Why Your Ads Convert Badly (It's Usually the Landing Page)

You have paid for the click by the time anyone sees your page. Everything before that point is about winning attention. Everything after it decides whether the money did anything. Yet the page is usually the last thing anybody touches, because campaign settings are quick and pages are work.

First, work out whose fault it is

Before rebuilding anything, establish whether you have a traffic problem or a page problem. They look identical in a monthly report and need opposite fixes.

Look at click-through rate and conversion rate side by side, per ad group.

• Healthy click-through rate, poor conversion rate. People find your ad appealing and your page disappointing. That is a page problem, and it is the good kind because you control it entirely.
• Poor click-through rate and poor conversion rate. You are probably reaching the wrong searches. Fix targeting before you touch the page — a better page shown to the wrong audience is still the wrong audience.
• Good on both, still no business. The problem has moved past the website to what happens when an enquiry arrives. More on that at the end.

Check the numbers by device separately while you are there. A page that performs adequately on desktop and badly on mobile will average out to mediocre, and for most local service searches the mobile figure is the one that matters.

The homepage is not a landing page

Sending paid traffic to your homepage is the most common and most expensive default.

A homepage answers “who is this company”. Somebody who searched for a specific service, in a specific town, has already asked a much narrower question and paid attention for about three seconds. Making them navigate to the answer loses a proportion of them at every step, and you paid for all of them.

The rule is one page per intent, not one page per keyword. If three keywords describe the same job and would be answered by the same page, they share a page. If a keyword implies a different job, a different urgency or a different buyer, it needs its own.

What the first screen has to do

Before any scrolling, a visitor should be able to answer four questions without effort.

Am I in the right place? The headline should echo the language of the ad, which should echo the language of the search. If someone searched for emergency drain unblocking and your headline says “Comprehensive Property Solutions”, they are gone. This is the single highest-return change on most pages and it costs nothing.

Do you cover me? For anything local, name the areas. Explicitly. A visitor should not have to scroll to a footer to find out whether you come to their town.

Why you? Two or three concrete points. Not “quality service” — a real qualification, a real guarantee of response time, the accreditation your trade actually recognises, how long you have been doing it.

What do I do now? One primary action, repeated. If phone calls close better than forms — and for urgent trades they usually do — the number should be large, tappable and above the fold, not tucked in a header.

Everything else on the page supports those four. Anything that does not is competing with them.

The specifics that move the number

Form length. Every field is a chance to leave. Ask for what you need to respond usefully and qualify the lead, and nothing more. Name, contact detail, one line about the job is often enough. Postcode earns its place for anyone with a service area. Fields that exist because a CRM wants them should be filled in afterwards by you, not by the customer.

Speed. Google’s page experience signals and plain human patience point the same way. A page carrying a slideshow, three tracking scripts and a chat widget it does not need will lose people before it renders on a phone on mobile data.

Proof that survives scrutiny. Reviews with a name and a source beat anonymous quotes in a carousel. Photographs of your own work beat stock imagery of people in headsets. If you have a company number, an address and a trade body membership, show them — those are the things a cautious buyer checks.

Price signals. Counterintuitive but reliable: giving an indication of cost, even a “from” figure or a typical range, usually reduces the number of enquiries and improves the number worth having. If you are drowning in unqualified leads, this is the lever.

Match the page to the funnel stage. Somebody searching a comparison question is not ready for a quote form. A page built for that search should teach first and ask second, or you are pushing for a commitment the visitor has not reached.

Make sure you can see what happened

None of this is improvable if the measurement is wrong. Three things to verify:

1. The conversion fires on the right event. A confirmed submission or a thank-you page view — not a click on the submit button, which counts failed validations as successes.
2. Calls are tracked, if calls are how people contact you. A page whose main action is a phone call and whose reporting only counts form fills will always look broken.
3. One person is not two conversions. Someone who calls and then fills in the form is one enquiry. Check the count settings so bidding is not chasing a doubled number.

Set this up properly once and the rest of the work becomes measurable. Set it up badly and you will be arguing about opinions. Analytics is the least glamorous part of this and the part everything else rests on.

Testing, honestly

Split testing needs volume. If a page produces a handful of enquiries a month, no test will reach a conclusion before the market has moved on, and running one anyway will just give you a result you believe for no good reason.

At low volume, do not test. Apply what is well established — message match, one clear action, proof, speed, a short form — and build the best single page you can. Save the testing for when you have enough traffic that a difference can be seen.

When it is not the page

Sometimes the page is fine.

If the searches you are buying are informational, no page turns them into buyers this week. If your price is genuinely uncompetitive, a better page just gets you more people who then leave. If your service area on the page is wider than the area you will actually travel to, you are collecting enquiries you will decline.

And the most common one of all: nobody rings back quickly enough. Somebody who searched, clicked and filled in a form is contacting several businesses in the same ten minutes. The one who answers first has an advantage no landing page can compensate for.

Fix the page, certainly. Then check the thing that happens after it.