Why Is My Website Not Converting Visitors Into Calls?

Start by Checking That You Are Counting Right

Before you rewrite anything, prove the number is real. A surprising share of quiet websites are not quiet at all. They are miscounted, and the owner spends three months fixing a page that was already working. Take one week and reconcile by hand. Open your analytics, note every conversion it recorded for those seven days, then compare that list against your real inbox and your real phone log. Three things usually turn up. Some sites count double, because a form event fires on submit and the thank you page counts again. Some count nothing on calls, because the phone number is plain text with no tap event behind it, so every call from a phone is invisible. And some count junk, where spam form fills and eight second wrong numbers sit in the data next to real customers. Each of these sends you in a different wrong direction. Double counting makes a failing page look fine. Untracked calls make a working page look dead. Junk makes a bad channel look great. While you are in there, send yourself a test form from a phone on cell data, not office wifi, and time how long it takes to arrive. If it never arrives, you have a delivery problem rather than a conversion problem, and that is a different repair job.

Your Site Average Hides the Real Problem

One number for the whole site is almost always the wrong number. It is an average of groups that behave nothing alike, and averages hide the half that is broken. Split it three ways and the problem usually names itself. Split by device first. Here is a hypothetical that shows up constantly: a site sits at 2.4 percent overall, which looks unremarkable. Split it and desktop is at five percent while phones are at 1.2 percent. Since phones bring most of the traffic for local service work, that gap is the entire problem, and no amount of desktop polish touches it. Split by channel second. Traffic from a map listing is people ready to call. Traffic from a blog post about how much a repair costs is people still reading. Mixing them produces a number that describes nobody. Split by landing page third. Most sites have one or two pages carrying most of the traffic and one page quietly converting well. Find the page with real visits and a poor rate. That is where an afternoon of work is worth the most. If you only ever look at the site average, you can ship five good changes and watch the number sit still, because the failing group is big enough to swallow all of them.

Is It the Traffic or Is It the Page?

Not every conversion problem lives on the website. Sometimes the visitors were never going to call, and no page change fixes that. Check the searches that brought them. If you run ads, the search terms report shows the actual queries that triggered them, and it is where the answer usually hides. Broad match sends a drain cleaning company people looking for drain cleaning jobs, drain cleaner products, and how to clear a drain themselves. All three click. None of them hire you. For search traffic, look at the split between people searching your business name and everyone else. Brand searches convert well because those people already decided. If your growth came entirely from brand searches, the rate goes up without the marketing doing much. If it came from wide informational queries, the rate falls even though nothing got worse. Check geography too. A page that ranks in a city you do not serve produces visits that can never become jobs. This matters most after any big traffic jump. Doubling visits with people who are three counties away or comparing prices for a service you do not offer will drop your rate every time, and the site did not change at all.

Most Lost Calls Die on the Mobile Contact Path

If the counting is right and phones are the weak group, walk the path yourself. Pick up your own phone, turn off wifi, and search for your service the way a customer would. Then count the taps it takes to reach a human. Four things usually go wrong. The phone number is not tappable, because it lives inside a header image or was typed as plain text. The number is there but sits below three screens of content, so nobody on a phone reaches it. A cookie banner or a chat bubble covers the button on smaller screens, which nobody notices on a desktop monitor. Or the page takes so long to load on a normal cell connection that people leave before the button ever appears. Speed matters here more than people expect, and Google measures it from real visits rather than lab tests. Time each step while you do this. If finding the number takes more than a few seconds, or the tap target is small enough to miss, you have your first fix and it needs no test. Then check what happens on the call. A tracking number that misroutes, a full voicemail box, or an after hours message with no way to leave a callback all look identical to a bad website in your reports.

When the Page Answers None of the Questions Callers Ask

Visitors leave when a page will not answer the questions they came with. The fastest way to learn those questions is to stop guessing and ask the person who answers your phone. Every trade has three or four that come up on nearly every call. What does this usually run. Do you come out to my town. How fast can somebody get here. Are you licensed and insured. If your page answers none of them, the visitor leaves to find a site that does, and the site they find gets the call. A price signal is the one people fight hardest about. You do not have to post a price list, and for most trades you should not. A range, a minimum, a service call fee, or a plain sentence about what drives the number all do the job. The point is that the visitor stops wondering. Service area is the same. Naming towns beats writing that you serve the greater area, because a customer scans for their own town and nothing else. Timing matters just as much. Someone with water on the floor is not comparing craftsmanship. They want to know who can come today. Say it near the top, on the pages that carry the traffic, in the words your callers actually use.

The Boring Breakages Worth Ruling Out

Before anyone rewrites a page, rule out the dull failures. They cost more leads than any headline ever will, and each takes minutes to check. Do it from a phone, on cell data, as an outsider rather than the owner. The list is short and it repeats across almost every site.

• A form that rejects a phone number typed with dashes or spaces
• A required field the visitor cannot answer, like a full street address
• An old phone number still hiding on one page or in the footer
• A contact button that points at a page that no longer exists
• A form that shows no confirmation, so people submit it twice or give up
• A chat widget or banner covering the call button on small screens
• Notification emails landing in a folder nobody has opened since spring

Fix in This Order, and Give Each One 30 Days

Order matters because each step decides whether the next one is worth doing. First, fix the counting. Until leads are counted once and calls are tracked, every later decision is a guess dressed up as data. Second, split the numbers by device, channel, and landing page, and pick the single worst group with real traffic behind it. Third, fix the breakages on that group. These are not tests. A broken form or a buried phone number gets repaired the same day you find it. Fourth, answer the three questions your callers ask, on the two or three pages that carry your traffic. Fifth, and only now, start changing things that are choices rather than faults, like which offer leads or where the reviews sit. Then wait. Give each change 30 days on a normal site and longer on a slow one, and compare against the same stretch last year if your trade swings with the season. If the number drops, undo it and write down what you learned. Nobody can promise a specific result from any of this, and be suspicious of anyone who does. What this order does reliably is stop you spending three months on the fifth step while the first one is still broken.

Frequently asked questions

Sources

• Google Ads Help: About the search terms report(opens in a new tab) The search terms report shows the actual searches that triggered your ads, which is how you tell a traffic problem from a page problem.
• Google Search Central: Understanding Core Web Vitals and Google Search results(opens in a new tab) Core Web Vitals are measured from real user visits, which is why page speed should be judged on a phone connection rather than a lab test.
• Microsoft Clarity(opens in a new tab) Free session recordings let you watch the visits that did not convert, which is the evidence behind the ten recording rule.

More on this topic