Don't measure the wrong crowd, very accurately.

Almost everything written about web3 marketing, plenty of it by us, is about measurement. Did the campaign work, and can you prove it to whoever decides on the budget. It's good work and it matters, it's why we've spent years on it. But all of it happens after the money's already gone out, and it ignores the setting that decides where the money goes in the first place.

That setting is targeting. Who the ad gets shown to. It's where it all begins and decides whether a campaign even has a chance to start with. Everyone wants better targeting. Almost nobody has a reliable way to actually do it, and we don't see many write about it, us included, so that's what we're getting into.

I don't think targeting is hidden so much as hard. Doing it properly in crypto is genuinely difficult, which is why what passes for targeting is so often a guess.

For measurement you're staring at a dashboard with data and that does feel legit. But targeting is really the choice you make at the beginning and then it sets up everything the dashboard shows you later. You can run flawless attribution and a clean incrementality report but still be spending the budget on people who were never going to matter to you, or you to them.

We're on a lot of advertiser calls, and you can tell the ones who've been doing this for a while, because they start by asking the targeting question straight out. Will the wallets that actually move money see this.

It's exactly the right question, just make sure you're asking it in the right place. Google and Meta can't see wallets or onchain activity at all. The best they can do for a crypto audience is guess from what someone searched or the sites they visited, something like "this person seems into crypto."

And that's a guess at best.

That guess definitely can't tell whether the person moves real money through a product like yours, or farmed one airdrop back in 2021 and hasn't done a thing since.

So it serves your ad to both and leaves you to sort them out afterwards in the report. No amount of measuring fixes that because you just measured the wrong crowd very accurately.

Targeting on what a wallet has actually done is a different game. You pick who sees the ad by behaviour: the protocols they've used, the chains they're active on, and whether they've touched a competitor.

For us that runs across more than 6,000 protocols and 15-plus chains, which is really just a way of saying the behaviour you can target on is wide enough to get specific. And it shows in who converts.

On a recent perp-DEX campaign, conversions averaged over $25k in volume each, which is a number that tells you the ad reached the right people, who move money.

So before you tune another report, ask who your ad is even reaching.

If the audience was picked by a platform that can't see a single wallet, then no matter how good the report, it's always going to be tainted, because it was aimed at the wrong people.

Sort out where the ad points first, then measure.