👋 Hi, I’m Andre and welcome to my newsletter Data Driven VC which is all about becoming a better investor with Data & AI. Join 35k+ readers from firms like a16z, Accel, Index, Sequoia, and more.
Brought to you by Affinity - Campfire 2025 is coming to London!
Private capital teams are drowning in emails, data, and shifting markets—just as competition for the best deals reaches new heights. On October 28, 2025, Affinity Campfire London will bring together leading investors, dealmakers, and operators to turn that noise into actionable insight.
Join peers for an afternoon of networking and hands-on learning. From smarter sourcing and fundraising to practical AI integration, you’ll take home proven tactics, hands-on demos, and real-world success stories you can deploy immediately.
Measuring True Efficiency in Venture Exits
John Rikhtegar introduces the Exit Velocity Index (EVI) that ranks 3,317 North American VC-backed exits above $10M between 2010 and 2025, factoring in exit value, total primary equity raised, and time to exit. It highlights how efficiently enterprise value was built, rather than just how big the exit was.
Long Duration, Lower EVI: Some of the most famous exits like Yahoo (23 years), StubHub (20), Reddit (19), Procore (19), and ServiceTitan (17), took unusually long to exit, resulting in lower efficiency scores despite their large valuations.
Low Efficiency Multiples: Well-known companies like Scopely (2.9x), pony.ai (3.3x), Postmates (4.5x), Oscar Health (4.8x), and Uber (5.3x) delivered lower returns on invested capital relative to their size, falling short on capital efficiency.
Exceptional Efficiency Leaders: At the top, Mir achieved an EVI of 100 ($400M exit in 2 years on $2M raised), outperforming even WhatsApp ($22B exit in 5 years on $60M raised). Only 27 exits exceeded an EVI above 20, or 0.81% of all exits analyzed.
✈️ KEY TAKEAWAYS
EVI shows that speed and efficiency are stronger indicators of performance than headline valuations. Many large exits underperform when adjusted for time and capital, while a small subset of highly efficient companies drive outsized value creation.
Should Founder Vesting Last 6 Years Instead of 4?
Peter Walker from Carta argues that the standard 4-year founder vesting schedule is outdated. With IPOs now taking 12–14 years, he suggests extending vesting to (6 years or even 8) to better align incentives and reduce “dead equity” when co-founders leave early.
Better Alignment with Reality: A 6-year vesting schedule matches the median startup journey toward IPO more closely than 4 years, which often ends around a Series A stage when much of the company is still unbuilt.
Investor Signaling and Retention: Longer vesting reduces the risk of dead equity and signals long-term intent to investors, though critics argue that doing so merely to please VCs misses the point of founder ownership.
Key Challenges and Counterpoints: Opponents note issues like founder re-vesting at new rounds, loss of leverage in equity negotiations, and fairness for founders who self-fund or exit early. Suggestions include milestone-based vesting or extending exercise windows as alternatives.
k
✈️ KEY TAKEAWAYS
While longer vesting can reduce equity waste and align with modern exit timelines, it risks disincentivizing founders and complicating future fundraising. The debate highlights a broader shift toward treating equity as both a retention tool and a governance mechanism.
Upgrade your subscription to access our premium content
LP list with 2500+ contacts
350+ deep dive articles
100+ master classes
Prompt database
AI copilots
& more
Does “Below Expectations” Mean You’re Out?
Matt Schulman from Pave analyzed performance ratings from over 40,000 employees to see how they correlate with turnover. Using data from Q1 2024 merit cycles, he measured the share of employees who left their companies by February 2025, revealing clear survival differences by rating bucket.
Turnover by Rating: Employees rated “Below Expectations” saw a 58% turnover rate within a year, the highest among all groups. “Meets Expectations” employees had a 22.4% turnover rate, while “Exceeds Expectations” dropped to 13.2%.
Promotion Effect: Only 11.4% of employees were promoted, but they had the lowest turnover, 11.5%. The link between strong performance and retention suggests that top performers and those recognized through advancement are least likely to leave.
