Check, ad by ad, that the first screen of the page repeats what the ad showed: the same product, the same promise, the same price. If people have to hunt for the connection, they're already gone.
The benchmark
At a store the authors work with, only 1.4% of visits from Facebook reached a product page.
They read this as a break between what the ad promised and the page it led to, and that's often how the problem shows up: as a low number nobody links to the ad.
To see where you stand, build a table for your ten biggest-spending ads. The Meta Ad Library shows every active ad on a page exactly as the public sees it, so you don't have to wait for an export.
Then open each link on your phone and note what fits in the first screen.
In a recent audit for a clothing brand, the first ad the authors reviewed said "discover our jackets" and showed an accessory. The click did lead to that accessory's product page, but the copy had promised something else.
Copy this table, then fill it in with your own ads:
| Ad (screenshot) | Landing page (URL) | First screen on mobile | Gap spotted |
|---|---|---|---|
| Visual, hook, price or offer | Full link, parameters included | Image, title and price visible without scrolling | Product, promise, price, variant or stock |
Once the table is filled in, you can tell leadership:
"Of our ten highest-spending ads, this many lead to a page that repeats neither the visual nor the price. I propose that the advertising team and the site team review this table together before every new batch of creatives."
The rule
The first-screen test: if someone sees only the top of the page on their phone, do they recognize the ad they just clicked?
The steps
- The visual. Reuse the ad's main image at the top of the page.
- The promise. Repeat it in the headline, in the same words.
- The price. Show the advertised offer, already applied. On Shopify, a discount link in the form `/discount/CODE?redirect=/products/product-name` applies the code to the cart, but the price shown on the product page depends on the theme: check that the discounted price actually appears.
- The variant. Preselect the color or model shown by adding `?variant=` followed by the variant ID to the product page address.
- The stock. Check availability before you ramp the budget back up.
The trap
The first trap is about pace. Ad creatives often change every week, while the landing page can stay unchanged for months, and after a few batches the ad touts an argument the page no longer mentions.
Nobody made a mistake, though. The team or agency running the ads reads its results in Meta or Google, while the site team reads its own in Shopify, and the gap between the two falls under no one's remit.
The authors run into this boundary too: when they prepare a landing page test for a brand, they have to build it with the people who run its advertising and read the result in the ad platforms, because their testing tool can't see the ad.
The second trap is even easier to miss. A brand the authors work with wanted to automate the removal of sold-out products, and the rule fit in one sentence: "if the warehouse says there's no more stock, archive the product."
On the site side, the setting makes sense. But on Shopify, an archived product is no longer displayed in the store, so an ad that still points to its product page sends every paid click to a page that can't be found.
The authors had discussed it with the brand in terms of display and back-in-stock alerts, and one line was missing from that list: pause or redirect the ads for the affected products before automating anything.
