Promotional Response Curves: Measuring Diminishing Returns in Pharma Spend
What promotional response curves show, how they are estimated in pharma, how to read saturation and carryover, and how to reallocate budget.
The short answer
A promotional response curve shows how outcomes, such as new prescriptions, change as spend or exposure in a channel increases. Most channels show diminishing returns: each extra dollar adds less than the one before. Curves are estimated from marketing mix models, experiments at different spend levels, or HCP-level frequency analysis. They are used to move budget at the margin, from channels near saturation to channels still on the steep part of their curve.
Average ROI hides the most useful information for budget decisions: what the next dollar will do. Response curves show that. They are how you find channels that are over-funded and channels that could use more.
Shape of a typical curve
Most channels follow a curve that rises steeply at first, then flattens:
- Low spend: each dollar adds a lot.
- Middle: returns decline.
- High spend: saturation; extra dollars add little.
Some channels have a threshold where little happens until enough reach or frequency builds.
Average vs. marginal return
| Spend level | Total incremental Rx | Average Rx per $10k | Marginal Rx per extra $10k |
|---|---|---|---|
| $500k | 1,500 | 30 | 20 |
| $1.0M | 2,300 | 23 | 10 |
| $1.5M | 2,700 | 18 | 6 |
Hypothetical figures. At $1.5M, average return still looks fine, but the next $10k adds only about 6 prescriptions. Budget decisions should use marginal return.
How curves are estimated
- Marketing mix models. Estimate saturation from variation in spend over time. See explainable MMM.
- Spend-level experiments. Run different spend levels in comparable regions.
- HCP-level frequency analysis. Compare outcomes by exposure level, with caution about selection bias. See effective frequency.
Combining methods gives more confidence than any one alone.
Carryover
Some effects last beyond the week of spend. Models capture this with carryover (adstock). A channel with long carryover may look weak in short windows and strong over months.
Using curves to reallocate
- Compare marginal returns across channels at current spend.
- Move budget from the lowest marginal return to the highest.
- Move in steps, not all at once.
- Re-measure after each move.
Stay within ranges the data has observed. Curves are unreliable far beyond tested spend levels.
Uncertainty
Show curves with uncertainty bands. If bands overlap heavily between channels, the data cannot tell you which is better at the margin. Plan a test.
Common mistakes
- Using average ROI to decide where to add budget.
- Extrapolating curves to untested spend levels.
- Ignoring carryover.
- Large reallocations from one model run.
Practical takeaway
Ask your analytics partner for each major channel's marginal return at current spend, with uncertainty ranges. Move a modest share of budget from the lowest to the highest and measure before moving more.
Frequently asked questions
What is a response curve in marketing?
A curve showing the relationship between marketing input, such as spend or exposures, and an outcome, such as prescriptions.
What does saturation mean?
The point where additional spend in a channel adds little extra outcome, because the audience is already well covered or frequency is high.
How reliable are response curves?
They have uncertainty, especially at spend levels the brand has never tried. Experiments at different spend levels improve reliability.
Sources
- Google for Developers, Meridian
- Meta Open Source, Robyn (GitHub)
- NIST/SEMATECH, Confidence Intervals
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.