🔥How to Predict Startup Success… | Data Driven VC Summary

Summary of newsletter by Andre Retterath, April 26, 2026

1. Key Themes

Theme 1: Founder Team Dynamics Trump Conventional Wisdom on Co-Founder Selection

Counter-intuitively, co-founders who are strangers outperform those with prior working relationships. The data shows that "co-founders who never worked together saw 26% higher exit outcomes than former coworkers." This suggests that the conventional VC heuristic of preferring teams with pre-existing working relationships may be systematically mis-calibrated.

Theme 2: Prior Founding Experience and Founder Demographics Are Strong Predictive Signals

Repeat founders and diverse founding teams carry measurable alpha. The article highlights that "previous founding experience increases success rates by 41%, and why teams with a previously exited female CEO outperformed all-male teams by nearly 2x." This makes prior-exit experience — especially in female-led teams — one of the highest-signal data points available to early-stage investors.

Theme 3: Time-Series Behavioral Data Beats Static Pedigree Signals

Resume credentials are weak predictors; operational momentum data is strong. The article notes that "pedigree (think top-tier schools) barely moves the needle, while time series data like early headcount growth and hiring disciplines offers strong predictive signal." This is a significant market shift: investors over-indexing on brand-name schools and logos are likely flying blind, while those tracking hiring velocity gain a durable edge.

Theme 4: AI Commoditizes Data Processing — Durable Alpha Lives in the "Why"

As AI levels the playing field on data execution, the interpretive layer becomes the moat. The article states that "AI commoditizes the 'how' of data processing, and why durable alpha lives in the 'why'." For investors and operators alike, the implication is that proprietary frameworks for interpreting alternative data — not just accessing it — will be the source of lasting competitive advantage.

Theme 5: Data Should Drive Attention Prioritization at Early Stages, Not Winner Selection

At pre-seed and seed, the role of data analytics should be reframed from picking winners to allocating GP bandwidth. The article makes "the case for using data to prioritize GP attention at pre-seed and seed rather than trying to pick winners." This is a meaningful operational reframe: data as a triage and routing tool rather than a prediction oracle.

2. Contrarian Perspectives

Contrarian 1: Co-Founders Who Never Worked Together Produce Better Outcomes

The conventional VC playbook prizes founding teams with shared history and prior working chemistry. The data directly contradicts this: "co-founders who never worked together saw 26% higher exit outcomes than former coworkers." The likely explanation is that unfamiliar co-founders bring complementary skill sets rather than overlapping ones, and may engage in more rigorous role definition early on. This should cause investors to rethink how they weight team cohesion history in diligence.

Contrarian 2: Elite Educational Pedigree Is a Noise Variable, Not a Signal

The VC industry has long over-weighted top-tier school credentials as a proxy for founder quality. The article explicitly states that "pedigree (think top-tier schools) barely moves the needle," while operational metrics like headcount growth and hiring discipline carry real predictive power. Investors and founders who've structured their pitches around pedigree-as-signal should recognize this as a weak variable that crowds out more predictive data.

Contrarian 3: Soft Traits Like Grit and Humility Outperform Anything on a Resume

Against the data-obsessed framing of the piece, a key finding is that intangible character traits remain the most durable predictors: "grit, humility, and the ability to iterate on new data outweigh anything you can read on a resume." This is contrarian within a data-driven investing context — suggesting that even with sophisticated alternative data models, qualitative founder assessment remains irreplaceable.

3. Companies Identified

Tribe Capital

Description: Data-driven venture capital firm
Why mentioned: Featured as a case study contributor; Jake Ellowitz presented research on predicting startup success with alternative data
Quote: "experts from four leading VC and PE firms presented their latest research on what truly drives investment performance and predictable founder outcomes"

Basis Set

Description: AI-focused venture capital firm
Why mentioned: Featured as a case study contributor; Rachel Wong presented research on alternative data signals for startup success
Quote: "Learnings from Tribe Capital, Basis Set, Outcast Ventures, and Bertram Capital"

