How to Tear Down a Competitor's Landing Page

Step zero: tear down the right page

Most teardowns fail before they start because they pick a page at random: whatever the first ad in the library linked to. The page behind a three-day-old ad tells you what a competitor is testing, not what works.

The board gives a better filter. Winning ads don't spread evenly across the web; they concentrate on a small number of pages.

Excludes ad servers, link shorteners and marketplaces, where the domain says nothing about the page.

69 domains take 59% of the winning ads on the board. When a brand points 10, 30 or 60 winning ads at the same destination, that page has been tested with a lot of money. Those are the pages to study.

Three signals to look for before you open a page:

• Many ads, one destination. One digestion brand sends 61 of its ads to the same dedicated domain, with a 414-day median run. Nobody keeps feeding a page for a year if it doesn't pay.
• Many Facebook Pages, one destination. A meal-kit domain receives 50 winning ads from 11 different Pages: the brand, creators, a celebrity. Several partners all sending traffic to one page means it converts for everyone's audience.
• Long-running ads pointing at it. Check the days live of the ads, not the page. An ad live for 200+ days into the same page is a page that survived several rounds of testing.

The teardown checklist

Open the page on a phone. Read it once as a buyer, then a second time with this list. Score each point 0 (missing), 1 (present) or 2 (strong).

The handoff (first screen)

1. Does the headline repeat the ad's promise, number or key word?
2. Is the product or the ad's main visual visible without scrolling?
3. Is the offer from the ad (discount, gift, guarantee) on the first screen?
4. Is there one primary button above the fold?
5. Is it obvious within three seconds what the page sells and to whom?
6. If the page looks like an article, is it labeled as advertising near the top?

On point 6: the FTC's native advertising guidance asks that article-style ads be identifiable as ads, with labels such as "Advertisement" where readers look first. Note whether the competitor complies; don't copy a page that doesn't.

The argument

1. What page type is it: product page, dedicated page, listicle, quiz, advertorial? (See DTC landing page examples.)
2. What's the first claim after the headline: a problem, a story, a benefit, a price?
3. Is there a named cause or mechanism ("the real reason X happens")?
4. Where does the product first appear: first screen, a third of the way down, or after the argument?
5. How many distinct sections, and in what order? Write them as a list.
6. Is there a comparison (us vs them, us vs the usual fix)?
7. Who is speaking: brand, founder, expert, customer, "editor"?

The proof

1. What kind of proof is closest to the first claim: a demo, a number, a testimonial, a study, a credential?
2. How many testimonials, and are they specific (names, situations, numbers) or generic?
3. Are there before/after images or demos?
4. Are claims qualified, or absolute? (Absolute health claims are a compliance risk for the advertiser, not a template.)
5. Is there a guarantee, and how long?

The offer

1. What's the price structure: single unit, bundles, subscription, quantity discounts?
2. Which option is visually pushed (badge, "most popular", default selection)?
3. Is the discount applied automatically or behind a code?
4. Is there a free gift or bonus, and how is it framed?
5. Is there urgency, and is it believable (stock, date) or generic (countdown that resets)?
6. Does the price match the ad?

The friction

1. Load time on mobile data, in seconds.
2. Number of redirects or domains between the ad and the page.
3. Pop-ups before reading starts?
4. Navigation menu present or removed?
5. How many taps from first screen to checkout?
6. Is the CTA wording consistent from ad to page to button?

A scoring sheet

Put each page you study in a row. Patterns show up after five or six pages in the same niche.

The point isn't the total. It's the column where your page scores lowest against pages that keep getting traffic.

Read the ads and the page together

A page makes sense only next to the ads that feed it. Tear down two or three ads that point to the page, then check where the page picks up their thread.

Turn the teardown into tests

A teardown is only useful if it ends in a test list. For each gap you found:

1. Write the hypothesis in one line. "Our first screen doesn't repeat the ad's number; adding it will lower cost per purchase."
2. Change one thing per test. Headline, proof placement, offer framing, or page type. Not all four.
3. Send the same ad to both versions. Otherwise you're testing the ad, not the page.
4. Judge on cost per purchase, after enough spend to cover a few purchases per side.

For the specific gaps that cost the most, see the post-click experience and message match. If you find the competitor's page is a presell, our presell pages guide covers when to build one yourself.

Ad Radar shows the landing page type and domain behind every winning ad, and keeps a snapshot of the page with one click to rebuild it for your product. It's a faster starting point than clicking through the Ad Library one ad at a time (features).

Figures marked as estimated spend come from Ad Radar's model of engagement on public Meta Ad Library ads. They are estimates, labeled as such, and are best used to rank ads against each other.