HypeLab vs Specify: two takes on wallet-native advertising

If you are comparing HypeLab and Specify, the first thing worth saying is that you have narrowed the field well. These are the two serious networks built around the same insight: that the best place to reach a crypto user is inside the products they already open, not on a banner beside an article. Most of the crypto ad market is still selling display slots on media sites. HypeLab and Specify are both somewhere else entirely, and they are more alike than either is like the rest of the category. This piece is about the one place they genuinely diverge, and how to tell which bet fits your team.

Disclosure: this is Specify’s blog, so read us with that in mind. We have kept every claim about HypeLab to what is in their published materials, and we mean it when we say they are a good network. This is two peers comparing notes, not a takedown. The usual caveat applies to both of us: a network’s claims about itself are claims, so wherever you can, prefer the numbers you can verify. Specify’s resolve to onchain transactions; that is the habit we would recommend bringing to every vendor conversation, including this one.

What the two networks share

Start with the overlap, because it is large. Both HypeLab and Specify were built after 2021 around native, in-product placements: ad units that live inside wallets, explorers, and dapps rather than as banners on crypto news sites. Both target on wallet-aware audiences rather than the crude contextual guess of “this person is reading a crypto article, so they must be a crypto user.” If your reason for shopping is that you want your message in front of people while they are actually using onchain products, either network will do that for you, and both do it well.

HypeLab has built real scale on this model. Their published materials describe more than 1 billion monthly ad requests, 200+ publishing partners, and 20M+ wallets reached, from a team founded in 2022 out of Oakland. Specify runs native placements across Coin98, Blockscout and 40+ block explorers, Turtle, Collab.Land, and Outposts, with audiences built from observed onchain behavior across 6,000+ protocols and 15+ chains. The surfaces differ in the particulars, but the philosophy is the same one: meet the user in the product, address them by what their wallet has actually done.

The honest fork: how you buy, and how you verify

Where the two networks part ways is billing, and it follows straight from two different answers to a single question: what should an advertiser be buying? Both answers are coherent. They just point at different kinds of certainty.

HypeLab bills the way most performance media does. Their materials describe transparent CPM pricing, so you pay per thousand impressions, and their reporting has been moving toward outcome metrics: they promote cost per wallet and onchain ROAS as the numbers to judge a campaign by, which is genuinely ahead of most of the market. The model has real virtues. Pacing is predictable, you can plan a budget against a known unit cost, and it scales cleanly, because impression inventory is abundant in a way that verified conversions never quite are. If your team wants a strong in-product media buy with outcome reporting on top, this is exactly that instrument.

Specify bills on the other side of the funnel. You pay a cost per conversion verified onchain, and every conversion carries a transaction hash you can audit for yourself on a block explorer. Billing is tiered on value, so a larger conversion costs more than a small one. Attribution runs on a 14-day window, campaigns are measured against incrementality holdouts rather than last-click, and converted wallets are tracked for lifetime volume so you see what a cohort did after its first action. The trade-off is the mirror image of HypeLab’s: pacing is less predictable and volume is bounded by how many real users actually match and convert, but you never pay for an impression that did nothing, and the results are verifiable without trusting anyone’s dashboard.

Neither of these is the smarter choice in the abstract. Impression billing buys you predictable scale, with outcomes arriving as reported metrics on top of the media. Conversion billing makes the outcome the billable unit itself, and hands you a ceiling set by the size of the real matching audience. They are different bets on what a buyer wants to walk away holding.

The CoW overlap, out in the open

There is a client both networks can point to, and it is worth addressing directly rather than pretending it away. HypeLab publicly cites CoW Protocol as a case study, reporting 4x lower CAC over five-plus months. Specify’s CoW Swap campaign produced 1,500+ verified conversions, more than $50M in attributed volume, and 6x lift against a holdout of identical users who never saw the ads. Both are real results reported by each network for the same protocol.

The honest reading of that is not a scoreboard. It is that serious protocols run several channels at once and measure each one, because no single network is the whole answer for a team of that size. CoW got genuine value from both, which is the point. What the two case studies actually illustrate is the difference we have been describing: HypeLab reports a CAC improvement through its platform, and Specify reports a lift number that only exists because an ad was withheld from a comparison group and the transaction hashes were counted afterward. Same client, two ways of knowing what you got.

How to choose

Here is the even-handed version. If you want scale first, judged on the outcome metrics the platform reports, HypeLab is a strong, well-established place to buy it. If you want to pay for outcomes and have those outcomes verified independently, down to the transaction hash and the holdout, that is the bet Specify is built for. And if you are genuinely unsure, the sensible move is not to argue it on paper but to run both and compare what comes back. Specify’s way into that is a $1,000 done-for-you test campaign: a calibration flight run for you, ending in a report you keep either way. Set it beside a comparable HypeLab flight, ask both the same closing question, and let the two reports settle it. Plenty of teams should simply do exactly that, and many will end up with room for both.