How to write a competitor ad strategy analysis report

A competitor ad strategy analysis report is a dated document built from six fixed sections: creative inventory, longevity and testing cadence, offer and positioning, spend signal, channel mix, and funnel stage coverage. It closes with a prioritised action list of three to five moves. The sections stay identical every cycle, so reports stay comparable.

Most competitive ad research dies in a folder of screenshots. The report is what survives, because a report has a fixed shape: same sections, same order, same units, every cycle. That repetition is the whole point. A stakeholder reading the third report in a series sees movement instead of a fresh pile of observations.

What a competitor ad strategy analysis report is and is not

It is not a tool export and not a one-off audit. It is a standing brief on how a defined competitor set buys attention, structured so findings can be compared across dates. A tool export shows what exists today. A report shows what changed, what the change implies, and who owns the response.

It is also not a budget estimate. Sizing a rival's spend is a separate exercise with its own error bars, and our guide on estimating a competitor ad budget covers that calculation end to end. A strategy report uses spend as one graded signal among several, never as the headline.

Two audiences read the same document. An internal performance team wants the operating detail: which angle is fatiguing, which format is underused, which landing page is being outbid. A client stakeholder wants the top three findings and the decision each one implies. One report serves both when the first page is written for the second audience and the appendix carries the first.

What sections should a competitor ad strategy analysis report contain?

Six sections earn their place, in this order: creative inventory, longevity and testing cadence, offer and positioning, spend signal, channel mix, and funnel stage coverage. Each one answers a question the next one depends on. Skip the inventory and the longevity analysis has nothing to age. Skip the action list and the document reads as research rather than strategy.

Every section carries the same three labels, whichever competitor set you track:

• Observed: what the ad library or platform API actually returned, plus the date it was pulled.
• Inferred: what that observation implies once you allow for the platform's known gaps.
• Estimated: anything you modelled, with the assumptions written next to the number.

Confidence grading is the part most templates leave out, and it is the part that keeps a report honest. When a client asks how you know a competitor tripled spend, the answer needs to point at a dated observation, not at a dashboard colour. Three labels do that work: observed, inferred, estimated.

Section 1: creative inventory

The inventory is the foundation. It logs every live and recently live ad in the competitor set with platform, format, first-seen date, last-seen date, and a thumbnail. The dates matter more than the creative itself, because everything downstream is calculated from them.

The scale of the problem is measurable now. Motion's Creative Benchmarks 2026 analysed $1.29 billion in Meta ad spend across 578,750 creatives from 6,015 advertiser accounts, covering a data window of September 2025 to January 2026.

Only 4 to 8 per cent of ads are winners across every spend tier in Motion's 2026 sample of 578,750 creatives. Half lose outright and the other half survive without scaling, while Enterprise accounts concentrate 64 per cent of spend on the winners. Source: Motion, “Creative Benchmarks 2026,” data window September 2025 to January 2026.

That hit rate reframes the inventory. You are not cataloguing art. You are building the denominator that lets you say whether a competitor is testing enough to find winners, or simply producing more ads that will never scale.

Section 2: longevity and testing cadence

Volume tells you what a competitor is testing. Longevity tells you what won. A brand running forty variants of one concept for eight straight weeks has found something and is milking it. A brand cycling forty fresh creatives every five days is still in testing mode, and copying this week's version spends the client's budget on an ad that has not earned its place yet.

Bucket every unit by age and the pattern surfaces on its own: under 14 days is active testing, 14 to 45 days is a validated performer, and 45 days or more is a control the competitor trusts enough to leave alone. One date range does more analytical work than any traffic estimate in the document, and it is auditable because the raw dates sit in the inventory beside it.

Testing cadence is the second half of this section. The same 2026 dataset found the top quartile of Enterprise accounts testing 54 new creatives per week against an average of 19, and producing ten winners a month against four. That is roughly 2.9 times the testing volume, and the winner output scales with it.

Volume is a structural advantage, not an instruction to spray. A solo founder or a ten-person team cannot match Enterprise volume, which is why the hook rate, hold rate and format benchmarks in our ad creative performance benchmarking breakdown are worth reading as a floor rather than a target.

Section 3: offer, positioning and channel mix

Creative without the offer is decoration. This section records what the competitor actually sells inside the ad: the promise, the price anchor, the risk reversal, and the destination. Group the entries across the set and you get a positioning map rather than a wall of screenshots.

Group hooks by the psychological angle they run, not by the words on the ad. Problem and agitate, social proof, mechanism reveal, urgency, and authority are angles. A single line about doctors is just one instance inside the mechanism-reveal bucket. Clients act on the pattern, not on the individual line, so the report should present the pattern.

A share-of-active-creative column turns the grouping into a decision. If a competitor puts 70 per cent of live creative behind social proof and the client is running mechanism reveals, that is a strategy finding, and it belongs in the report with the counts printed underneath it.

Compliance flags belong in this section too, and most templates skip them. A hook that implies a guaranteed outcome, an unsubstantiated before-and-after, or a testimonial with no verifiable source is a liability signal for the client's whole category, not a curiosity about one competitor. Flag it, date it, and note whether the ad is still running.