Inflation vs. Reality: Despite HR pressure to “fix” inflated ratings, data shows that most teams don’t harbor many low performers. Only 9% of employees received a “Below Expectations” rating, implying many managers already manage out underperformers before reviews.
✈️ KEY TAKEAWAYS
Performance reviews remain a key retention predictor. While poor ratings often precede exits, the data shows that not every “Below Expectations” employee is immediately out (42% stay beyond a year) suggesting that context and culture shape outcomes more than ratings alone.
Are Software Companies Cutting R&D to Grow GTM Spend?
Tomasz Tunguz analyzed six years of public software company data to test whether firms reduce R&D to fuel sales and marketing during downturns. Contrary to expectations, the data shows R&D spending has risen while GTM investment has fallen as a percentage of revenue.
R&D Spend Rising: Across all public software companies, R&D spend grew from 24% to 28% of revenue since 2016. Product-led growth (PLG) companies led the trend, increasing R&D intensity from 27.5% to 33%.
Sales & Marketing Down: Both PLG and sales-led companies reduced S&M spend relative to revenue, especially during the COVID-era efficiency surge. This shift suggests a structural rebalancing toward product investment rather than top-heavy GTM expansion.
Efficiency vs. Growth: Despite reduced GTM spend, overall sales efficiency has also declined, hinting that lower S&M outlays might be slowing growth. The last three quarters, however, show a modest rebound in GTM spending, potentially signaling a cyclical correction.
✈️ KEY TAKEAWAYS
Public software companies now favour sustained R&D investment, particularly those with PLG models, while S&M budgets tighten. The pattern suggests long-term bets on product quality and self-serve growth over sales-driven expansion, but with early signs of rebalancing.
Share this article with others who might benefit.
Series A Valuations and Unicorn Probability
Stanford professor Ilya Strebulaev shared new data linking Series A valuations to the likelihood of becoming a unicorn. The analysis, based on data from the Stanford GSB Venture Capital Initiative, shows a clear correlation between higher early valuations and eventual billion-dollar outcomes.
Probability by Valuation Tier: Companies valued below $10M at Series A have a 1% chance of becoming unicorns. That probability climbs to 4% in the $20–30M range and stabilizes around 3–5% for mid-tier valuations.
Breakout Threshold: The largest jump occurs at valuations above $100M, where the likelihood of reaching unicorn status rises to 9%, a ninefold increase over the lowest tier. These valuations likely reflect exceptional investor conviction tied to tangible early traction or a massive market.
Signal, Not Causation: Strebulaev emphasizes that a high valuation doesn’t cause success—it signals market validation. Companies earning large Series A valuations typically already show breakout growth, product-market fit, or network effects that justify investor confidence.
✈️ KEY TAKEAWAYS
A high Series A valuation is less a vanity metric and more a reflection of strong fundamentals. While only a small fraction of startups reach unicorn scale, early valuations remain one of the clearest predictors of eventual success.
Does Communication Matter in Technical Interviews?
Aline Lerner analyzed over 100,000 coding interviews on intervierwing.io, scoring each across coding ability, problem-solving, and communication. The data shows that communication plays a smaller role in hiring outcomes, except for senior engineers.
Technical Skills Dominate: Candidates scoring 4 in both coding and problem-solving but only 2 in communication passed 96% of the time. Meanwhile, those with strong communication but slightly weaker technical scores advanced only 88%.
Relative Weighting: Dropping one point in coding or problem-solving increased rejection odds by roughly 7X. Dropping one point in communication only doubled rejection odds, suggesting technical skill carries 3–4 times more weight overall.
Experience Shifts the Balance: Among senior engineers, communication mattered more. At that level, it correlated strongly with advancement since clear reasoning and decision-making become more important than raw problem-solving.
✈️ KEY TAKEAWAYS
Communication alone rarely compensates for weaker technical ability, but its importance grows with seniority. For junior and mid-level engineers, passing still depends almost entirely on code and problem-solving performance.
Thanks to Jérôme Jaggi for his help with this post.
Stay driven,
Andre
PS: Check out Affinity’s Campfire Conference in London here
