Are we paying back supply acquisition profitably across horizons?
№ 046 · Supply LTV/CAC
Definition
MetricSupply-side lifetime value over supply-side acquisition cost, margin-adjusted
Unitratio (x)
Bar showing seller LTV (lifetime fees generated) divided by seller acquisition cost. Below 3 is questionable. Build by computing trailing fees-from-seller and dividing by cost-to-acquire-seller. Channel comparison reveals where to scale supply spend. Example: $1,800 LTV / $400 CAC = 4.5x.
- LTV/CAC
- Lifetime value to acquisition cost ratio. LTV divided by CAC. Above 3x is the common health threshold.
- LTV
- Lifetime value. Average gross profit per customer per period divided by that period churn rate, or summed over expected life.
- CAC
- Customer acquisition cost. Fully loaded sales and marketing spend in a period divided by new customers won in that period.
Benchmarks
| Bottom 30% | Median | Top 30% |
|---|---|---|
| below 3x margin-adjusted, or ratios computed on GMV instead of net revenue (which overstates LTV by the inverse of the take rate), or blended across sides hiding a loss-making side | approx 3x floor treated as viable for marketplaces; supplier LTV must be computed on net revenue (take rate on the seller's GMV) and matched to the shorter of the two sides' lifespans | 4x+ on the supply side, the ratio a16z cites investors looking for on both sides of a marketplace; achieved through long supplier lifespans and GMV expansion of retained sellers |
consumer marketplaces, VC diligence norms 2019-2026; LTV on contribution after take rate and variable costs, CAC fully loaded per acquired transacting supplier, lifespans matched across sides per the Valentini methodology · The 4x figure is a stated diligence bar, not a measured distribution; no large-sample supply-side LTV/CAC dataset exists. Supply CAC is often dominated by onboarding and activation cost, not media. When supplier lifespan exceeds buyer lifespan, only a fraction of supply CAC should be charged to a given demand cohort, so naive ratios understate supply economics.
Category split omitted: No two independent Tier 1-2 sources publish supply-side LTV/CAC distributions by category.
When it looks bad
The ratio declines cohort over cohort because supply CAC rises (bounties, sales headcount) while early-tenure GMV per supplier falls; or the curve of cumulative supplier contribution never crosses CAC within the modeled lifespan.
Supply CAC rises $180 to $310 per activated seller over a year while m12 contribution per seller falls $520 to $390; the ratio drops 2.9x to 1.3x on the newest cohorts.
What to do about it
- Recompute the ratio properly before optimizing it: LTV on net take-rate revenue less variable costs, CAC per transacting (not signed-up) supplier, lifespan capped at the shorter side per the two-sided methodology; many bad ratios are measurement artifacts.
- Cut supply CAC by paying for outcomes: switch recruiter and referral bounties from signup to first-sale, and let seller-to-seller referral carry more of the mix, since referred sellers arrive pre-qualified.
- Raise supplier LTV through the expansion toolkit (demand analytics, promoted listings, financing); a16z ties best-in-class supply economics to retained sellers expanding GMV 2-3x, which compounds LTV without touching CAC.
- Segment the ratio by supplier type and stop acquiring the segments that never pay back; a blended 3.5x can hide a professional-seller segment at 6x and a casual segment at 0.8x.
Sources
- Andreessen Horowitz (Olivia Moore) investors calculate LTV vs CAC for both sides of the marketplace and in many cases look for a 4x+ ratio; higher GMV retention directly raises LTV and marketing headroom a16z.com ↗
- Gianluca Valentini (Equity Pitcher Ventures) total marketplace CAC is supply CAC plus demand CAC weighted by conversion frequency and lifespan; e-commerce style single-sided ratios do not apply linkedin.com ↗
- Gianluca Valentini (Gringotts Ventures) when supply lifespan exceeds demand lifespan only a fraction of supply CAC should be allocated per demand cohort; worked ride-hailing example with 5-year driver vs 3-year rider lifespans medium.com ↗
- Foundry CRO marketplaces treated at a 3:1+ floor with two-sided tracking; improving marketplace LTV:CAC from 2:1 to 3:1 materially rerates valuation per a16z guidance foundrycro.com ↗
- Revenue Map corroboration: marketplace benchmark tables set the good threshold above 4:1 vs 3:1 for single-sided models because acquisition is paid on both sides revenuemap.app ↗
- Eightx payback context: public marketplace buyer economics modeled at very short paybacks due to brand-search demand; supply-side paybacks run longer and must be modeled separately eightx.co ↗