Outcast Ventures

Description: Venture capital firm
Why mentioned: Featured as a case study contributor; Amy Lin presented research at the Virtual DDVC Summit 2026
Quote: "Learnings from Tribe Capital, Basis Set, Outcast Ventures, and Bertram Capital"

Bertram Capital

Description: Private equity firm
Why mentioned: Featured as a case study contributor; William Mataker presented on deterministic vs. probabilistic models for investment decisions
Quote: "Learnings from Tribe Capital, Basis Set, Outcast Ventures, and Bertram Capital"

Wispr Flow (Sponsor)

Description: AI voice-to-text productivity tool
Why mentioned: Paid sponsor; converts voice dictation into polished text across apps
Quote: "89% of messages sent with zero edits. Used by teams at OpenAI, Vercel, and Clay."

4. People Identified

Rachel Wong

Description: Investor/Researcher at Basis Set
Why mentioned: Panelist at Virtual DDVC Summit 2026; presented research on alternative data and startup success prediction
Quote: "a unique deep dive on 'How to Predict Startup Success With Alternative Data' with Rachel Wong from Basis Set"

Jake Ellowitz

Description: Investor/Researcher at Tribe Capital
Why mentioned: Panelist at Virtual DDVC Summit 2026; contributed research on data-driven investment performance
Quote: "Jake Ellowitz from Tribe Capital"

Amy Lin

Description: Investor at Outcast Ventures
Why mentioned: Panelist at Virtual DDVC Summit 2026; contributed research on founder outcomes
Quote: "Amy Lin from Outcast Ventures"

William Mataker

Description: Investor at Bertram Capital
Why mentioned: Panelist at Virtual DDVC Summit 2026; contributed research on deterministic vs. probabilistic modeling
Quote: "William Mataker from Bertram Capital"

Andre Retterath

Description: Author of Data Driven VC newsletter; GP at Earlybird Venture Capital (per newsletter context)
Why mentioned: Author and curator of the newsletter; host of the DDVC Summit
Quote: "Hi, I'm Andre and welcome to my newsletter Data Driven VC which is all about becoming a better investor with data and AI."

5. Operating Insights

Insight 1: Use "Weather Reports" on Sector Supply/Demand to Drive Negotiation Posture

Valuation discipline should be informed by real-time market dynamics, not just deal-specific fundamentals. The article advocates for "why 'weather reports' on sector supply and demand should drive your negotiation posture and valuation discipline." For operators raising capital and investors deploying it, tracking sector-level momentum data can shift negotiating leverage and prevent over/underpaying in heated or depressed markets.

Insight 2: Deploy Deterministic Models to Codify Human Judgment, Probabilistic Models for High-Liquidity Sectors

Model selection is not one-size-fits-all — context determines architecture. The article highlights "when to deploy deterministic models to codify human judgment vs. probabilistic models for high-liquidity sectors." Investors building internal data infrastructure should match their modeling approach to the liquidity and information density of their target sectors, rather than applying a universal framework.

Insight 3: Early Headcount Growth and Hiring Discipline Are Among the Strongest Operational Signals

Time-series people-data is more predictive than credentials. The article notes that "time series data like early headcount growth and hiring disciplines offers strong predictive signal." For founders, this underscores the importance of building talent infrastructure intentionally and rapidly post-funding — those patterns are being observed and modeled by sophisticated investors.

6. Overlooked Insights

Overlooked Insight 1: Previously Exited Female CEOs Nearly 2x All-Male Teams

Buried within a broader point about founding experience is a highly specific and actionable data point: "teams with a previously exited female CEO outperformed all-male teams by nearly 2x." This is not framed as a DEI argument but as a pure performance signal — and it's striking. Most investor pattern-matching still skews toward all-male technical founding teams, meaning this cohort may be systematically undervalued and underpriced.

Overlooked Insight 2: The Model Type Choice (Deterministic vs. Probabilistic) Is Sector-Dependent

The article briefly distinguishes between two fundamentally different modeling approaches depending on sector conditions — an often-overlooked nuance in data-driven investing discussions. The distinction between "deterministic models to codify human judgment" versus "probabilistic models for high-liquidity sectors" implies that applying the wrong model to the wrong market context could produce systematically flawed investment signals, even with high-quality underlying data.