How to Evaluate an AI-Generated Pharma Media Plan
A red-team checklist for AI-generated pharma media plans covering invented facts, hidden assumptions, source quality, math consistency, and feasibility.
The short answer
AI tools can produce a complete-looking media plan in minutes: objectives, audiences, channels, budgets, KPIs, and timelines. The output is often plausible. It may also contain invented benchmarks, platform features that do not exist, budgets that do not add up, and assumptions that do not fit pharma's constraints. Treat an AI-generated plan as a first draft that needs a structured review.
The red-team checklist
1. Invented facts
- Are there specific statistics, benchmarks, or CPMs? Where did each come from?
- Are named platforms, partners, or features real and described accurately?
- Are regulatory statements correct and current?
Ask the tool for sources, then check them yourself. A cited source that does not exist or does not say what was claimed is a red flag for the whole plan.
2. Hidden assumptions
- What audience size does the plan assume? Is it realistic?
- What does it assume about match rates, CPMs, and conversion rates?
- Does it assume creative flexibility the brand does not have?
Write each assumption down and label it. See the assumption ledger approach.
3. Math consistency
- Do channel budgets add up to the total?
- Do impressions, CPMs, and budgets agree?
- Are reach and frequency consistent with impressions and audience size?
AI tools frequently produce numbers that do not reconcile. Recompute key figures.
4. Pharma fit
- Does the plan account for medical, legal, and regulatory review timelines?
- Does it include required risk information in formats?
- Does it use data sources that are permitted?
- Does it avoid audiences or contexts that would be inappropriate?
5. Feasibility
- Can the proposed partners actually deliver this?
- Is the timeline realistic?
- Is the measurement design capable of answering the plan's questions? See the measurement design guide.
6. Strategic logic
- Does the plan connect the business problem to the tactics?
- Would it pass the investment thesis test?
Score the plan
| Area | Issues found | Severity |
|---|---|---|
| Invented facts | ||
| Hidden assumptions | ||
| Math | ||
| Pharma fit | ||
| Feasibility | ||
| Strategic logic |
A plan with any high-severity issue in invented facts or pharma fit should be rebuilt, not patched.
Where AI plans help
AI-generated plans can be useful for structure, for surfacing options the team had not considered, and for speeding up first drafts. The value comes from the review, not from the draft itself.
Practical takeaway
Never forward an AI-generated media plan without running the six-part checklist. Keep a record of issues found; patterns show where the tool is least reliable for your work.
Sources
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.