Which channels produced fee-paying clients?

A small-business client engaged this January will file again next January, and probably add bookkeeping and advisory along the way. Judging the campaign that won them on the first invoice alone is a systematic undercount.

Because the match is on the client rather than a session, every later fee credits the channel that produced them. Measured that way, the channel with the highest cost per lead can turn out to have the lowest cost per retained client.

A $400 return and a $30,000 advisory engagement arrive from different places and are usually reported as one lead count.

A service-type column mapped as Product / service splits them, showing fees by service line with how much of each traces to a channel, so the firm can see whether marketing is reaching the practice it wants to grow or only the one that fills January.

A tax practice takes on most of its new clients inside a narrow window, and the conversations that produced them started months earlier. A small business owner who decides in September that the bookkeeping has become a mess engages in January. Somebody who fell out with their previous accountant in May comes on board in time for an extended return in August. The inquiry and the fee sit in different quarters almost by definition, and in a monthly report they sit in different stories altogether, with nothing connecting them.

Comparing consecutive months in a practice with that shape measures the calendar rather than the marketing. The report is built to compare a period against the same period a year earlier, which is the only comparison that means much here, and because the match on the client carries no window, an October inquiry that becomes a January engagement is a single match rather than two disconnected records in two systems. The gap between them costs nothing and needs no explanation in the file.

It also answers the question firms ask every autumn, which is whether the off-season advertising is doing anything at all. Ranked on fees collected in the following season rather than on inquiries received at the time, the quiet months frequently look considerably better than a monthly report suggests. That is the argument for keeping a budget running when the phones are slow, and it is an argument most practices have never been able to make with a number attached to it.

Realization is the number that governs whether an engagement was worth taking at all. Fees get written down, scope gets absorbed, a client argues about an invoice or pays it eleven months late, and a practice's billed figure and its collected figure end up describing two different businesses. Attribution built on billed amounts will rank a channel highly for work that was never fully paid for, which is a particularly unhelpful error when the report is being used to decide next year's spending.

Exporting collected fees instead, dated when the money arrived, produces a ranking on what the firm actually received. This can move the referral-led sources upward and price-led search traffic downward, because a client who came through a professional introduction tends to accept a scope and pay for it, while a client who arrived comparing prices tends to argue about both. The difference is rarely small and it is almost never visible in the systems a firm already runs. Most practice management systems will produce a receipts report as readily as a billing one.

Write-offs and disputes do not have to be categorized for any of this to work. The revenue file contains what was collected, so unrecovered work is simply absent from it. That keeps the report free of judgment calls about whether a particular write-down was the client's fault or the firm's, and it anchors the ranking to the only figure the partners reliably agree on, which is the money that reached the bank account in a given month. The partners never have to agree on whose fault a write-down was for the ranking to hold.

Most practices are built on a handful of professional relationships: a commercial banker who introduces borrowers needing accounts prepared, an attorney who sends business owners through at a transaction, a bookkeeper who passes the compliance work upward, a software partner directory producing a steady trickle of inquiries. Each of those is described internally as word of mouth, each is maintained with real partner time, and none of them is measured against anything the firm spends money on. The result is a pipeline the firm depends on and cannot describe in numbers to anybody.

A list naming each referrer alongside the clients they introduced can go into the lead file, with each client tagged by referrer. Each referrer then ranks on collected fees, which converts an impression into an amount. The finding may be uneven in a way that is immediately useful: one or two relationships producing most of the referred value, and several others producing almost nothing while consuming a similar amount of lunches, updates and reciprocal attention across the year. Partner time is the scarcest thing a practice spends, and this is the only report that prices it.

What the report will not do is invent the connection. Where nobody recorded who introduced a client, the fees stay in Direct / Unknown rather than being distributed across the referrers on a guess. Capturing that detail after a first report can be the highest-return change a firm makes to its marketing data, and it amounts to one extra field on an intake form and a habit of filling it in. One field, filled in consistently, is worth more here than any amount of additional advertising data.