The video works. Cost per click is good, the comments are there, the traffic arrives. And the page converts at the same rate as before, sometimes worse.
It is almost always the same cause, and it has nothing to do with the design of the page: a break between what the video promised and what the page shows. The person clicking arrives with a sentence in mind; if they do not find it, they leave.
The page does not start over, it continues
A classic product page is written for someone arriving from a search, that is, for someone who knows nothing. A page receiving video traffic welcomes exactly the opposite: someone who has just spent thirty seconds with a person explaining one precise thing.
Explaining the product to them from scratch is not neutral, it is a loss. It tells them they are in the wrong place, or that what they watched does not commit anyone.
So the rule holds in one sentence: the page picks the promise up where the video left it, in the same words, and moves forward.
This guide covers that page. Where to put a video across the rest of the site, on a product page or in an email, is a separate subject, covered by reusing UGC on your site and in your emails.
Continuity is checked on three points
Vocabulary
If the creator said "hold", the page must not say "retention". If she spoke of "two minutes in the morning", the page must not speak of "a simplified routine". The exact word is what the person retained, and it is the one their eyes look for.
It is the easiest point to fix and the most often missed, because the page was written by someone else, earlier.
The promise
A video promises one thing, rarely more. If it promised the product holds without touch-ups all day, the first screen must answer that, and not present the six qualities of the product of which that is the fourth.
Then there is the image. The person has just seen a face, a setting, a gesture. Finding even one shot from that video at the top of the page confirms they are in the right place, in a fraction of a second and without reading.
That is also the point that most often runs into rights: reusing a still taken from a video on a commercial page is a distinct use from advertising distribution, and it has to sit inside the usage window granted. Checking it before building the page saves having to rebuild it.
What the first screen must carry
Four elements, and the order matters as much as their presence.
- The promise picked up in one line, in the words of the video, not those of the catalogue.
- The product visible, from an angle recognisable by someone who has just seen it in motion.
- The price, because hunting for it is the leading cause of abandonment on this traffic.
- One possible action, and only one, since two competing buttons on that first screen divide attention instead of directing it.
What has no place above the fold: the full menu, the brand's history, a list of six arguments. All of that can exist further down, where someone who has already decided to stay will go and read it.
What the page must carry that the video cannot
This is where continuity stops and the page takes up its own job.
The information owed to the buyer, essential characteristics, price, delivery times, withdrawal terms, sits on the page and nowhere else. A video does not replace it and should not try: it is neither consultable at the right moment nor binding.
Reassurance signals come next, and in a precise order. First whatever settles the doubt the video created, typically size, material or duration. Then reviews, which confirm. Finally return conditions, which reassure whoever has nearly decided.
Inverting that order, that is, opening with return conditions, suggests that returning is likely.
One page per question, not one page for everything
A single page cannot serve five creatives answering five different objections. Each brings someone with a different sentence in mind, and a page trying to pick up all of them picks up none.
The right granularity is one page per dominant promise, not one page per creative. Three pages are nearly always enough: one per major objection in the category.
The usual objection is that multiplying pages complicates maintenance. That is true, and it is the right trade-off to state: if you cannot keep three pages up to date, make one and align every creative on the same promise. What does not work is five promises and one page.
What to measure, and in what order
One measure first: the bounce rate of that page on that traffic, isolated from the rest. Mixed with search traffic it says nothing.
If it is high, the problem is continuity, and it is fixed in the first screen. If it is low and conversion stays weak, the problem is elsewhere on the page, further down, and that is where to look.
Time on page is a poor indicator here, and it is worth knowing: on this traffic, a long dwell often signals someone hunting for information they cannot find.
The cheapest test before rebuilding the page
Before committing to a redesign, one check takes ten minutes. Open the page on a phone, having watched the video just before, and ask whether the first visible line picks up what the video just said. If the answer is no, you already know the fix, and it needs neither mockup nor development.
That test works badly on your own product, because you know what the page means to say. Having someone who does not know the product do it takes the same time and gives a usable answer.
Sources
Checked on 12 September 2026. This guide is not legal advice. Where this guide and the official source disagree, the source prevails.
