relying on a single model. The best investors regularly ask: “What’s another way to interpret this situation?”

3. Systematic invalidation: Cognitive science research shows we naturally seek confirmation. Counteracting this requires deliberately structured processes to test assumptions. For every investment thesis, ask: “What evidence would invalidate this pattern?”

4. Learning loops: Research on expert performance emphasizes the importance of structured feedback. Regular portfolio reviews focused on pattern validation/invalidation rather than just performance metrics improve pattern recognition over time.

Practical application for your investment process

The research points to specific, implementable changes to improve pattern recognition:

1. Pre-mortem analysis: Before investing, conduct a structured exercise where you assume the investment has failed and analyze why. Research shows this counteracts overconfidence and reveals blind spots in pattern matching.

2. Decision journals: Document investment rationales at the time of decision, including specific pattern matches that influenced the choice. Review these periodically to refine pattern accuracy.

3. Pattern diversity: Research on team decision-making shows diverse experience bases improve pattern recognition. Ensure investment discussions include voices with different pattern libraries.

Beyond intuition

The most sophisticated venture investors today are moving beyond pure intuition-based pattern matching toward a more calibrated approach, using the power of pattern recognition while guarding against its limitations.

This isn’t about abandoning intuition. Research on expert performance shows that intuitive pattern recognition remains valuable, but its accuracy improves dramatically when complemented by structured processes that test and refine these patterns.

The question for your firm: Which of your current investment patterns deserve reexamination? And what process could you implement to continuously refine them?

My PhD research focuses at decoding the behavioral processes and decision-making in venture assessment. Are you a VC investor and interested in the results of this study? Send an email to participate in the survey.