The questions, not the services
Pharma leaders don't wake up needing audience planning. They need to know whether they're reaching the right people, at the right moment, whether it worked, and where the next dollar goes. The same claims foundation sits under all four, so the answer to one informs the next.
Are we targeting the right HCPs and patients?
What you get: Less media against audiences that were never going to convert, and a target list your brand team can defend line by line.
Most target lists are built from specialty and geography, because that is what list vendors can supply. Claims support a more specific question: which prescribers have the eligible patients, how many of those patients are not yet on your therapy, who is gaining volume and who is losing it, and who has adopted early in this category before.
In one study, prescribers with an upward pre-campaign trajectory converted four percentage points higher than flat or declining ones. Trajectory is not visible in a specialty file. It is visible in claims.
- Scored target lists with the scoring logic exposed
- Opportunity gap by prescriber, not just volume
- An honest statement of where the data was too thin to support a claim
Are we reaching them at the right point in the care journey?
What you get: Venues validated before you commit budget, and message sequencing built on observed patient timing rather than assumed timing.
A media plan is a set of assertions about where patients are and when. Those assertions are testable. Pathway analysis traces patient flow through claims in both directions: do patients at this venue go on to reach your target prescribers, and do your target prescribers’ patients pass through this venue? The two answers mean different things and imply different creative.
In one analysis, the majority of patients reached a target prescriber within 30 days of the venue visit and over 80% within 90. In another, most venue visits turned out to happen after the prescriber visit rather than before, which meant the placement was reaching patients already in treatment rather than approaching a decision. The plan had assumed the opposite. Same audience, same venue, different message entirely.
- Pathway rates by venue, measured both directions
- Journey timing distributions that define the sequencing window
- Signals that drive creative selection: indication, line of therapy, journey stage
Did the campaign actually change prescribing behavior?
What you get: A causal read your brand team can take into a budget conversation and defend on the methodology, not just the number.
Campaign and control cohorts propensity-matched on pre-period prescribing, specialty, geography, and trajectory. Symmetric pre and post windows, so conversion metrics are not inflated by a longer post-period. Balance validated before any read is reported: if the cohorts do not track before the campaign starts, the design is rebuilt. Difference-in-differences and ANCOVA run in parallel, so two methods with different failure modes have to agree.
And the plan is documented before launch. Not to defend the study afterward, but because a campaign designed to be measured produces a better answer than one measured after the fact.
- Lift with confidence intervals, dual-method validated
- Cohort balance evidence you can inspect
- New-to-brand and total volume read separately, because they mean different things
Where should the next dollar go?
What you get: A reallocation recommendation with the evidence behind each part labeled, so you know which parts are measured and which are informed judgment.
Lift decomposed by segment, specialty, geography, and tactic turns a single number into a decision: where the effect concentrated, where it did not, and what moving budget between them is projected to produce.
We label evidence by tier: measured, benchmarked, or assumed. Early in a relationship, more of it is benchmarked than measured, and the recommendation says so. As reads accumulate across brands and channels, the tiers upgrade. Overstating the precision of an early allocation is the fastest way to lose a client who understands modeling, and most of our clients do.
- Lift decomposed by segment and tactic
- Scenario comparison against a stated objective
- Every input labeled measured, benchmarked, or assumed
Start with a decision you need to defend
The most useful first conversation is usually about a specific choice a brand team is about to make, or one they made and can't explain.
Start a conversation