For perps DEXes, trading frontends, launchpads and prediction markets
Attribution for products where the customer is a trader
Your referral and KOL programmes bring volume. What no dashboard tells you is which of them brought traders who keep trading, and which brought farmers who claimed once and left. Same headline signup count, completely different economics.
The problem, stated plainly
A referral programme reports 4,000 new wallets this month at $12 each. Finance signs it off. Two months later the cohort accounts for 0.4% of volume, and nobody can say which referrer was responsible because the only number anyone kept was the signup count.
The failure is not the spend, it is the measurement. Signups are counted at the point where they are cheapest to fake, and quality is assessed — if at all — long after the payouts have cleared.
Measured in your metrics, not DeFi's
Cost per funded trader
Referral spend divided by wallets that actually deposited and traded — not by wallets that connected once.
Volume per acquired trader
Trading volume attributable to each acquisition source. The metric that decides whether a KOL was worth the fee.
Per-referrer revenue
Fees generated by the traders each referrer brought, set against what that referrer was paid.
D7 / D30 trader retention
How many of a cohort are still trading a week and a month later. The number that separates traders from farmers.
Notably absent: TVL. It is the headline number for lending and yield protocols and close to meaningless for a product whose users are trading, not depositing.
Why not just build a Dune dashboard?
A Dune dashboard is genuinely good at what it does: it shows you what happened on-chain. What it cannot show you is why — the campaign, link or referrer that caused a wallet to show up in the first place. That information never touches the chain, so no amount of SQL recovers it.
The practical costs follow from that. Someone has to write and maintain the queries. Every new campaign means new work. And the join you actually need — this wallet came from that KOL — has to be reconstructed by hand from spreadsheets, if it is possible at all.
Attribution records the click and the wallet together, at the moment they happen. On the analytics side the two approaches are complementary; on the acquisition side only one of them answers the question.
Start with the wallets you already paid for
Before changing anything, take the wallets from your last KOL payout and run them through the free wallet quality checker. No account and no integration — paste the addresses and read the low-quality share. If two referrers you paid the same amount come back 40 points apart on average quality, you have found the problem and the size of it in about a minute.
FAQs
Why is this not just another DeFi analytics tool?
DeFi analytics tools measure TVL and protocol revenue. Trading products live or die on a different set of numbers: cost per funded trader, volume per acquired trader and whether a trader is still trading on day 30. Those are acquisition metrics, and almost nothing in the web3 stack reports them per referrer.
How is a funded trader different from a signup?
A signup is a wallet that connected. A funded trader deposited and traded. Referral programmes are almost always paid on the first number and judged on the second, which is why cost per acquisition looks excellent right up until the cohort stops trading.
Do I need to change my contracts or add an SDK?
No. Tracking links carry the campaign, referrer or KOL, and on-chain activity from the connecting wallet is matched back to that link. There is no contract change and no SDK requirement.
Which chains does this cover?
Ethereum, Base, Solana and TON for wallet activity and quality scoring. Base and Solana cover most of the current perps, launchpad and prediction-market activity.
What about KOLs who bring volume from wash trading?
Wallet quality scoring is the check. A referrer whose wallets are new, thinly funded and interact with nothing but your own contracts is a different proposition from one bringing wallets with years of DEX history. You can see that split before the next payout rather than after it.
Can I try it without talking to sales?
Yes. The wallet quality checker runs free with no account, and the product itself has a free tier. There is no demo call gate.
Know which referrers bring traders before the next payout
Free tier, no demo call, no SDK. Tracking links go out today and wallet quality is attributed per referrer as conversions land.