How I Invest with David Weisburd

BITKRAFT's Jerry Ye: the Anthropic talent signal, card-panel data and the science of picking winners

Key points

Key takeaways from a How I Invest with David Weisburd interview with Jerry Ye, partner at BITKRAFT Ventures (November 2026):

Talent as a signal. Employees leaving highly paid roles at big firms to join a startup is a powerful investment signal — each senior hire forgoing a six-figure salary for equity is effectively betting their career, comparable to how a top venture fund on the cap table signals conviction.

The Anthropic read. Talent data showed Anthropic on average hiring people with stronger pedigree markers — better schools and experience — and surpassing OpenAI on that measure years before its rise, with lower engineer churn; Ye flags the art-versus-science caveat around weighting senior versus junior hires.

Card data as ground truth. A 3–5% sample of US credit and debit card spend, resold by panels such as Yodlee and Consumer Edge, can reveal revenue trends and retention for consumer companies before the media or other VCs notice — hedge funds have used it for years and largely exhausted the alpha in public markets, while venture is still early.

Grammarly via cohorts. Card-panel data flagged Grammarly as inflecting with SaaS-like retention, supporting SignalFire's early investment; years later Grammarly used the same cohorts for a reported billion-dollar non-dilutive user-acquisition financing facility.

User acquisition financing. For post-product-market-fit companies, UA financing funds monthly growth spend in exchange for a capped revenue share repaid as cohorts generate revenue — thresholds such as 40% ROI in month one, 50% in month two, with funding paused if targets are missed.

Sector-dependent founder maths. PhDs and domain experience correlate with better outcomes in enterprise, but matter far less for consumer companies where traction is what counts; the composition of the founding team and later hires is itself a mini-round of conviction.

Measuring the manager. Venture firms should track which deals they saw, missed or passed on, and LPs could assess a manager's true access using CRM records, emails and meetings — mirroring how public investors scrutinise process.

Data maturity gap. Most venture teams struggle to find signal in billions of data points when they are used to Excel data rooms; crossover funds run eight-figure data budgets with 20–30-person engineering teams, leaving alternative data largely untapped on the venture side.

Read more: How I Invest with David Weisburd (YouTube)

More top stories