Which ads produced closings?

A buyer inquires in February and closes in September. By then the cookie is long gone, the ad account has closed seven monthly reports, and the campaign that produced the transaction is showing a cost per lead and nothing else.

Matching on the person rather than the session means the September commission is credited to the February inquiry. For a business where the gap between first contact and closing is routinely six months, this is the difference between measuring marketing and guessing at it.

Portal fees are often the largest single marketing cost in a brokerage, and the reporting comes from the portal itself. Checking it means matching the buyers a portal says it sent against the transactions that actually closed.

That check tends to be uncomfortable in both directions: portals sometimes produce far less than claimed, and sometimes produce the highest-value transactions while looking mediocre on inquiry count. Either way it is a number worth having before the renewal conversation.

Every CRM in this industry has a lead source dropdown, and in most offices it is filled in by whoever happened to pick up the inquiry. Zillow, website, sign call, sphere, other. The entry is a guess made in fifteen seconds by somebody whose job is to book an appointment rather than to keep records for the marketing budget, and it is never revisited when the buyer turns out to have first seen the property somewhere else entirely. Nobody audits it because nothing downstream depends on it, which is precisely why it drifts.

The commission figure, by contrast, is trustworthy. It sits in the transaction management system, entered by a coordinator at closing, checked against the settlement statement and reconciled against what the office actually banked. Two records of the same deal therefore exist at different levels of quality on opposite sides of a system boundary that nobody owns, and no report in the building joins them. The marketing conversation runs on the unreliable one because it is the only one that mentions a channel at all.

Exporting both files and matching on the client's phone number or email closes that gap without anybody retyping anything, and without changing how intake works. The dropdown stops being the office's official account of where business comes from and becomes one more input to check against the closings that actually happened. Where the two disagree, the closing wins. The disagreement itself is usually the most interesting thing in a brokerage's first report, because it tends to show that the sources everybody names in the weekly meeting are not the sources carrying the year.

A valuation request and a request to view a property are one lead each in every report an office receives, and they have almost nothing in common. A listing is inventory. It carries its own signage, it generates viewing inquiries at no extra cost, it produces a second commission when the seller buys again, and it is far scarcer than buyer demand will ever be. A viewing request is abundant, usually early, frequently unrepresented, and converts on somebody else's listing as often as on your own, months after the inquiry was logged and forgotten.

The channels behind the two sides diverge accordingly. Searches about what a property is worth, about a street's recent sales, or about selling without an agent draw a different audience from searches naming a development, a school catchment or a price band, and the cost of winning a listing appointment typically runs several times the cost of a viewing inquiry. Averaged together they produce a blended cost per lead that describes neither half of the business, cannot guide a decision about either, and quietly shifts as the mix between them moves from one month to the next.

Adding a side column to the closing export, mapped as Product / service, splits the revenue cleanly: listing-side and buying-side commission each get their own line, with how much of each traces to a channel. That alone can show that one side of the business is well explained by the advertising and the other barely at all, and that the office has been funding both as though the spend were doing the same job at the same price. The channel ranking itself still covers both sides together. Knowing which half the advertising actually explains is what lets a principal move money on evidence rather than on the loudest recent anecdote, or on whichever portal representative last came in with a renewal quote.

Offices count transactions because transactions are what the leaderboard shows, what the franchise reports, what agents talk about in the parking lot and what the annual awards are given for. A source producing eleven closings beats one producing five in every conversation that happens on the sales floor, and the budget quietly follows that conversation from one quarter to the next. It is an entirely reasonable habit, it requires no extra work from anybody, and it is wrong often enough to be genuinely expensive over a full year of spending.

Price point is the reason it goes wrong. Five closings in the upper part of the market can carry more commission than eleven at the entry level, and the share the office retains differs again, because splits and caps apply per agent rather than per transaction. A channel reaching first-time buyers therefore accumulates rows quickly while contributing modestly to what the brokerage keeps, and it will win any comparison built on counting. The agent working those leads may also be the one furthest from their cap, which widens the gap further.

A first report makes the divergence visible almost immediately, because the channel table shows the money and the number of deals behind it side by side rather than one without the other. The ordering by revenue and the ordering by row count are frequently different lists, sometimes reversed outright, and the gap between them is the clearest single argument for running the reconciliation at all rather than continuing to manage a marketing budget on lead volume and the impression it leaves.