Luca Barberis

When does each supply cohort pay back its acquisition cost, and how does the return compound over time?

№ 048 · Supply ROAS curve

01

Definition

MetricCumulative net revenue from a supply cohort over the cohort's acquisition spend, by month since onboarding

Unitcumulative ratio (x) by month N; breakeven month

Cumulative supplier-side revenue divided by supplier acquisition cost over months. Newer or better channels reach 1.0x faster. Build by tracking per-cohort fees from sellers minus their CAC monthly. Example: paid acquisition crosses 1.0x at M9; partnership channel at M4 (partnerships are the winner).

ROAS
Return on ad spend. Revenue attributed to a campaign divided by the spend on that campaign.
CAC
Customer acquisition cost. Fully loaded sales and marketing spend in a period divided by new customers won in that period.
02

Benchmarks

Bottom 30%MedianTop 30%
curves that flatten below 1.0x, meaning the cohort never repays its acquisition; each newer cohort's curve sitting below the last at the same tenurebreakeven inside the first year for healthy consumer marketplace supply, consistent with 80-95% early GMV retention plateauing near 45-50% at m12; slope flattens as the cohort maturessupply cohorts crossing 1.0x well inside 12 months and compounding after breakeven because retained sellers expand GMV 2-3x; curve slope increasing or stable after m6

consumer marketplace supply cohorts, revenue defined as take-rate net revenue on cohort GMV, spend fully loaded including bounties and onboarding cost; derived from GMV retention shapes and diligence payback norms rather than a published ROAS distribution · No source publishes supply-side ROAS curves at sample; the shape is derived from the a16z GMV retention benchmarks plus two-sided CAC methodology, so treat the breakeven month as model-specific. This is supply ROAS on take-rate revenue in a marketplace; it is not the channel ROAS of the Marketing section and numbers must not be copied across.

Category split omitted: No independent Tier 1-2 sources publish supply-side ROAS curves by category.

03

When it looks bad

Each successive cohort's cumulative curve launches lower and flattens earlier, and the newest cohorts visibly asymptote below the 1.0x line while acquisition spend per cohort keeps rising.

The Q1 cohort reached 1.0x at m8; Q3 sits at 0.5x at m8 and its slope has already flattened; projected terminal value 0.7x on $240K of cohort spend.

04

What to do about it

  • Plot every cohort against a stage-gate: minimum ratio by m3 and m6 derived from the historical curves of cohorts that eventually crossed 1.0x, and cut spend in channels whose cohorts miss the gate twice.
  • Bend the curve up with activation, not spend: time-to-first-sale is the highest-leverage input to early cohort revenue per the a16z retention drivers, so week-one boosted placement and listing rescue directly steepen the ROAS curve.
  • Extend the compounding tail by investing in the retained sellers: expansion tooling and financing for the surviving 50-70% multiply post-breakeven slope, which is where supply ROAS is won.
  • Amortize honestly: include bounties, onboarding operations and early liquidity subsidies in cohort spend; a curve that only counts media will cross 1.0x on paper and never in cash.
05

Sources

  1. Andreessen Horowitz (Olivia Moore)GMV Retention: The Marketplace Metric Most Ignore · 2022 · a16z seed to Series B consumer marketplace dataset (approx 18 months of company data) plus 16 public marketplace filings supply GMV retention 80-95% early plateauing 45-50% at m12, best-in-class 100%+; retention level fixes how fast cumulative take-rate revenue climbs against acquisition cost a16z.com ↗
  2. Gianluca Valentini (Gringotts Ventures)The Secret Formula to Match Demand CAC vs. Supply CAC in Marketplaces · 2019 · VC diligence methodology supply CAC amortization must match supplier lifespan; long-lived supply justifies longer breakeven than the demand side medium.com ↗
  3. EightxAverage ecommerce CAC payback by business model 2026 · 2026 · modeled from public 10-K disclosures (Etsy, Chewy, Warby Parker, FIGS, BARK) plus published benchmarks business-model payback modeling from public filings: marketplace models can reach very short paybacks when acquisition is organic; the method transfers, the numbers do not eightx.co ↗
  4. Foundry CROLTV:CAC Ratio Benchmarks 2026 · 2026 · synthesis of Skok framework, a16z marketplace guidance and 2026 model benchmarks per-market ROAS and payback tracking beats blended; local supply density changes curve shapes market by market foundrycro.com ↗
  5. Everything Marketplaces (Marketplace Capital)Fundraising Guide and Benchmarks for Marketplaces in 2026 · 2026 · stage benchmarks synthesized from marketplace fund deal flow investor lens on GMV vs net revenue applies to cohort curves: supply ROAS must be presented in net revenue terms, take-rate dependent everythingmarketplaces.com ↗