Competitor ad analysis template: from source evidence to a test brief

Start with the decision your next creative must support

Write one research question before you collect ads. For example: “How do advertisers show that this product fits a small space?” This gives you a reason to save an example and a reason to reject an unrelated one.

Set the product category, audience, market, and intended action. Compare ads for similar offers within each platform first. A subscription trial and a one-time purchase can use similar images while asking for different customer commitments.

Your output is a creative hypothesis with traceable evidence. A public ad record can show a message or format. It does not, by itself, prove spend, conversions, return on ad spend, or profit.

Use the right source for YouTube, Meta, and TikTok

YouTube: Google’s Ads Transparency Center supports searches by advertiser or website, plus date and targeted-location filters. It includes YouTube ads. Save the source record and the filters you actually use. Keep the advertiser identity separate from an assumed consumer brand.

Meta: start with Meta Ad Library for Facebook and Instagram creative research. Save the Page identity, visible ad details, and search conditions. If a field or page is unavailable, record that limit. Do not apply political-ad spend disclosures to ordinary commercial ads.

TikTok: Top Ads in Creative Center contains selected ads that meet performance thresholds and have advertiser permission. Its video analysis uses indexed metrics. Use those signals to locate a creative moment for review; they do not give you a complete market sample or a verified profit figure.

Keep the source name on every record. A TikTok selection and a Google advertiser lookup do not represent the same sample. Compare the creative choices you can see, with the source limits attached.

Sources: Google: Ads transparency · Meta Ad Library · Meta: Ad Library background and disclosure distinctions (March 28, 2019) · TikTok: Top Ads selection and indexed video analysis

Keep the placement context for NewsBreak, Rumble, and native ads

NewsBreak documents several formats and placements, with availability that depends on the selected objective and platform. Record the actual format and visible context. A captured image does not establish every placement where the advertiser delivered it.

Rumble’s campaign guide distinguishes creative format, publisher inventory, and optional targeting. A video observation cannot reveal those campaign settings. Record a visible channel or publisher only when the source provides it.

For other native networks, name the network and publisher when known. Compare the headline, thumbnail, page promise, and next action as separate elements. If the destination fails to load, leave it unavailable. An advertiser homepage is not a substitute for the missing ad destination.

SwipeHarbor provides captured-ad research for NewsBreak and Rumble. This method also applies to other sources, but it does not imply a SwipeHarbor library for YouTube, Meta, TikTok, Taboola, Outbrain, or MGID.

Sources: NewsBreak: Ad formats and placements · Rumble Advertising Center: Video campaign settings

Copy this competitor ad analysis template for each ad

Keep one evidence record per source ad. Put these fields in your notes or in the brief builder’s Evidence / source notes field. Use “unavailable” for a missing field and “not checked” for a step you did not complete.

• Research question: the audience question this example helps you study.
• Source: platform, library name, record URL, and visible ad identifier.
• Search context: query, country, date filter, format filter, and your observation date.
• Identity: displayed advertiser, Page or account, and the product shown. Flag any uncertain connection.
• Creative: opening image, exact headline or first line, product demonstration, and call to action. Add video timestamps where useful.
• Claim and support: what the ad promises and what evidence it displays. A displayed claim is not a verified fact.
• Destination: declared URL, reached URL, page headline, offer terms, requested action, and page observation date.
• Dates and status: copy each source label with its value. Keep first seen, last seen, upload date, and your review date separate.
• Evidence limits: unavailable media, blocked pages, uncertain identity, source selection rules, and unknown campaign results.
• Interpretation and test: your explanation for the creative choice, an alternative explanation, and one original change to test.

Compare creative patterns without a false performance ranking

Use the same descriptive labels across your records: problem shown, demonstration shown, claim explained, or offer introduced. Describe the sequence before you judge it. “The product appears before the price” is an observation. “The audience trusts it more” is a hypothesis.

Group similar versions within an advertiser before you count distinct ideas. Repeated assets can make one concept look like many independent examples. Keep the source records so another buyer can check your grouping.

Choose a test candidate for its relevance to your audience question, the available evidence, and your ability to create an original version. Do not assign a performance score from ad age or duplicate counts. Continued visibility does not establish continuous delivery or profitable results.

Across platforms, transfer a question or creative principle rather than an assumed result. A clear product demonstration can inspire a new NewsBreak concept. Its presence in TikTok Top Ads does not predict that concept’s NewsBreak conversion rate.

Worked example: a compact desk product

This is a fictional planning exercise, not a captured campaign or customer result. It shows how to complete the existing brief fields. Replace the example notes with evidence from your own research before you use it for a real decision.

For this purchase-focused example, define the primary measure as cost per purchase from your own campaign records. Set the acceptable cost, measurement window, and review rule before launch. Use your own economics to set them; the source ad cannot supply those values.

Use a platform experiment where appropriate. Unequal delivery can complicate a simple ad-to-ad comparison. If the test lacks enough evidence for your planned decision, record it as inconclusive. Do not select a winner solely because one opening receives more clicks.

• Project / product: a compact desk organizer.
• Research question / goal: does the opening explain the available desk space clearly?
• Audience / market: people who need a small work area; set the actual market before the test.
• Offer / promise: the same organizer and purchase terms in both versions. Use only product claims you can support.
• Evidence / source notes: fictional concept A opens with a cluttered desk; fictional concept B opens with a size demonstration. No source ads or campaign results support this example.
• Evidence / source notes (interpretation): a size demonstration might answer a space concern sooner. Alternative explanation: the cluttered desk might make the problem easier to recognize.
• Next test / success metric: compare those two original openings on one platform. Keep the remaining video, offer, destination, and intended audience consistent.

Check the evidence before the creative handoff

A second person should be able to find each source and distinguish its visible details from your interpretation. If they cannot, resolve that gap before you present the idea as research-backed.

1. Check each ad link and the recorded search conditions.
2. Label each statement as an observation, an interpretation, or an unknown.
3. Confirm that every date retains the source’s meaning.
4. Remove unsupported spend, sales, and profit claims.
5. Specify the original creative change and the decision the test must support.

Turn the template into a free ad research brief

Use the free ad research brief builder linked below. Enter the project, question, audience, offer, evidence, and next test. Choose the platforms you researched. Then review the checklist and edit the final brief.

The tool needs no account. It organizes your text; it does not fetch source links or verify the evidence. Copy or download the brief before you leave. The tool does not retain your work after a reload.

For a paid research library, check the current coverage before purchase. SwipeHarbor currently covers NewsBreak and Rumble only. This article’s wider research method and the free tool’s platform choices do not expand that capture coverage.

We reviewed the linked platform documentation on October 12, 2026. The template and fictional exercise are editorial guidance. They are not a report of an advertiser’s test results.

Sources: Free ad research brief builder — no account required · SwipeHarbor: current capture coverage