--- title: "Buyer subsidies outran the repeat GMV they were meant to unlock" description: "Acquisition subsidies were justified on lifetime GMV that most subsidized buyers never delivered. Marketing & margin audit: $66k/yr in margin identified." url: https://marginfix.ai/cases/buyer-subsidies-outran-the-repeat-gmv-they-were-meant-to-unl/ lastmod: 2026-10-28 --- Buyer subsidies outran the repeat GMV they were meant to unlock. [Request an audit →](/book/) Marketplace · B2B wholesale # Buyer subsidies outran the repeat GMV they were meant to unlock. This marketing & margin audit surfaced **$66k in annual margin** in a Marketplace business, evidenced, senior reviewed, and delivered in 7 days. $3–8M annual revenue B2B wholesale Focus: Buyer subsidies +26% buyer payback +26% repeat GMV payback 7 days to findings Payback on subsidized buyers Before audit 0.8× → After fix 1.3× ## The business A B2B wholesale marketplace acquired new buyers with onboarding subsidies and first order incentives, justified on the lifetime GMV each buyer was expected to generate. Buyer growth was strong, and the lifetime value assumption underpinning the subsidies had been set early and applied broadly, without ever being tested against how those buyers actually behaved over time. ## What triggered the audit Buyer acquisition looked efficient against projected lifetime GMV, but cash didn’t reflect it, hinting the LTV assumption was too generous. The audit tested the actual repeat GMV behavior of subsidized buyers against the lifetime value the subsidies had been justified on, cohort by cohort. ## What the audit found The subsidies were being justified on a lifetime GMV that most buyers never delivered. The early LTV assumption, set when the marketplace was young and applied uniformly ever since, assumed subsidized buyers would return and reorder at a rate the actual cohort data flatly contradicted. A large share placed one subsidized first order and rarely came back, so the incentive was effectively a one off cost with no lifetime value behind it, while the healthy projected payback existed only on the spreadsheet. Because buyer counts kept growing and the assumption went unchallenged, the marketplace continued scaling acquisition on economics that simply weren’t real. Measured against genuine repeat GMV, the subsidized segment paid back far worse than assumed, and realigning it improved buyer payback by 26%. ◉ How we produced this finding To reach the buyer subsidies finding, MarginFix tracked real marketplace cohorts over time rather than trusting day one numbers, exposing how the economics actually behaved. The evidenced read was checked against the client’s own data and approved by a named senior auditor before release. Data sources: Cohort level acquisition, retention and revenue data tracked over time, joined to fully loaded acquisition cost, so buyer subsidies is judged on genuine lifetime behavior and where it actually breaks down rather than a day one snapshot. Key frameworks: Cohort retention curve and LTV:CAC modeling, incrementality testing and full cost payback analysis, framed by the Bain CMO Effectiveness Framework for contribution and full funnel efficiency. Human validation gate: Every cohort read is rerun against your own data and signed off by a named senior auditor before it ships. No model output is ever presented unreviewed. Verified against Seller central settlement Returns & refund logs Ad console exports Finance P&L The buyer payback was measured like for like over a matched period, reconciled to invoiced margin in the P&L, and signed off by a named senior auditor before publication. Client identity redacted to protect their commercial position. WORKING PAPER ████████ B2B wholesale Representative Redacted Buyer cohort Buyer payback Repeat buyers 1.5× Referral buyers 1.3× Subsidized buyers 0.8× buyer payback +26% Durable, the improvement holds every year the fix stays in place, not a one off. Working paper: payback on subsidized buyers traced line by line and reconciled to invoiced margin in the P&L over a matched period. Line items representative and redacted; the recovered figure is the reconciled audit finding. ## What we changed ✓ Reset the lifetime GMV assumption to actual measured repeat behavior, cohort by cohort, rather than an early projection that had been applied across the whole base. ✓ Capped subsidies on the buyer segments that don’t reorder, and concentrated them on the segments with proven, repeatable GMV behind them. ✓ Switched acquisition reporting to genuine repeat GMV payback, so no cohort can be scaled on projected lifetime value alone. ✓ Set guardrails tying subsidy levels to demonstrated buyer retention rather than the optimistic early forecasts that no longer held. ## The result The published chart shows payback on subsidized buyers moving from 0.8× to 1.3×. The separately reported 26% buyer payback improvement is not presented as the percentage change between those two endpoints. Buyer payback improved 26% by subsidizing only the buyers who actually come back. For any B2B marketplace, the trap is the early LTV assumption: applied broadly and never tested, it lets acquisition scale on economics that exist only in the model. If your subsidies rest on projected lifetime GMV rather than measured repeat behavior, you may be buying one off orders at a lifetime price. It takes cohort level repeat analysis to see it. A fixed fee audit does that in days, before you fund more of the gap. From kickoff to signed off findings: 7 days. This +26% buyer payback gain is ≈$66k/yr at the client’s revenue scale. ![Portrait photograph of Alex Pop](/assets/team/alex-pop.jpg) Reviewed & signed off by: Alex Pop Senior Auditor · MarginFix · 10+ years of auditing experience ✓ Anonymized to protect the client · senior reviewed findings · Audit completed between 2024 and 2026, contracted through Cristian Bragau's practice before the MarginFix name existed · Published in the MarginFix library July 21, 2026 · Last reviewed October 28, 2026 [Request an audit →](/book/) [Prefer to talk it through first? Talk to an auditor →](/book/call/) ## More Marketplace findings [Paid search was bidding on supply the platform already ranked for, −$66k/yr](/cases/paid-search-was-bidding-on-supply-the-platform-already-ranke/) [Courier incentives overlapped organic supply in dense zones, −$74k/yr](/cases/courier-incentives-overlapped-organic-supply-in-dense-zones/) [CAC looked fine, until we split it by cohort, +$82k/yr](/cases/cac-looked-fine-until-we-split-it-by-cohort/) [Supply side incentives outlived their purpose, −$63k/yr](/cases/supply-side-incentives-outlived-their-purpose/) [Take rate was leaking through unmanaged refunds, +$44k/yr](/cases/take-rate-was-leaking-through-unmanaged-refunds/) [Promotional fee waivers were never switched off, +$58k/yr](/cases/promotional-fee-waivers-were-never-switched-off/) [The benchmark: what audits recover across 74 cases →](/benchmark/) [All case studies](/case-studies/) [Model a Marketplace budget in the wasted ad spend calculator →](/estimator/marketplace/)