Thinking · Newsletter
Landing page vs product page: why your ad traffic is probably going to the wrong one
We read live ad libraries for a living. The most common and most expensive defect we find is not in the creative. It is that every specific ad points at a generic page.
This piece is now part of a longer guide: Landing page vs product page for paid ads: the full guide.
There is a version of this article on most agency blogs, and it is usually a CRO post about button colour and above-the-fold copy. This is not that. This is about a defect you can measure across a whole account in an afternoon.
The defect
Pull any consumer brand's live ad library. Read the promise each ad makes. Then open the page each one points at.
In almost every account we read, the ads are specific and the destinations are generic. One brand we teared down recently was running 168 live creatives across nine destinations, and zero percent of that traffic went to a presell page or a quiz. Every ad opened on a different reason to care. Every page opened on a catalogue.
The click was paid for. The promise was made. Then the page changed the subject.
Product page vs presell page
| Product page | Presell page | |
|---|---|---|
| Assumes | The visitor already wants this product | The visitor wants an answer |
| Opens with | Price, variants, add to cart | The question the ad raised |
| Works when | Traffic is warm, branded or returning | Traffic is cold and ad-led |
| Fails when | The ad made an argument the page never repeats | The offer is simple and the visitor is ready |
| Typical use | Almost all paid traffic, by default | Rare, and usually unbuilt |
The default is not a decision anyone made. It is what happens when the ad team and the site team are different people working from different briefs.
The counter-argument, which is fair
Adding a page adds a tap, and adding a tap is normally bad for conversion. That objection deserves data rather than reassurance.
Published evidence, all third party:
- A presell page versus product page A/B test showed +47% conversion (ConversionWise).
- Presell pages have been associated with +77% ROAS in a Tier 11 case study.
- Advertorial-style presell formats have produced 32 to 40% lifts in documented cases.
There is no published benchmark for presell-to-offer click-through, which means anyone quoting you one is guessing. Make follow-through your own pilot's KPI instead.
You will also see "96% of ad clickers aren't ready to buy" repeated everywhere. We do not use it, because we cannot find a primary source for it.
How to run the audit yourself
This takes an afternoon and costs nothing.
- Export your top 20 creatives by spend.
- For each, write the promise in one sentence. Not the product, the promise. "Your hair isn't falling out, it's thinning at the part."
- Open the destination. Write the first sentence a visitor actually reads.
- Put the two columns side by side.
- Count the rows where they do not match.
Every mismatched row is a paid click landing on a page that does not answer the question the ad raised. That count is your opportunity, and it is usually larger than any creative optimisation available to you.
What to build first
Not nine presell pages. One.
Take the single highest-spending mismatched pair. Build one page from the same brief as that ad, so the promise continues rather than restarts. Run it against the existing destination with the same creative, same budget, same window.
The test is clean because only one thing changed, and the result tells you whether this is worth doing nine more times.
The structural fix
The reason this defect is so common is organisational, not creative. The ad is briefed by one team and the destination is owned by another, and no brief spans both. Fix the brief and the symptom stops recurring.
That is why we produce the ad and its destination from a single brief and test them as one unit. It is a harder thing to promise, because it makes us accountable for what happens after the click.
Questions people ask
Isn't a presell page just an advertorial? Advertorial is one format of presell page. Others include quizzes, comparison pages, listicles and problem-diagnosis pages. The format should follow the promise the ad made.
Does this apply to gaming? The equivalent exists and it is the store listing. An ad promising one experience against a listing showing another produces installs that churn in the first session.
Where does the page have to live? On your domain, not ours. On AppLovin specifically the platform hosts only the end card; the landing page is your own URL and it has to carry the pixel and pass the click parameters through, or you lose the measurement that makes the test worth running.
How do we know it is not just novelty? Run it long enough to clear the learning period, and measure on purchases rather than on clicks.