Which channels produced enrolled members?

A member who stays five years is worth $12,000. A channel judged on the first month's payment appears to cost more to acquire than it earns, and gets cut.

Because the match is on the member, every renewal credits the original channel. That is the difference between a marketing budget that can be defended and one that cannot.

Members arrive through existing patients, physicians and employers, and practices rarely have any measurement of which of those actually produces enrollments.

Referred members can go into the lead file, tagged with the referrer. Ranked on membership revenue, the referral sources may show that one or two relationships produce most of the growth — a fact worth knowing before renewing an advertising contract.

The whole proposition rests on the panel being small. Six hundred members instead of two and a half thousand patients is what buys the same-day appointment and the unhurried consultation, and a practice that grows past its cap has sold something it can no longer deliver. So marketing in this model is not a growth engine in the usual sense. Once the panel is full, its only job is to replace the members who leave, which is a materially different brief from the one most agencies are given.

That changes which number matters. Cost per inquiry is close to irrelevant when the practice needs thirty new members a year rather than three hundred; what matters is whether a source produces members who stay, because a member recruited and lost within a year has cost a seat and returned one annual fee. A channel producing twelve inquiries and eight long-tenured members is worth more than one producing two hundred inquiries and a churn problem. That comparison is not close, and nothing in a monthly acquisition report will ever make it, because both channels look identical in the month the member signed.

Measuring it requires nothing beyond exporting membership payments with a contact detail and running the report across several years. Every renewal joins to the same person and therefore to the source that recruited them, so a channel whose members stay keeps accumulating revenue while one whose members leave stops. Sources may turn out to be sharply differentiated on what they earn over time and barely differentiated on acquisition cost. It also means the honest reporting period here is years rather than quarters, which suits a model whose whole argument is about continuity.

One inquiry frequently becomes three memberships. A spouse joins at the same time, an adult child is added, sometimes a parent follows a few months later, and the practice books each of them as a separate member with their own annual fee. The inquiry file holds one name and one phone number. The revenue file holds three names, and possibly three different contact details, only one of which the advertising ever touched. On any acquisition report the campaign that produced the household is therefore credited with a third of what it actually delivered.

Left alone, that produces a systematic undercount. The source is credited with one membership and the other two fall into Direct / Unknown, which is honest but unhelpful — it makes every acquisition cost in the practice look roughly three times worse than it is for exactly the channels that recruit families. Since family enrollment is one of the strongest signals a practice has that its proposition is landing, penalizing it in the report is a particularly unfortunate way to be wrong.

The fix is in the contact detail, not an identifier. Each additional member is a separate row, and it is credited to the household's original inquiry only if it carries a contact detail from that inquiry — putting the paying member's email on every family member's row does exactly that, while the phone field can still carry the member's own number. A household or account identifier is not a matching key. Without a shared detail, the extra memberships appear as unmatched revenue rather than being guessed at, because inferring a family relationship from a shared surname and an address is not something this product will do. That is a deliberate limit rather than a gap: a contact detail is a fact the practice holds, and a shared surname is a guess.

Members recruited in one month renew in that month for as long as they stay. After three or four years a practice's billing calendar has hardened into a pattern set by whichever campaigns happened to be running in previous Januaries, and a monthly revenue report is largely a record of past recruitment rather than current performance. A channel that has not delivered a new member in two years can still dominate a given month's figures and look like the practice's best source.

This is not a small distortion. In a subscription model with high retention, new business is a thin layer on top of a thick base, and any report that does not separate the two will be read as though the base were a result. Practices make real decisions on this, usually to keep funding whatever the report says is producing revenue, which is the source that produced it several years ago. The base is also the most reassuring number in the practice, which is part of why nobody looks closely at what it is made of.

Separating new enrollments from renewals is a column in the export, mapped as Product / service, and it changes what the page is for. The report then shows new-enrollment and renewal revenue as separate lines, with how much of each traces to a channel, so both are visible instead of one blended figure. It does not rank channels within each line. Running the comparison annually rather than monthly helps as well, because an annual view is not hostage to which members happen to renew in a given two-week period. The new-business layer may turn out thinner than expected, and it is better to find that out while the panel is still full rather than a year after it stopped being.