Why “340 meetings this quarter” was never a measure of anything

Medical affairs KPIs built on activity — meetings held, healthcare professionals reached, materials shared — measure how much a field medical team did, not what any of it achieved. The distinction is not academic, because a metric that cannot see quality will eventually reshape the work around volume.

Most field medical reporting still opens with a count. It is the number leadership expects, the number the CRM produces without being asked, and the number that appears in every quarterly deck. The argument for why Medical Affairs has no owned outcome metric is a structural one, and it is worth reading separately. This piece is narrower: what the activity number actually tells you, what it cannot tell you, and what to put beside it in the meantime.

Where did the activity target come from?

Activity targets did not originate inside Medical Affairs. They arrived from commercial field organisations, where a call-rate target is a defensible measure of capacity: a sales representative has a territory, a set of accounts, and a finite number of hours, and call frequency correlates reasonably with coverage of that territory.

Transplanted into a medical function, the same metric measures something quite different. An MSL’s value is not proportional to the number of HCPs reached, because scientific engagement is not a coverage exercise. Six substantive conversations with the clinicians who shape practice in a therapeutic area may matter more than sixty introductory ones. The metric has no way to express that, because it was never designed to.

It persisted for a practical reason worth acknowledging. Activity data was already being captured, already structured, and already reportable. Nobody had to build anything. When a function is asked for a number and has one available, the available number wins.

What can a volume metric not distinguish?

Three distinctions matter, and a count collapses all three.

Substance. A conversation that confirmed what an HCP already believed and one that surfaced an evidence gap nobody in the organisation knew about are the same entry in the system. So are a two-hour scientific discussion and a fifteen-minute corridor exchange logged on the way out of a hospital.

Difficulty. The sceptical academic who challenges a trial’s applicability takes three attempts to secure a meeting with and produces an uncomfortable conversation. The supportive clinician takes one email. Both count once. The metric quietly prefers the second, and so, over time, does the diary.

Direction. Activity counts what went out. It has no field for what came back — the unmet need described from practice, the treatment barrier, the question nobody anticipated. Those are the outputs that justify the function’s existence, and they do not appear in the number at all.

None of this makes the activity number false. It makes it silent on everything that matters.

What happens when a proxy becomes a target?

This is Goodhart’s law: when a measure becomes a target, it stops being a good measure. It applies here with unusual force, because the people generating the data are also the people the data is used to assess.

The mechanism is not cheating. It is drift. Faced with a quarterly number to reach, a rational MSL schedules the meetings most likely to happen — the receptive HCPs, the shorter conversations, the accounts where access is easy. Every individual decision is defensible. The aggregate effect is a field team optimised for the conversations that were never the point.

The second-order effect is worse. A team that knows the number does not capture quality stops trying to record quality. The free-text field gets shorter. Insight capture becomes a compliance exercise. And the organisation loses not just the measure but the underlying information, because nobody has a reason to log it.

That is why replacing the metric is not only a reporting question. The reporting shapes what gets recorded, which determines what can ever be measured.

What should sit next to the activity number?

The practical answer is not to delete it. Activity data is genuinely useful for capacity planning, territory design and workload management — the things it was built for. The error is reporting it alone, as though it were an outcome.

Two measures belong beside it, and the argument for both is set out in detail in talking-point coverage and insight yield.

Coverage answers whether the science planned for a given HCP was actually discussed. It is derived from the organisation’s own approved content, which makes the benchmark defensible rather than arbitrary.

Yield answers what came back — whether the conversation produced anything the organisation did not already know.

Reported together, the three numbers say something a count alone cannot. Two hundred interactions with high coverage and low yield describes a team delivering messages efficiently and learning nothing. Sixty interactions with high yield describes something considerably more valuable, and would look like underperformance on any activity report.

How do you change the reporting line without losing it?

Reporting lines are institutional. A metric that has appeared in every quarterly deck for a decade cannot simply be dropped, and attempting it usually fails on politics rather than logic. Three moves make the transition survivable.

Add before you remove. Report the new measures alongside the activity number for at least two quarters. Leadership needs to see them move together before trusting either alone.

Report distribution, not totals. A single quarterly figure hides everything. The same total looks completely different depending on whether it comes from a handful of deep engagements or an even spread of shallow ones, and the distribution is where the interesting question lives.

Start at cohort level, not individual. Team and therapeutic-area reporting first. Individual-level reporting of a quality measure invites exactly the defensive behaviour that degraded the last metric, and it is unnecessary — the useful questions at this stage are about topics and evidence, not people.

Do those three things and the activity number stays where it belongs: a measure of capacity, reported next to measures of substance, no longer asked to do a job it was never designed for. What the deck looks like once it does is a quarterly review of the science rather than the volume.

RocketMSL generates coverage and yield from your own scientific content, alongside the activity data you already have. See how the platform measures scientific engagement →