Medical group and MSO marketing for multi-specialty organisations

One P&L, several specialties. A blended CAC is a number that lies — averages hide the winning service lines and protect the losing ones, so capital allocates to noise.

CC — PRESENTING COMPLAINT

Referrals are being placed. Nobody can say whether the patient was seen.

A multi-specialty group's largest source of demand is itself. A primary care physician refers into cardiology; cardiology refers into imaging; the specialist refers back. On paper this is the whole argument for owning several specialties under one organisation.

In practice, almost no group can say what share of those referrals resulted in a completed visit inside the organisation. The referral is placed in the record and then it stops being anybody's number. Some patients are seen. Some go outside because the internal wait was three weeks and an outside practice could see them Tuesday. Some are never contacted at all.

That leak — a referral loop nobody has closed — is usually larger than anything available in the ad account, and it costs nothing in media to close: it is scheduling, access and follow-up, instrumented and reported. We measure the internal referral as a funnel with named steps: placed, contacted, scheduled, seen. Groups that start reading that number generally stop asking about cost per lead within a quarter.

SUBJECTIVE

What we hear from MSO and medical group leadership

"Every specialty has its own agency and none of the numbers reconcile." · "Derm subsidizes primary care and nobody can prove it." · "The board deck takes two weeks to assemble."

DX — PRODUCTION AGAINST REVENUE

Work RVUs per FTE are up; revenue per FTE is not

This is the finance-side version of the same complaint, and it is the one a CFO opens with. Physicians are working at least as hard as last year by any production measure, and the revenue those physicians generate has not moved with it.

When those two lines diverge, the cause is a mix, a rate or a leak, and marketing is only implicated in one of the three. Mix: the growth arrived in the service lines with the lowest contribution. Rate: the payer contracts moved and nothing about volume will fix it. Leak: the work is being produced and not fully collected.

The reason this belongs on a marketing page is that a group without service-line-level instrumentation cannot tell the three apart, and will spend a year buying volume against a rate problem. So we instrument by service line before recommending anything: demand, capture, and cost per new encounter, per specialty, per site — and if the finding is that this is not a marketing problem, that is the finding.

DX — DEMAND TYPES

The specialties inside a multi-specialty group

A multi-specialty organisation is not nine marketing problems. It is one instrumentation problem and a small number of genuinely distinct demand types, and the useful split is by where the patient comes from rather than by specialty name.

Some lines are panel- and payer-driven — primary care and internal medicine are the clearest examples. Growth there is a function of contracts, panel size and access, and consumer advertising moves it very little. Some are referral-driven, including most of the procedural specialties: the work is won from the referring physician, not from the patient, and the instrument is the referral, not the campaign. And some are genuinely patient-chosen and shopped, which is where direct acquisition by service line earns its place.

We sort a group's book into those three before proposing anything, because the budget question is not how much to spend but which lines can be moved by spending at all. A group whose largest specialty is payer-gated should be told so, in writing, before an ad account is opened.

OBJECTIVE — what we find on intake

What we find on intake

A handful of attributed channels across many service lines · one blended acquisition cost covering specialties whose true costs differ by a wide margin · no shared instrumentation stack · monthly reporting assembled by hand

ASSESSMENT

The reporting nightmare isn't a dashboard problem — it's an instrumentation problem. Until every service line reads on the same instruments, specialty-level CAC is unknowable and every budget meeting is a negotiation instead of a readout.

PLAN

What we do: internal referral instrumentation, service-line acquisition, site-level reporting

The internal referral is instrumented first, as a funnel with named steps — placed, contacted, scheduled, seen — reported by referring specialty and by receiving specialty. It is the cheapest volume in the organisation and it is almost never measured.

Acquisition runs by service line, never blended. Each line gets its own demand read, its own capture instrumentation and its own cost per new encounter, so that a budget conversation can be about marginal return rather than about who argued best in the meeting.

Reporting is site-level and assembled by system rather than by staff. One instrumentation stack across every line, one definition set, and a monthly document a CFO can read without a translator and without waiting two weeks for someone to build it.

Access is treated as a marketing input, because in this audience it usually is: a specialty that cannot see a referred patient within a reasonable window will lose that patient regardless of what the demand side does.

WHAT WE MEASURE

What we measure: referral capture rate, revenue per FTE physician, cost per new encounter

Internal referral capture rate — placed against seen inside the organisation
Revenue per FTE physician, read alongside work RVUs per FTE
Cost per new encounter, by service line, never blended
New patient volume by service line and site, with access time beside it
Reporting cycle time — days from month-end to a finished document
Instrumentation stacks in use, counted
Months to system

REFERENCE RANGES

Reference ranges for multi-specialty groups

REFERENCE RANGES FOR THIS SPECIALTY →

RELATED CASE FILES