Competitor Research for Ad Campaigns: What to Look For, Where to Look, and What to Ignore

Competitor research before an ad campaign has one job: to tell you how your market is already being addressed, so that your ads either say something different or say the same thing better. It is not a way to copy winners. An ad you can see running is evidence that someone chose to run it, not evidence that it is profitable, and the difference between those two matters for every conclusion that follows.

Infinall AI includes competitor research as the first step of its Full Campaign: from a product URL it researches the market, then drafts strategy, copy, and creative, which land in an Approval Queue for you to edit before export. On the Free plan research is Standard only; Pro ("$30/month") runs Standard mode with real-data research and can go up to Deep or Max, with paid web data sources. The Library's Ad Spy lets you browse real competitor ads directly. None of that replaces your own reading of what you find, and it does not tell you what any competitor spends or how their campaigns perform.

This guide describes a method you can follow by hand or use to check a tool's output. The aim is a short document you can act on, not a spreadsheet of screenshots. If your research does not change a decision about your angle, offer, or format, it was not research, it was browsing.

Start with sources that show what is actually running.

Public ad libraries from the major platforms show active ads for a given advertiser, often with start dates. [NEEDS VERIFICATION: current scope and fields of Meta Ad Library, Google Ads Transparency Center, TikTok Commercial Content Library, and LinkedIn Ad Library; availability differs by region and ad category.] Infinall's Ad Spy in the Library is another way to browse real competitor ads.

Then follow the ads to their destinations. The landing page tells you what offer sits behind the creative: free trial, demo, discount, waitlist, lead magnet. An ad's promise and its landing page often differ, and the gap is itself a finding.

Check the places that are not ad libraries: the competitor's homepage headline, pricing page, app store listing, and newsletters they send. These show how they describe themselves when they are not constrained by an ad format.

Finally, look at who else solves the same problem differently. Your closest competitor in search is not always your real competitor in the customer's mind; a spreadsheet, an agency, or doing nothing may be the actual alternative. Include one or two of these so your research does not assume the category boundaries your competitors drew.

Limit the set. Five to eight competitors, chosen to include direct, adjacent, and one indirect alternative, is usually enough to see patterns without drowning.

For each competitor ad, resist saving it as a screenshot and moving on. Record five things.

Angle: the one idea the ad leads with. Is it speed, cost, a pain, a status claim, a comparison? Name it in a few words.

Offer: what you are asked to do. Start a trial, book a demo, buy now, download something. Note any incentive.

Proof: what the ad uses to be believed. Customer quotes, numbers, logos, product screenshots, a demonstration. Note whether claims are specific or vague. Do not assume a number in a competitor's ad is true; it is their claim.

Format: static image, carousel, short video, long video, text-heavy, UI-heavy. Note the platform.

Longevity: how long has the ad been running? A long-running ad suggests the advertiser kept paying for it, which is weak evidence it was worth keeping, but it is only a hint, since some advertisers leave ads running without review.

After a dozen ads, patterns appear: most competitors lead with the same angle, nobody shows the product in use, everyone offers a free trial. Those patterns are the output. The point of the exercise is to identify what is crowded and what is empty.

A positioning gap is an angle, audience, or proof type that customers care about and competitors are not using. Finding one is a judgment call, not a calculation, and it is worth being honest about the uncertainty: an empty angle may be empty because it does not work.

Use three tests on any candidate gap. First, is it true for your product? An angle you cannot support in the product is a liability, not a gap. Second, does it matter to the buyer? Check this against customer conversations, support tickets, and reviews rather than assuming. Third, can you say it in a way that is visibly different from what is already running? If your version reads like theirs, there is no gap in practice.

Reviews of competitors (on app stores, review sites, or social posts) are an underused source here. Recurring complaints are candidate angles: if customers repeatedly say a competitor's setup is slow, that is a claim you might make, if it is true for you.

Write the result as one paragraph: who we are speaking to, what they are choosing between, what we will say that competitors are not, and what we can show to back it up. Infinall's strategy step produces a draft of this kind of plan from research, which you should check against what you know. Then test it with a small number of ads rather than committing the whole budget.

Be clear about the limits so the research is not over-trusted.

You cannot see a competitor's spend, conversion rate, cost per acquisition, or audience targeting from public ad libraries. Third-party estimates exist, and their accuracy varies; treat them as rough at best. [NEEDS VERIFICATION: accuracy of third-party ad spend estimators.]

You cannot tell whether an ad is profitable. A brand with a large budget may run many mediocre ads. Copying their creative copies their mistakes along with their wins.

You cannot see what is being tested privately, in unpublished audiences, or through channels with no public library.

You also cannot see context: a competitor's ad that works for them may rely on a brand name, a large existing audience, or a price position you do not share.

For these reasons, competitor research should narrow your first tests, not substitute for them. Take two or three angles that look differentiated, produce a small set of ads for each, and run them in the ad platform yourself with a budget you can afford to learn from. Infinall prepares the research, strategy, and creative drafts, held in an Approval Queue for your edits before export; it does not buy media or measure results, so the learning still happens in your own accounts.