Channel mix sits beside offer and positioning because the two travel together. A brand pushing a free trial on Meta and a demo request on LinkedIn is running two funnels with two different economics. Mapping which offer appears on which platform is often the most useful page in the whole document.

Judging individual creatives on Meta and Google needs its own method, and our breakdown of competitor ad creative analysis across Meta and Google covers the platform-specific checks. This report section only needs the output of that method: angle, format, and the age bucket each unit sits in.

Section 4: spend signal, graded by confidence

Spend is where reports lose credibility. Ad library tools publish a range, not a figure, and the true number usually sits outside that range once agency fees and non-platform channels are added. Print the range you worked from and label it estimated, every time.

Use corroborating signals instead of pretending to precision. Creative count, geographic spread, and the depth of a tracking setup all scale with budget. A competitor running the same offer across six English-speaking markets with fifteen or more active variants per market is very likely spending mid five to six figures monthly on that offer alone, but that sentence belongs next to the word inferred, not stated as fact.

Trend beats snapshot. A single spend estimate is nearly meaningless on its own. Six months of the same estimate, produced the same way each time, shows whether a competitor is scaling or pulling back, which is why our guide to competitor ad spend tracking builds the section as a series rather than a number.

Section 5: funnel stage coverage

Funnel stage coverage is what stops a report becoming a creative gallery. For each angle and each format, record the stage it appears to target: cold acquisition, warm retargeting, or conversion. A competitor with forty ads at the top of the funnel and none at the bottom is leaving money unclaimed, and that gap is a recommendation with a number attached.

Building the stage map takes a method of its own, and our competitor ad funnel analysis walkthrough covers the four-step version. The report needs the finished map plus a count of live units per stage.

The format mix adds a second reading of the same data. If a competitor's retargeting layer is entirely static image while the client's is video, the difference is rarely taste. It usually signals a production constraint on one side, and knowing which side carries the constraint changes what you recommend.

How do you structure findings for a client or an internal stakeholder?

Lead with the decision, then the evidence. An executive summary of two or three paragraphs names the two or three competitive threats, the biggest opportunity, and one recommended action. Decision-makers read this section first and sometimes only, so it has to stand alone.

After the summary, keep data and interpretation in separate callouts. A chart followed by a sentence starting with what this means is readable in five minutes. A chart followed by three tables of raw numbers is a data dump with a header.

Date-stamp everything. Every figure needs the date it was pulled and the tool it came from. Competitive landscapes move weekly, and an undated number in month three is worse than no number, because it invites a decision on stale evidence.

Automation has moved the bottleneck. Forrester and the 4A's reported in 2026 that 70 per cent of agencies now apply generative AI to research and competitive intelligence, with 74 per cent using it to summarise documents and communications. Two years earlier the leading use case was ideating creative concepts. The collection step is no longer the differentiator, so a 2026 agency AI adoption roundup puts the analysis, the grading and the recommendation at the centre of the fee.

If the collection step is still manual, our walkthrough of automating competitor ad research covers the discovery and extraction layers. The report structure does not change when collection is automated. It arrives faster and with fewer transcription errors.

A report and a dashboard are different deliverables and clients benefit from both. A live dashboard is for monitoring, and our guide to building a competitive ad intelligence dashboard for clients covers that build. A report is a dated artifact with an argument in it, and it is the thing a client forwards to their board.

How often should you refresh the report?

Match the cadence to the competitor's testing velocity, not to the calendar. A rival shipping fifty creatives a week invalidates a monthly report before it lands. A rival running the same three ads for two quarters does not need weekly attention. Measure the churn rate first, then set the cycle.

Three bands cover most cases. High-churn categories such as ecommerce supplements and mobile gaming justify a weekly cycle. Most B2B and considered-purchase categories work on a monthly cycle. Slow categories with long procurement cycles are fine quarterly. Whatever you pick, freeze the section order and the column definitions so the versions stay comparable.

The highest-value section in any repeat report is the diff. What changed since last cycle, what did not, and which recommendation from last time the client actually shipped. That last column is uncomfortable, and it is the reason clients renew.

The report template: a checklist you can run this week

Run this checklist before the report leaves your desk. It takes ten minutes and catches the errors that make a deliverable look careless.

1. Scope line on page one: the competitor set, the date range, and the platforms covered.
2. Every creative logged with first-seen and last-seen dates, not just a screenshot.
3. Each age bucket stated with its boundary, so 14 days and 45 days mean the same thing next cycle.
4. Angles grouped as patterns with a share-of-active-creative column, not listed as individual lines.
5. Every number labelled observed, inferred, or estimated.
6. Funnel stages covered for both the client and the competitor set, with counts.
7. Executive summary on page one, capped at three findings and one action.
8. A dated action list of three to five moves, each tied to a specific finding.
9. Version number and pull date on the cover, plus a one-paragraph diff against the previous cycle.

The template is the easy part. The hard part is holding the structure while the findings change, and refusing to let the report slide into a folder of screenshots with a cover page. Six sections, three confidence labels, one dated action list. Repeat until the client can read the diff without you in the room.

Frequently asked questions