Which channels produced closed-won contracts?

The person who downloaded the report is not the person who signed. Between them are eighteen months, a procurement process and a committee, and no click-based model survives any of it.

Matching on the account's contact records rather than a cookie means any of those touchpoints can be reconciled against the contract when it finally closes, whichever quarter that lands in.

Trade shows and field events are frequently the largest line in an enterprise marketing budget, and are justified with badge counts and stories.

The scan list goes into the lead file, tagged with the event. Ranked on closed-won ACV, events either justify themselves against paid channels or they do not — and either answer is worth having before the next renewal.

An agreement that closes at four hundred thousand dollars is a first payment on a relationship that will run for years and add modules, users, regions and support tiers along the way. The renewal cycle and the expansion motion together usually produce more revenue than the original signature, and they produce it against an account the marketing team was credited for exactly once, in a quarter that is now several fiscal years in the past and nobody reports on. The figure marketing defends itself with is therefore the smallest one that relationship will ever generate.

Because there is no attribution window and the match is on the contact, every subsequent contract line can credit the channel that produced the first conversation. That reframes the question from what a campaign cost per closed deal into what a campaign has earned since, which is the only version a multi-year enterprise business can answer honestly. It is also the only version that stands up when the marketing budget is put next to the revenue it is supposed to have produced.

There is a discipline required here. A renewal signed by a procurement officer who appears in no lead file is not handed to the acquiring channel because it would be convenient. It matches on the contact or it does not, and where it does not, the revenue goes to Direct / Unknown. That is a weaker claim than a lifetime-value model makes and a stronger one than a model can defend, which is precisely the trade this product is built around.

Enterprise buyers do a great deal of work before they identify themselves. They read an analyst evaluation, ask three peers at a user group, sit through a colleague's account of a failed implementation elsewhere, and arrive at a vendor's website already carrying a shortlist. The first interaction anything records is often a demo request or an invitation to tender, by which point the decision that mattered was taken somewhere nobody measured and nobody could have. Treating the demo request as the origin of the deal credits the form rather than whatever filled it in.

No product can attribute a conversation that left no record, and this one does not pretend to. What it can do is rank the activities that did leave a list. Analyst inquiry days, user group attendance, executive roundtables, customer advisory boards, webinar registrations, target account lists: all of them are people with contact details, all of them can go into the lead file tagged with their source, and all of them rank on closed-won contract value beside paid media. Each list you can produce converts a defended assumption into a row with a revenue figure next to it.

What remains after that is the genuine unknown, and its size is itself a finding. A vendor discovering that half of closed-won revenue traces to nothing identifiable is learning that the market talks about it more than it clicks on it. That argues for a different shape of budget than a click-based report implies, and it is a far more useful conclusion than an attributed share that has been inflated until it looks reassuring to whoever commissioned it. The unattributed share is a description of how this market buys, and it deserves to be read rather than removed.

Enterprise software revenue is concentrated. A handful of deals carry the year, and one unusually large contract can make a channel look outstanding on a total while every other deal it produced sat well below quota. Reading a blended channel ranking in that situation is close to reading nothing at all, because the ranking is describing a single contract rather than a pattern anybody could repeat deliberately next year. A single contract is an event; a pattern is what a media plan can be built on, and a blended total cannot tell you which of the two you are looking at.

A segment, region or territory column, mapped as Location or as Product / service, splits the report so each cut shows its own closed-won value and how much of it traces to a channel. It does not rank channels within each cut, but a territory where most revenue sits in Direct / Unknown is one the campaigns are not reaching, whatever the global ranking says. Funding a global rollout on a number one strong territory produced is a mistake this cut makes visible quickly. It also stops a strong territory from quietly subsidizing the appearance of every other one on the same report.

Deal count beside revenue is the other guard. A channel with one large win and a channel with nine consistent ones may sit adjacent in a revenue ranking while representing completely different levels of confidence. Both figures are shown, so anybody reading the report can tell the two apart, which matters a great deal more when the report is going into a board pack than when it is being discussed inside the marketing team. Confidence and magnitude are different things, and a ranking showing only one of them invites the wrong decision.