Which lead sources produced commission you actually collected?

A team lead knows precisely what each portal and campaign costs per lead, because everyone selling leads reports that number. What nobody supplies is what those leads closed, after the buyer's agent took their split and after the ones that took nine months are counted.

So the decision gets made on cost per lead and close rate estimates that came from the vendor. Reconciling the closed-transaction export against the lead sources replaces both with the commission actually banked.

In practice this reorders the list more often than not, because the cheapest leads convert slowest and the most expensive ones sometimes convert best.

Response time and follow-up discipline vary enormously across a roster, and portal leads punish both harder than referrals do. One agent turns a source into a good channel and another turns the identical source into wasted spend.

An agent column mapped as Product / service shows each agent's closed commission and how much of it traces to a channel. It does not cross agents with sources, but an agent whose closings sit mostly in Direct / Unknown while the team's paid leads close elsewhere is a routing conversation worth having, and on many teams that is worth as much as the channel ranking itself.

Teams that buy leads at volume almost always employ somebody to work them. An inside sales agent dials the new inquiries, chases the ones that went quiet, books the appointment and hands a warm buyer to whichever agent is up on the rotation. That salary is a real cost of acquisition, and it is entirely absent from the cost per lead the vendor quotes and from the spreadsheet the team lead keeps. A source producing inquiries that take eleven attempts to reach is far more expensive than the same source producing inquiries that answer the first call.

The dial log holds that difference and nothing in the team's reporting ever reads it. It knows which numbers connected, which went to voicemail four times running, which turned out to be wrong or recycled, and which converted to a booked appointment with a named agent. Set beside the closings, it turns a flat cost per lead into something much closer to a real cost per client, which is the figure the team is actually spending against every month. Vendors quote the first number because it is the one that flatters them, and teams repeat it because it is the only one available.

Adding the desk's own contact list to the lead file, tagged as the desk, is the simplest version of that comparison, and it needs nothing more than an export of who was called. Because the earliest record is credited, the desk earns closings from people it reached first, ranked alongside the closings that arrived directly to an agent, on commission rather than on appointments set; its call count sits beside them. The team lead can see how much of the pipeline genuinely starts with somebody sitting at a phone eight hours a day and how much would have arrived regardless. Teams rarely have a considered answer to that before they look, and the answer decides whether the desk grows or shrinks.

A team's lead spend usually buys both sides, and the two behave nothing alike. A homeowner asking what their property is worth is expensive to reach, slow to commit, often twelve months from moving, and produces a listing that generates its own buyer inquiries and frequently a second transaction when the seller buys again. A buyer inquiry is cheap, plentiful, frequently already working with somebody else, and worth nothing at all until it is under contract on a property the team may not have listed. One is inventory and the other is demand.

Because both arrive through the same vendors, land in the same CRM queue and get worked by the same desk, the team's reporting averages them without anybody deciding to. The resulting cost per lead sits somewhere between two very different economics and corresponds to neither, which means every decision taken on it is taken on a number that does not describe anything the team actually buys. Worse, the average moves whenever the mix moves, so a change in vendor allocation reads as a change in performance.

A side column on the closing export, mapped as Product / service, shows the two as separate lines for good, and it costs one field in a report the transaction coordinator already runs. Each side then shows its own commission and how much of it traces to a channel, which goes some way toward showing whether the listing spend is filling the inventory the buyer agents work on, or whether the team is paying premium prices for seller leads that never convert into signed instructions. It does not rank sources within each side, but it is a split many team leads want to see, and one their current reporting cannot produce at any price.

Every established team has closings that came from the database. Somebody inquired in a previous market, went quiet for two years, stayed on the newsletter through a rate cycle and a price correction, opened a property alert on a Tuesday evening and transacted six weeks later. When that closing is logged, the source field gets set to database, or to sphere, or to whatever the agent believes at the time, and the campaign that originally bought that person disappears from the record permanently. Nobody is being careless; there is simply nothing in front of them that remembers.

This is not an edge case in an established team. It is a substantial share of the annual commission and it grows every year the database does, because the list only ever gets longer and the people on it only ever get closer to moving. Attributing all of it to email flatters the newsletter enormously and quietly erases the acquisition spend that filled the list in the first place, which is the spend actually under review at the budget meeting. The channel being cut is the one that bought the closings being celebrated.

Because the match runs on the person rather than on a session or an attribution window, that first inquiry keeps its claim however long ago it was made, provided it appears in the file you upload. Loading three or four years of inquiry history can shift the ranking more through this one effect than through anything else in the exercise, with the oldest, cheapest sources gaining the most ground. It also gives the team a defensible answer to the perennial question of what the database is worth, which until now has been a matter of opinion.