How to Reverse Engineer a Brand's Ad Strategy

When I research a competitor, I want to leave with a clear reason to make an ad. Saving a beautiful reference is easy. Explaining which part belongs in your next campaign takes more care.

To reverse engineer a brand's ad strategy, I work through its public ads in a consistent order: define the sample, map the angles, read the copy, study the visuals, and form gap hypotheses. Here's the method you can use with your next brief.

Step one: define the public ad sample before you reverse engineer the strategy

Start with a customer question. What makes someone consider a product like yours? What objection keeps them from trying it? Choose a rival whose ads address that question, then record which ads you're studying and when you found them.

Keep the ad's source, opening, format, offer, and observed dates together. In StealAds, find a rival's long-running Meta ads and choose references relevant to your question. You can also start with the Meta Ad Library. Keep your selection rule consistent so a colleague can follow the research.

The typical ad in our research ran 16 days, so anything that's been live for months stands out. Start there. Keep what you can see and what you think it means in separate notes, so you can test the second part later.

Step two: map the angles

Read the promise before judging the design. What feeling or practical change is the ad asking the customer to care about? Give each reference a primary angle and write down the words or imagery that support your label.

• Identity: the ad connects the product to who the customer wants to be. Record the role or aspiration it shows.
• Fear and relief: the opening raises a concern, then presents the product as a response. Note the concern and the proposed relief.
• Social proof: the ad uses a testimonial, rating, or endorsement. Record what the person actually says.
• Authority: the ad offers expertise, a certification, or a documented process. Identify the evidence it presents.
• Exclusivity: the creative emphasizes access, membership, or availability. Record the specific offer language.
• Objection response: the ad answers a concern about fit, setup, price, or another obstacle.

Leave an unclear label when the creative doesn't give you enough evidence. I'd rather have an honest open question than force every ad into a neat category.

Then look across your references. Does the same customer concern keep appearing? Do different products use the same promise? Write a plain observation such as "these ads all show the product fitting into a rushed morning." That's something your team can inspect and discuss.

Step three: break down the copy

Break the copy into its jobs: opening, promise, explanation, proof, objection, and action. Note which jobs the headline does and which are left to the image, video, or landing page.

For a short ad, ask what the reader has to understand immediately. For a longer ad, ask what extra explanation earns its place. Don't assume more detail makes an ad more persuasive. Read the actual argument and check whether each sentence adds something useful. The Empty Copy Paradox offers a framework for comparing how much the words and visuals each explain.

Compare the openings. Are they questions, demonstrations, personal stories, or direct offers? Look for repeated language around the customer's problem. If every headline names an ingredient but the body keeps returning to convenience, you've got a useful question for your own brief: should your opening lead with the everyday use case?

Keep exact quotes in your research notes, with their source. Write your own copy for the draft. The ad hook generator can help you try alternative openings once you've decided what your customer needs to hear.

Step four: study the visuals

Look at the creative before reading your notes again. What do you notice first? Can you tell what the product does? Where does the proof appear? For video, follow the opening through to the demonstration and the closing action.

Record the visual choices you can see:

• Format: image, video, or carousel, and what the format allows the ad to explain.
• Product visibility: whether the product appears immediately, in use, or after the problem is introduced.
• Composition: where the person, product, headline, and background sit relative to each other.
• Text overlay: which message gets the largest type and whether it repeats or extends the caption.
• Trust signals: where the creative puts a testimonial, rating, demonstration, or other evidence.

Compare ads with similar promises. A close product shot and a customer demonstration can make the same claim feel very different. Your research note should describe that difference clearly enough for a designer to sketch it.

For category context, use the skincare industry guide to think through demonstrations and customer concerns. Keep the visual principle you want to test separate from another brand's colors, photography, or identity.

Step five: find gaps to test

Now ask what your selected ads leave unanswered. Maybe they repeat a lifestyle promise but never show setup. Maybe the copy addresses experienced buyers while your customers are still trying to understand the product.

Turn that observation into a hypothesis you can act on:

• Angle gap: "Our customer worries about effort. Let's show the simplest everyday use case."
• Format gap: "The references explain this in captions. Let's try showing the process."
• Copy gap: "The product needs an explanation. Let's compare a direct opening with a short customer story."
• Audience gap: "These ads address experienced users. Let's answer a first-time buyer's question."

An absence from your sample doesn't tell you everything the rival has tried. Make the hypothesis about your own customer and explain why it's worth testing. That keeps the brief useful even if tomorrow's research reveals another ad.

For context, in our sample 29% stayed live for at least 30 days; 9% reached 90 days.

Among ads that had stopped, 25% ended within a week. Most stopped ads therefore lasted more than a week; an early end alone doesn't explain the advertiser's decision.

Turn the ad-strategy research into your own creative brief

I finish by writing the customer problem, the angle, the opening, the proof we can supply, and the action we want someone to take. Attach a reference and explain the specific structure you're borrowing.

For example, you might keep a sequence that introduces an everyday frustration, demonstrates the product, and answers a practical objection. Your version needs your own setting, product evidence, and language. A rival's customer quote or promise doesn't belong in your ad.

Choose the reference in StealAds, then make your version with your brand, product, images, and message. Review the draft against the question you started with. If the visual looks good but the customer concern has disappeared, revise the direction before production.

Test the idea in your own account and keep a note on what you learned. The next research session should help you answer a more specific question. If you're choosing a tool for this process, compare ad spy tools or read StealAds vs Foreplay.

FAQ

What does it mean to reverse engineer a brand's ad strategy?

It means mapping the angles, copy, visuals, and offers visible in its public ads, then turning those observations into hypotheses for your own creative. Keep source notes so another person can follow your reasoning.

Why does run time matter when studying ads?

I use run time to choose references worth inspecting. A long observed run shows persistence, but it doesn't reveal why the advertiser kept the ad or how it performed.

How do you find gaps in competitor advertising?

Compare the customer questions, formats, and proof in your selected ads. Write an unanswered question as a hypothesis for your own test; keep the claim about what your sample shows.

What should an ad research brief include?

Include the customer problem, your angle, an opening, the proof you can support, and the intended action. Attach a reference and explain the structure you want to adapt.

How we counted

Method note: the run-length findings use 10,950 measured public ads across 234 brands. Runs span each ad's start date to its last observation; ads still live have unfinished runs. We can see what ran and for how long, never what a brand paid or what an ad brought in.