Retail Media Campaign Measurement Report
Turn a finished retail media campaign into a performance readout.
What is the Retail Media Campaign Measurement Report?
The Retail Media Campaign Measurement Report is a free AI skill that turns the raw results of a finished retail media campaign into a clear performance readout for food and beverage teams. You give it the campaign's objective, the platform, and the metrics you pulled once it ended; it returns a performance summary against the original goal, a ROAS and efficiency read, what likely drove the result — creative, targeting, pacing, or competitive pressure — and specific learnings for the next campaign. It is built for shopper and brand teams who have a spreadsheet of Amazon Ads or retailer.com metrics after a campaign closes and need it turned into a report leadership can act on. Because it works strictly from the numbers you pulled, the readout survives a skeptical question about whether the campaign worked. It flags any missing metric as a gap. Pairing the driver analysis with live food and beverage demand data separates a real campaign effect from a category-wide swing that would have happened anyway.
Who it's for
- Shopper and e-commerce managers reporting on a closed retail media campaign
- Brand managers explaining retail media results to leadership
- Sales teams sharing a campaign readout with a retailer partner
- Founders deciding whether to repeat or change a retail media investment
What you get
- A plain-language performance summary against the original goal
- A ROAS and efficiency read from your pasted metrics
- The likely drivers behind the result (creative, targeting, pacing, competition)
- What underperformed and why, where the data shows it
- Specific learnings to carry into the next campaign
- A leadership-ready summary of results and recommendation
How to use it
- 1Copy the prompt into ChatGPT, Claude or Gemini.
- 2Paste your campaign metrics and fill in the brackets (objective, platform, spend).
- 3Answer its clarifying questions about what the metrics do and don't include.
- 4Ask it to turn the learnings into next-campaign targeting or budget changes.
See it in action
Example input
Platform: Amazon Ads. Product: fictional 'NomNoms' dog treats. Objective was to grow hero-SKU sales and defend our brand term over a 6-week campaign. Results: $12,000 spend, $38,000 attributed sales (3.2x ROAS), brand-term impression share held at 91%, but new-to-brand orders were only 18% of total, lower than our 30% goal.
Example output
Retail media measurement report — NomNoms dog treats, Amazon Ads, 6-week campaign Performance summary against goal Mixed result: the campaign hit its efficiency and defense goals clearly but missed its growth-quality goal. A 3.2x ROAS on $12,000 spend against $38,000 attributed sales is a solid return, and holding 91% brand-term impression share means the defend objective — stopping competitors from winning your own brand search — succeeded. But the stated objective also included growing hero-SKU sales specifically through new buyers, and new-to-brand orders at 18% against a 30% goal is a real miss on that half of the objective. ROAS and efficiency read 3.2x ROAS is healthy on its face, but ROAS alone doesn't tell you whether this was new business or harvested sales you'd have won anyway. Given that only 18% of orders were new-to-brand, a meaningful share of that $38,000 in attributed sales likely came from shoppers who already knew NomNoms and would have searched for or bought the brand regardless of the ads — the brand-defense campaign doing its job, in other words, rather than the growth campaign doing its job. This is the central tension in the report: strong headline ROAS partly reflects efficient harvesting, not proof of new-buyer growth. Likely drivers - Brand-term defense performing well is straightforward: exact-match brand campaigns convert efficiently by design, which inflates blended ROAS even when the growth objective underperforms. This is expected, not a surprise finding. - The new-to-brand miss is more likely a targeting-mix issue than a budget issue: if the campaign structure leaned heavily on brand-defense and proven-converting terms rather than broader discovery or competitor-conquesting terms, the 18% new-to-brand rate is a predictable result of that mix, not a sign the product or creative failed to recruit. [Confirm the actual budget split between defend, harvest, and discovery campaign types to verify this read.] - Creative and competitive pressure cannot be assessed from the metrics provided — no creative performance or competitor spend data was included in this pull. What underperformed, and why the data shows it The new-to-brand rate is the clear underperformance: 18% against a 30% goal is a 12-point miss, not a rounding difference. The data available points toward campaign structure (likely too much brand-defense and harvest weighting, too little discovery/conquesting spend) as the probable cause, but this is a directional read from the aggregate numbers provided, not a confirmed diagnosis — a campaign-type-level spend and performance breakdown would confirm or rule this out. Learnings for the next campaign 1. Increase the discovery and competitor-conquesting budget share in the next campaign specifically to address new-to-brand growth, while holding the brand-defense spend that is clearly working. 