“The account is badly built, can you take it over?” I have heard that sentence often enough to be wary of it.
A case that keeps coming back: an advertiser convinced the campaigns were the problem. I take them over, restructure, tighten the keywords, rewrite the ads. Cost per click drops, clickthrough rate climbs. The quote requests do not move an inch. The account had never been the problem. The page the clicks landed on was not built to convert, and it could not be: it was the site’s product page.
You can have the best campaigns in the world. If the page doesn’t follow, you are paying for nothing.
1. A product page answers a different question
A product page, or a category page, speaks to a visitor already inside the house. They have browsed, compared, they know the brand, they arrived through the menu or through on-site search. The question that page answers is “which one do I pick?”.
The visitor coming from an ad is nowhere near that. They read a three-line promise and they clicked. Their question is “is this really what I just saw?”, and they ask it during the two or three seconds their thumb hesitates between staying and going back. Nothing on a standard product page is written for that question.
This is not a copywriting flaw, it is structural. A product page serves navigation: variants, sizes, related products, full menu, breadcrumb, trust bar, chat window. On an organic journey, each of those is a service. On a paid click, each one is an exit, and you have just bought it at full price.
To clients who hesitate, I ask a very concrete question: how many outbound links sit between the top of your page and your button? On most sites I open, it is somewhere between fifteen and forty. On a page built for advertising, we aim for one.
2. Continuity of the promise, not repetition of the keyword
Plenty of guides boil this down to a recipe: copy the keyword into the page title. That is a short reading, and Google says the opposite itself.
Its documentation on using Quality Score asks you to keep messaging consistent from ad to landing page, and to make sure the page follows through on the offer the ad made. It then adds that the exact wording of the search does not need to appear on the page, with an example to match: nobody expects a headline repeating “South Chicago Chihuahua-friendly budget hotels” word for word.
What has to be continuous is the promise, not the vocabulary. If the ad says “quote within 24 h”, the word quote and the deadline must be readable without scrolling. If it says “free delivery over €50”, the threshold shows up before the form, not at the bottom of the terms and conditions.
The test I run fits in one screenshot: I put the ad and the top of the page side by side, and I ask someone who knows nothing about the project whether the two are talking about the same offer. If they hesitate, the page goes back on the bench.
3. What Google does with your page, and what it doesn’t
Two ideas circulate here. One is overblown, the other is not taken seriously enough.
The overblown one is Quality Score. I am regularly asked to “push it up to bring the cost per click down”. Landing page experience is indeed one of its three components, along with expected clickthrough rate and ad relevance. But Google states in black and white that the score is not a key performance indicator, that it is not an input in the ad auction, and that it should not be optimised. It is a thermometer: it tells you where to look, it does not bring the fever down.
The underrated one is the destination policy. Google disapproves an ad whose page carries pop-ups or interstitials that stop users from seeing the content they asked for, that interferes with the browser back button, or that doesn’t load quickly on the most popular browsers and devices. A warning arrives at least seven days before any account suspension. At that point cost per click is beside the point: serving stops.
Plenty of sites drag along a full-screen consent banner, a newsletter pop-up timed at five seconds and a chat widget that opens by itself. On an organic page, that is a debatable choice. On a page receiving paid clicks, it is a conversion killer plus a disapproval risk.
One corollary catches a lot of people out: a dedicated page often gets pulled from the sitemaps, then blocked in robots.txt so it won’t pollute organic search. The ad gets disapproved right after, because Google requires the destination to stay crawlable by its AdsBot crawlers and names a restrictive robots.txt among the reasons. To keep a page out of search results, use a noindex tag, never a crawl block.
4. The three frictions I remove first
In that order, because they do not cost the same to fix and do not pay back the same.
| Friction | What I do | What it changes |
|---|---|---|
| Page weight | A standalone page, without the full site template or the shop scripts. Hero image sized properly, fonts hosted locally. | The gain lands exactly where you ask for something. |
| Number of fields | Three fields when the call qualifies afterwards, more only when the appointment costs more than the click. | Every field removed is one hesitation less before sending. |
| Exits | Menu, breadcrumb, related products, chat: all gone. What stays is the logo, one call to action repeated, and the legal notice. | The visitor has one thing to do, or not to do. |
On speed, the figure I bring out most often comes from the Milliseconds Make Millions study, run by Deloitte and the agency 55 for Google across 37 European and American brands and more than 30 million sessions. A 0.1 second improvement, repeated on every page of the journey, lifts retail conversions by 8.4 %. And above all, in lead generation, it raises by 21.6 % the number of users reaching the form submission page. That last figure is the one a landing page cares about: the benefit does not spread out, it concentrates on the final step.
On forms, the Baymard Institute measures an average of 11.3 fields in a checkout flow when eight would be enough for most sites, and observes in testing that 42 % of participants type their full name into the “First Name” field. On a landing page I tighten the rule further: every field I keep has to qualify or to recall, the rest gets asked later. I would rather have three fields filling the contact list with a bit of noise than eight fields filling nothing at all. The sorting happens on the phone.
5. When the product page is still the right target
I am not dogmatic about this. Building a dedicated page is sometimes exactly the wrong move.
- Shopping and Performance Max. The product feed points at the product page, that is how the channel works. The job is then to turn that page into a good landing page, with price, availability, delivery time and returns visible straight away, not to replace it. I go into it at length in the article on Performance Max.
- Retargeting on a product already seen. The visitor returns to a page they know. Cutting in with an intermediate page breaks the thread instead of picking it back up.
- Brand traffic. Someone typing your name wants your site, not an airlock.
- No measurement. If conversion tracking isn’t reporting, one more page will only add one more conviction. You fix the thermometer before arguing about the temperature.
The line I draw is simple: a dedicated page when the ad sells a promise, a quote, an appointment, a sign-up, a dated offer. The product page when the ad sells an identified product the visitor already has in mind.
6. Where to start if your campaigns are already running
A reading order, not a shopping list.
- Open the landing pages report in Google Ads and sort it by cost, not by number of clicks.
- Isolate the page that swallowed the most budget without converting. It is almost always a product page or the homepage.
- Load it on a phone over mobile data, and count the windows that open before the first readable word.
- Put the ad and the top of the page side by side, and run the same-offer test again.
- Build a dedicated page for that one ad only, and put it up against the existing one.
The last point matters most, and it is the one people skip. You do not rebuild forty pages. You build one, you watch it live, and you only go at the second one if the first has won. On an account with modest volume, calling the result takes a bit of method: I set mine out in a piece devoted to experimentation.
The reflex to keep
Before adding a euro to a campaign, I look at what happens after the click. A page that converts twice as well is worth a doubled budget, and it costs nothing in media. That is the work I do in conversion rate optimisation, alongside Google Ads and Meta Ads accounts.
The crucible and the vessel
A well-tuned campaign pouring into a product page is ore tipped into a cracked crucible. The refining happened, the material is good, and it drains out through the twenty openings left there for convenience. A landing page is not the ad’s packaging, it is the vessel that decides what you keep. Plug three holes before adding more ore.