Which channels produced merchants who transact?

Fintech acquisition is measured on signups because signups are instant and easy to attribute. Activation — the first real transaction — happens days or weeks later, and processed volume builds for months after that.

Ranking channels by processed revenue rather than signup count routinely inverts the picture. The cheapest signup source is frequently the one that activates least, and a channel with a third of the signups can produce most of the volume.

Your product analytics measures what happens inside the product, and does it well. It cannot tell you which advertising channel produced the merchant now processing six figures a month, because that connection was made before the account existed.

CloseRev joins the acquiring channel to the revenue the account went on to generate. The two views sit side by side and answer different questions; they are not alternatives and the numbers should not be summed.

A payments portfolio is a small number of large accounts and a very long tail of small ones. A restaurant taking twenty thousand a month and a marketplace taking four million are both merchants, both counted as one activation, and both produced by a channel the report will rank on activation count unless somebody insists otherwise. Average processing volume per account is a figure that describes almost nobody in the book it was calculated from. Ranking on it produces a league table of which channel found the most small businesses fastest.

Ranking on revenue rather than on accounts corrects that. The report's deal-size breakdown, or a merchant-category column mapped as Product / service, then shows how revenue is spread across small and large merchants and how much of each band traces to a channel. Beside it, each channel's revenue and account count in the ranking show the pattern that matters: a cheap acquisition channel can produce the smallest merchants, at a cost per account that looks excellent and a cost per dollar of processed volume that is the worst in the portfolio. Both of those things are true at once, and only the first one is currently on the dashboard. Which of the two you read determines whether that channel gets more budget or less next quarter.

Size mix also explains most of the internal disagreements about which channel is best. Sales looks at the accounts they were handed, marketing looks at the volume of signups, and finance looks at revenue recognized. When the same two files rank every channel on processed revenue with the account count beside it, those three views collapse into one table that everybody can read the same way without anyone having to concede a point first. One table, two files, and an argument that had been running on three incompatible definitions of success.

Merchants rarely announce that they are leaving. Volume tapers over a few months, a second processor appears in the mix, a seasonal business simply does not come back in the spring. By the time an account is formally closed it stopped producing revenue long before, and the channel that acquired it has been carrying credit for a customer that had effectively gone two quarters earlier without anybody noticing. The acquisition cost was paid in full and the revenue stopped without any system raising its hand, which is why attrition here is usually discovered late and by accident.

Because the revenue file is exported by account and period, this is visible without any churn definition being agreed first. A channel whose merchants keep processing accumulates revenue month after month; one whose merchants fade stops accumulating. Run the report across two or three years and the two separate cleanly, which is a far better measure of acquisition quality than an activation rate taken inside the first thirty days. Nothing has to be classified, flagged or agreed between teams for the difference to appear in the totals, because the absence of revenue is itself the evidence.

It changes what is worth paying for. A channel with a high activation rate and rapid attrition can be profitable across a quarter and unprofitable across a year, and a channel that activates slowly but produces merchants who stay can be exactly the opposite. Neither conclusion is available from a signup report, and both fall out of the same two exports with no additional instrumentation and no change to the onboarding flow. It also gives the partnerships and paid teams a shared basis for a conversation they currently have separately.

A large share of payments volume now arrives embedded in somebody else's software: a booking platform, a practice management system, a marketplace, a vertical application sold to one trade. Those merchants never saw an advertisement. They were onboarded inside a product they already used, and the partnership that put you there cost integration engineering, revenue share and a partnerships team that is funded from somewhere other than the marketing budget. The result is that a substantial part of the company's distribution never appears in a marketing report at all, and is therefore never compared with anything.

Adding the partner's merchant list to the lead file, each row tagged with the partner, puts that motion into the same ranking as paid acquisition. Each named platform ranks separately on processed revenue, so a partnership producing a thousand tiny merchants and one producing forty substantial ones stop being summarized as a single row labeled partners. That comparison is usually the one that decides where the next integration effort goes, and it is currently made on instinct. Two partnerships summarized as one row is a common way a platform business misreads its own distribution, and it is the easiest thing on this list to fix.

The same applies to referral programs, agent networks and bookkeeper introductions, all of which are lists of people with contact details. What none of it requires is a change to your onboarding flow or a tag inside the merchant dashboard. The files are exported after the fact, which also means a partnership signed two years ago can be evaluated on everything it has produced since rather than on how its first quarter happened to look. A first quarter is a poor guide to a partnership, and it is usually the only evidence anybody has kept.