2. Report ROAS and new-to-brand rate together going forward, never ROAS alone — this campaign shows how a strong blended ROAS can mask a missed growth objective. 3. Pull campaign-type-level performance data (not just the campaign total) next time, so a future readout can attribute results to defend versus harvest versus discovery directly instead of inferring it. Leadership-ready summary The NomNoms Amazon Ads campaign delivered a strong 3.2x ROAS and successfully defended 91% of brand-term impression share over 6 weeks on a $12,000 spend. However, new-to-brand orders came in at 18% against a 30% goal, suggesting the campaign's efficiency was driven more by defending and harvesting existing demand than recruiting new buyers. We recommend shifting budget mix toward discovery and conquesting terms in the next flight to directly target the growth shortfall, while preserving the brand-defense spend that is clearly protecting share. Want me to model a revised budget split for the next campaign based on this read?
The prompt
Here's the start of the prompt. Download the free bundle for the full, ready-to-paste version — plus the installable Claude Skill and Custom-GPT instructions.
# Role You are a senior retail media performance analyst for food and beverage brands. You turn closed-campaign metrics into an honest readout, never letting a strong ROAS number hide a missed piece of the original objective. # Context I'll provide - Platform: [PLATFORM e.g. Amazon Ads, retailer.com network] - Product: [PRODUCT] - Original campaign objective(s): [OBJECTIVE(S)] - Campaign length and spend: [LENGTH + SPEND] - Results, pasted as text: [PASTE METRICS — sales, ROAS, impression share, new-to-brand %, whatever you have] - Audience for this report (optional): [AUDIENCE] # Your task
Frequently asked questions
- What is a retail media measurement report?
- A retail media measurement report is a structured readout of how a retail media campaign — Amazon Ads, Walmart Connect, or another retailer network — actually performed once it closed, measured against the original objective rather than just a headline ROAS number. It names what drove the result and what specifically to change next time. This skill builds that report from the metrics you pull after a campaign ends.
- How is this different from the Retail Media Plan Builder skill?
- The Retail Media Plan Builder plans a campaign upfront, before it runs — objectives, keyword strategy, budget split, and bidding approach. This skill is its counterpart on the other end: it measures and reports what actually happened after the campaign has run, using your real results rather than projections. Use the plan builder to set the campaign up, and this skill once it closes to report honestly on how it performed.
- Which AI models can run this prompt?
- Any capable chat model — ChatGPT, Claude, or Google Gemini. It's model-agnostic, so paste your campaign metrics directly into the chat, save the prompt as a Custom GPT, or store it as a reusable skill so every campaign readout your team produces follows the same honest, objective-first structure.
- What metrics do I need to paste in for this to work well?
- At minimum, spend, attributed sales or ROAS, and whatever metric ties to your original objective — impression share for a defense goal, new-to-brand rate for a growth goal. The more granular your data (campaign-type breakdowns, not just campaign totals), the more specific the driver analysis can be; with only high-level totals, the skill will still report honestly but will mark deeper causes as directional reads pending more detailed data.
Related skills
Coupon & Offer Mechanic Designer
Design the right offer mechanic and the psychology behind it.
Get it freeCross-Shop & Basket Affinity Analyzer
Map which brands and categories your shoppers also buy over time.
Get it freeDigital Coupon & Rebate App Campaign Brief
Brief a cash-back rebate app offer from submission through redemption tracking.
Get it freeWant the live data behind sharper outputs?
These skills get better with real-time F&B intelligence. See what Tastewise can do for your team.