Key Takeaways

  • Hedge funds decided a decade ago that consumer credit card data held no fresh alpha, yet early-stage venture investors are only now tapping the same feeds.
  • Public market funds pay seven- to eight-figure annual fees for real-time transaction panels to beat quarterly earnings, while venture funds extract equal value from delayed, cheaper slices.
  • Tracking a 3% to 5% sample of all US debit and credit card spending lets outside investors calculate exact cohort retention and repeat purchase curves on private companies without founder permission.
  • Quantitative tracking revealed Grammarly was generating enterprise-grade SaaS retention curves on consumer software long before standard venture rounds priced it in.

The Decade-Long Lag Between Public and Private Panels

Around ten years ago, Jerry Ye sat at a dinner in New York with a room of hedge fund managers. The group reached a blunt consensus: public market alpha from credit card data was dead. Every major fund had already bought the panels, mapped the tickers, and priced real-time merchant revenue into quarterly forecasts.

When Ye showed up at an alternative data conference in New York shortly after, the divide was obvious. “I was literally the only one representing a venture fund there,” (play from 14:33) Ye recalls. “And you saw everyone else on the hedge fund PE side.” (play from 14:37)

Public equity desks treat transaction data as a high-speed arms race. They pay seven or eight figures a year for low-latency transaction streams because thirty seconds can make or break an earnings print. Venture firms do not need millisecond delivery. A venture investment lives on a seven- to ten-year horizon. For early and growth investors, a thirty-day delayed panel yields the exact same strategic clarity at a tiny fraction of the cost.

Reconstructing Retention from a 5% Spend Sample

Consumer transaction panels track between 3% and 5% of all credit and debit card spend across the United States. For an early-stage startup, that sample size is plenty to model trajectory. If a consumer app processes twenty thousand card transactions a month, several hundred will land inside the panel.

The real prize is not top-line volume. The prize is cohort retention. Private company pitch decks routinely present blended growth charts that mask high customer churn behind heavy paid acquisition. By isolating unique anonymized cards over six, twelve, and twenty-four months, an outside investor can reconstruct true cohort decay curves.

“Keep in mind that you can figure out what retention rates look like as well,” (play from 16:54) Ye notes. “We saw a company that was inflecting and not only was it inflecting, you saw really high retention rates which were similar to what you would see in a SaaS company. And that led to investment there.” (play from 16:59)

That dynamic played out directly with Grammarly. Standard consumer tech investors long treated writing assistants as low-retention utilities. The transaction panel proved the opposite: subscribers stayed on the product with net revenue retention figures matching enterprise business software. Investors who tracked the panel did not have to wait for audited metrics or a warm introduction to write the check.

The Structural Blind Spot in Private Data

The advantage works because private market data remains messy. Public merchants come cleanly tagged to stock tickers. Early-stage startups do not. A seed or Series A company often processes revenue through generic payment processors or changing legal entity names. Mining early signals requires manual entity resolution, cleaning up messy merchant strings before the rest of the street realizes which startup owns the billing tag.

The method has boundaries. It only catches direct card swipes, online checkouts, and app store transactions. B2B enterprise software with wire transfers, government contracts, and enterprise purchase orders bypass consumer card panels entirely. But for direct-to-consumer software, market inflections appear on card panels quarters before founders open a formal data room.

Why It Matters

Proprietary access in growth investing is shifting from founder networks to proprietary data processing. When an investor spots cohort stabilization and pricing power from the outside, the diligence phase flips from verification to preemptive origination. Funds relying solely on inbound founder decks and bank-run auctions will consistently pay peak prices for companies whose underlying retention inflection was visible to quantitative peers eighteen months earlier.