Measuring Pharma Marketing ROI by Channel Without Double Counting
How to measure pharma marketing ROI by channel without double counting: one value per Rx, incrementality, MMM, and a reconciled view.
The short answer
Channel ROI in pharma is often overstated because each channel's attribution report claims the same prescriptions. To measure it properly, use one value per prescription across channels, estimate incremental rather than attributed prescriptions, reconcile channel results so their sum does not exceed total incremental volume, and use MMM calibrated with experiments to split shared credit. Report ROI as ranges with stated assumptions.
Ask each channel partner for ROI and you will likely get strong numbers from all of them. Add them up and the total exceeds the brand's growth. That is double counting, and it is the most common problem in pharma ROI reporting.
Why double counting happens
- Overlapping exposure. An HCP sees programmatic, email, point of care, and a rep. Each channel's report credits the prescription.
- Attributed, not incremental. Reports count prescriptions from exposed HCPs or patients, not prescriptions caused by exposure.
- Different windows and definitions. Each partner uses its own lookback windows and counting rules.
Step 1: Use one value per prescription
Agree with finance:
- Net revenue per prescription after rebates and discounts.
- Expected persistence, so a new patient is valued over their likely time on therapy.
- Same value across all channels.
Step 2: Estimate incremental, not attributed
For each major channel, get an incremental estimate from:
- A holdout or matched-control study. See matched-control studies.
- Geographic tests where HCP-level holdouts are not possible.
See attributed vs. incremental prescriptions.
Step 3: Reconcile
Check that the sum of channel incremental estimates is plausible against total incremental volume. If it is not:
- Look for overlap in exposed groups.
- Use MMM to split shared credit.
- Adjust with a documented method.
Step 4: Calibrate MMM
Marketing mix models estimate each channel's contribution using time-series data. Calibrate them with experiment results so their estimates line up with what tests showed. See explainable MMM and MMM vs. Rx attribution.
Step 5: Report ranges
Hypothetical example of a reconciled view:
| Channel | Spend | Incremental Rx (range) | ROI range |
|---|---|---|---|
| HCP programmatic | $2.0M | 3,000 to 4,500 | 0.8 to 1.7 |
| Point of care | $1.2M | 1,800 to 2,800 | 0.8 to 1.8 |
| $0.4M | 900 to 1,500 | 1.7 to 3.5 |
Assumes a net value of $1,200 per incremental prescription, agreed with finance; ROI = (incremental Rx x $1,200 minus spend) / spend. Ranges reflect study uncertainty.
What to do with results
- Shift budget at the margin, not wholesale, based on response curves. See promotional response curves.
- Re-test channels whose ranges are wide.
- Keep a learning budget for channels without good estimates.
Common mistakes
- Summing vendor-reported ROI.
- Gross price per prescription.
- Different windows for different channels.
- Single-point ROI with no uncertainty.
Practical takeaway
Add up your channels' reported attributed prescriptions for last year and compare with total new prescriptions. If the sum is close to or above total volume, your ROI reporting is double counting, and reconciliation should come before any reallocation.
Frequently asked questions
Why do channel ROI reports add up to more than total sales?
Because the same patient or HCP was exposed to several channels, and each channel's attribution report counts the resulting prescription.
What value per prescription should be used?
Net revenue per prescription after rebates and discounts, adjusted for expected persistence, applied the same way to every channel.
Is MMM better than attribution for channel ROI?
MMM is designed to split credit across channels, but it needs good data and calibration. Attribution is useful for campaign reads. Together, with experiments, they give a better view.
Sources
- Veeva, Crossix
- Google for Developers, Meridian
- American Statistical Association, Statement on p-Values
External guidance and platform documentation change. Links were current at publication; check them again before relying on them for a decision.
Editorial note. Analysis and frameworks are the author's own and do not represent Acxiom or any current or former employer, client, or named platform. Examples labeled hypothetical or illustrative are not results from real campaigns. Nothing here is legal, regulatory, or medical advice.
New pharma programmatic breakdowns, occasionally
One email when I publish something worth reading. Benchmarks, measurement teardowns, and case studies with the caveats attached. No cadence promises, no reselling your address.
Unsubscribe any time. See the privacy policy.
Working through this decision on a real plan?
I work on health and pharma data, identity, and activation, after five years running HCP and DTC programmatic agency-side. Happy to talk through how this applies to your situation.