Which channels produced signed build contracts?

A custom build starts with somebody imagining a house. Between that first inquiry and a signed contract they usually buy land, engage an architect, iterate on a design, get planning approval and arrange construction financing. Eighteen months is normal and two years is common.

No advertising platform reports on that horizon. The campaign that produced the inquiry is long since paused, the account may have been restructured twice, and the builder is left crediting whichever touch happened last.

Reconciling signed contracts against an inquiry history removes the horizon problem entirely. Either the client who signed in September appears in the inquiry file from two Februarys ago, or they do not.

Custom builders receive a great many inquiries from people gathering ideas, testing budgets or planning several years out. A meaningful share have no land and no financing, and the builder's own qualification process is what separates them.

Ranking channels on inquiry volume therefore rewards the channels best at generating inspiration. Ranking on contracted build value — seven figures apiece — produces a shorter and much more useful list, and it is usually not the same list.

Architects, designers and past clients produce a large share of custom work, and those relationships cost time rather than media spend. They are normally left out of marketing reporting altogether.

Adding the clients they introduced to the lead file, with each architect as the source, puts them on the same axis as paid search, so the builder can see whether an architect relationship is worth more than the whole digital budget. It may well be, and that is a finding worth having in writing.

A custom build needs a site, and a substantial share of inquiries arrive without one. The family has a budget, a rough idea of the area and a folder of images, but no lot, and until they have one there is nothing to price and nothing to draw. Finding it can take a year, involve an agent who has no particular interest in the builder, and end with a plot whose slope, services or planning position rules out the house that started the conversation.

Clients who already own their land are therefore a different population altogether. They convert faster, more predictably, and with far less attrition along the way, because the largest unknown has already been resolved before the first meeting. So are clients who come through the builder's own lot inventory, or through a developer relationship where the site is secured and serviced. Those three groups convert at completely different rates and cost completely different amounts to acquire, and every inquiry report a builder receives shows them as a single undifferentiated number.

A land-status column on the contract export, mapped as Product / service, shows landowners and land-seekers as separate lines, with how much of each traces to a channel. It does not rank sources within each, but read beside each channel's inquiry count and contract value it can be more informative than the channel ranking alone. A source producing landowners at three times the cost per inquiry of one producing land-seekers may well be the cheaper source per signed contract, once the attrition and the eighteen months of waiting are accounted for. No builder can know that without matching the signed contracts back to where the inquiries originally came from, which is the one thing nobody has been doing.

Custom contracts move, and everyone in the business knows they move. Allowances are set for kitchens, joinery, tiling, sanitaryware and landscaping long before anybody has chosen anything, and the choices actually made nine months later frequently exceed them. Change orders accumulate steadily through the build as the client stands in the space and revises their thinking about a wall, a window or a staircase. The house contracted at one figure is invoiced at a meaningfully different one, and the direction of travel is very rarely downward.

How far it moves varies by client rather than by house, and clients vary systematically by where they came from. A family introduced by an architect who had already set honest expectations about cost behaves quite differently from one who responded to an advertisement quoting a price per square meter. Ranking channels on signed contract value therefore understates some sources and overstates others, consistently, in the same direction every time, and in a way that never corrects itself no matter how many years of data are added.

Because draws and variations are invoiced against the same client in the same system, they can be exported as additional rows and credited to the same original channel without any manual work at all. The report then ranks on what each client actually paid across the whole build rather than on what they agreed to at the outset, before anybody had chosen a tap. That is both a more honest number and the one the builder's own accounts were keeping all along, which is what makes it easy to defend in a management meeting.