Which campaigns produced placements you actually got paid for?

A placement agency's whole economics are a conversion business: free advice to families, a fee from the community when a family moves in, commonly around a month's rent. Roughly one enquiry in ten or twenty ends in a payable placement.

That ratio is exactly why enquiry-level reporting misleads. A channel producing plenty of families who tour and never move is expensive and looks cheap; a channel producing fewer, better-qualified families looks the opposite.

Matching invoiced placements back to the original enquiry reverses the ranking often enough to be worth doing once, whatever you conclude afterwards.

You refer a family in March. They tour four communities, choose in May, move in June, and the community's accounts pay you in July. Nothing in that chain is under your control and all of it sits between your advertising and your revenue.

Because the match is on the family's contact detail rather than a session identifier, a four-month lag is not a problem to be modelled. It is simply two rows in two files that name the same person.

An agency earns nothing from a family it cannot place into a contracted community. Enquiries arrive for towns outside the network, for care levels nobody in the network provides, and for budgets that no contracted community will accept. The adviser still takes the call, still spends forty minutes on it, and still refers the family somewhere — because that is the job — and the agency invoices nobody. The call still costs the agency an hour of skilled time, and the accounting system will never record it as anything at all.

Media buying rarely reflects this. Campaigns are bought against search terms and geographies that look like demand, and a source producing a steady flow of families in a county where the agency holds two agreements will look identical, on any lead report, to one producing families where it holds thirty. The cost per enquiry is the same and the revenue is nothing like it. It is one of the few places in media buying where two identical-looking line items differ by their entire value, and nothing in the platform can tell them apart.

Ranking sources on fees actually collected makes the coverage map visible in the marketing numbers for the first time. It also tends to reframe the network question: a geography producing a large volume of unmonetised enquiries is either a market to sign communities in or a market to stop advertising in, and that is a decision worth making deliberately rather than by default. Signing three more communities in one county can be worth more than doubling the budget there, and the report is what tells you which of those two situations you are actually in.

Placement agreements commonly carry a guarantee period. If the resident moves out or dies within a stated window, part or all of the fee is returned or credited against the next placement. Those adjustments land weeks after the invoice, in the accounting system, and they are never reflected in whatever the CRM says about the placement. By then the placement is closed, the adviser has moved on to other families, and nobody goes back to amend the original record. The pipeline report stays permanently a little more optimistic than the bank account.

That matters because the adjustments are not evenly distributed across sources. A channel producing families in crisis, placing quickly into the first available bed, will see more early move-outs than one producing families who researched for three months and chose carefully. On invoiced fees the two look similar. On collected fees, after clawbacks, they may not. The difference is structural rather than random: it follows from how the family was recruited and how quickly they were moved, which is precisely what a channel determines. That makes it a channel-level correction and not noise.

Exporting collected rather than invoiced amounts — or exporting both, with a status column — puts the real number in front of the person buying the media. It is the sort of correction that only ever moves in one direction, and agencies that make it usually find one source has been carrying a quiet discount against it for years. Most agencies find it in the direction they expected and are surprised by the size of it. A ten per cent clawback rate against one source and two per cent against another is not a rounding difference.

An agency's capacity is a number of advisers multiplied by the families each can genuinely support at once. Buying more enquiries than that capacity can absorb does not produce more placements; it produces slower responses, longer gaps between calls, and families who have already toured with somebody else by the time anybody rings back. Volume past the constraint is not merely wasted, it damages the conversion of everything else. The families who waited are not neutral about it either, and they tend to say so in the reviews the following month's enquiries will read.

This is invisible in any report measured in leads. Cost per enquiry falls as volume rises, so the dashboard improves at precisely the point the business gets worse, and the decision to buy more looks well supported right up until the placement count fails to follow. Nothing in the media account knows how many families an adviser can hold at once, so nothing in the media account can warn you. The warning arrives later, as a placement count that stayed flat while spend rose, and by then a quarter has gone.

Ranking sources on collected fees per enquiry rather than on enquiries delivered puts the constraint back in the picture. A source that occupies four hours of adviser time per placement and one that occupies forty can cost the same per lead, and only one of them can be scaled. An adviser column, where the export carries it, makes that comparison specific rather than theoretical. Capacity is what an agency is really buying when it buys media, and pricing it correctly is the difference between growth and a busier team producing the same number of placements.