Which channels produced signed contracts?

Buyers register at a show home, join a mailing list, watch a community fill, and sign a year later. Every platform has stopped claiming that click long before the contract exists.

Matching on the buyer removes the window entirely. In this category it routinely moves a seven-figure share of contract value out of Direct / Unknown and onto the campaigns that actually earned it.

A channel that fills a starter community can be almost useless for a move-up product two suburbs away, yet builders usually run one blended report across all of them.

A community column mapped as Location splits the revenue by community, with how much of each traces to a channel, so a release is not judged on a portfolio average that describes nowhere. It does not rank channels within each community.

Production builders convert on site. A family drives out on a Sunday, walks three model homes, talks to an on-site sales consultant for forty minutes and leaves a name on a registration card. That card is the single most important record in the entire funnel, and in a great many builders it is a paper form transcribed into a spreadsheet on Monday by somebody who was not there. The how-did-you-hear-about-us box on it is answered casually, because the visitor is standing in a kitchen they are imagining living in rather than filling in a research form.

Everything upstream of that visit is digital and measurable: the portal listing, the community page, the retargeting that followed them around for two weeks, the email announcing a new release, the video walkthrough they watched twice. Everything downstream is a contract in a different system, handled by different people, reported on a different cadence. The registration card sits between them as the only place the two halves of the funnel ever touch, and it is by a wide margin the weakest link in the chain. Builders spend heavily on both ends and almost nothing on the join.

Typed up as a spreadsheet with a name, a phone number and a visit date, that card becomes rows in the lead file like any other source, and it does not matter that it was filled in badly or answered vaguely. The contracts that eventually signed are matched back to it on the visitor's own contact details rather than on what they wrote in the box, so the walk-in traffic can finally be ranked against the campaigns that produced it instead of being treated as demand that arrived from nowhere and belongs to nobody.

Most production builders pay a co-operating commission to real estate agents who register a buyer and accompany them through the sales center. It is one of the very few acquisition costs in the business that is precisely known at the moment of sale, expressed as a percentage of the contract, and paid only on a completed settlement rather than in advance. Compared with a digital budget committed months ahead against uncertain results, it looks like the safer purchase, and builders lean on it for exactly that reason. The lean is rarely tested against anything.

What has never been available is the comparison itself. The co-op commission paid across a year is a large number sitting in the finance system, the advertising spend is a different large number sitting in an ad account, and no report has ever put a cost per settled home against both of them at once. Sales managers and marketing managers argue about where the next dollar goes on instinct and on whichever side produced the most recent memorable sale, which is not a basis for allocating six or seven figures.

Adding agent registrations to the lead file — each registered buyer, with a contact detail, tagged with the agent as the source — puts the agent network on the same axis as everything else in the mix. Settlements introduced by agents are ranked on contract value beside settlements produced directly, in one table, over whatever period the builder chooses. The builder can then see what its own demand generation costs per home against what the agent network charges for the same outcome. Both numbers may surprise somebody in the room.

A signed contract is not money. Buyers fail to sell their existing home, financing falls through on a re-qualification weeks before settlement, a rate move changes the payment beyond what a family will accept, or somebody simply changes their mind during a build that runs the better part of a year. Cancellation rates in this business are high enough that ranking channels on contracts written rather than homes settled is a materially different exercise, and the difference is not distributed evenly across sources. Some channels produce buyers who complete and some produce buyers who intend to.

The other half of the revenue arrives later still. Structural options are chosen before the slab goes down, design studio selections are made weeks after that, and a family that spends heavily on upgrades is worth considerably more to the builder than the base price of the plan suggests. Those selections are invoiced against the same buyer in the same system, so they credit the same original inquiry without anybody having to connect them by hand or remember which campaign brought that family to the community in the first place.

Export settlements at total contract value, options included, and keep contracts that were written but never settled out of the sales file, where every row counts as revenue. That produces a ranking the finance director recognizes rather than one only marketing can explain. With written contracts in the lead file, each channel's written contracts sit beside its settlements, and a channel that writes a great many and settles few is visibly producing buyers who cannot complete — a qualification problem rather than a demand problem, and the two call for opposite responses.