How to Choose an Rx Attribution Window

Separate four clocks

These clocks should not be collapsed into one “30-day window.”

Ground timing in the use case

HCP education for a chronic therapy, consumer awareness for an acute seasonal condition, and refill support have different plausible lags. Brand, analytics, and appropriate subject-matter experts should define the journey. Media teams should not make clinical assumptions independently.

Test window sensitivity

Report outcomes under a primary window and a small number of prespecified alternatives. Watch whether the conclusion depends on an arbitrary cutoff. Plot cumulative attributed outcomes by days since exposure and compare exposed and control patterns where available.

Hypothetical example

A campaign shows 600 attributed prescriptions at 30 days and 850 at 60 days. That does not mean the additional 250 were caused by media. If control outcomes grow similarly, incremental lift may not change. Figures are illustrative.

Prevent common distortions

• Apply consistent windows across comparison groups.
• Distinguish view-through from click-through rules.
• Deduplicate repeated exposures and channels under a documented rule.
• Separate new-to-brand, refill, and total outcomes.
• Wait for comparable claims maturity.
• Do not select the best-looking window after results arrive.

Questions to settle before launch

Agree on these in writing with the measurement partner and brand analytics before the first impression serves:

1. What is the primary window, and what two alternatives will be reported? Naming the alternatives in advance removes the temptation to pick the best-looking one later.
2. Does the window start at first exposure, last exposure, or each exposure? Each rule credits different prescriptions, especially for long flights with repeated frequency.
3. How are prescriptions that occur after exposure but before the outcome window opens handled? Some designs include a short lag to avoid crediting prescriptions already in motion.
4. When is the data considered mature? Claims arrive and adjust over weeks. Reporting on an immature window understates outcomes and makes short windows look worse than they are.
5. Will the comparison group use identical windows? It must, or lift estimates are not comparable.

If the brand wants a single headline number for a dashboard, report the primary window and footnote the alternatives. That keeps the simplicity leadership asks for without hiding how much the answer depends on timing. See sensitivity analysis for a structured way to show it.

Practical takeaway

Window choice belongs in the measurement protocol, not in post-campaign negotiation. The next step is a timing diagram that shows exposure, baseline, outcome, and claims-maturity periods with rationale for each.

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.