Click to call on mobile and the businesses still missing it

This is the least interesting fix in conversion optimization and one of the most reliably profitable. It takes minutes, it costs nothing, and a shocking number of local business sites still get it wrong somewhere.

The reason it survives is that owners test their own sites on a laptop. On a desktop browser a phone number that is not linked looks completely fine. It reads correctly, it is the right number, nothing appears broken. The failure only exists on a phone, which is where most of your local traffic actually is.

Why does a phone number need to be a link?

Because the alternative is asking someone to do work at the exact moment they were ready to act.

Picture the visitor. They are on a phone, they have decided to call you, and the number is text. Now they have to hold ten digits in their head, switch to the dialer, type them without a mistake, and hope they remembered the area code. If they get interrupted halfway, you lost the call and you will never know it happened.

A tappable number removes all of that. One thumb, one tap, ringing.

Three versions of this break most often. Numbers inside a header image or a logo graphic, which cannot be tapped at all. Numbers styled as text in a page builder that never wrapped them in a link. And numbers in the footer that are linked while the one in the header, the one people actually see first, is not.

On a phone, an unlinked number is not a phone number. It is a picture of one.

Why does this matter more in 2026 than it did?

Because the call buttons you did not own are disappearing, and the one on your own site is the one that is left.

Sterling Sky's field data for 2026 documents a decline in clicks to call from Google Business Profiles even for profiles whose rankings held steady. The underlying dataset, compiled with Jepto, covers 179 profiles across 34 US law firms over two years. Website clicks did not fall the same way, because website icons show on desktop while call icons dominate mobile. The loss is mobile specific and it is the call button.

The same report describes AI powered local packs, appearing on roughly 7% of tracked keywords at the time of writing, mobile only and US only. They show one or two businesses instead of three, and they have no call buttons at all.

That matters because of where local calls originate. Patient Prism's dental access report, covering 8,280 locations and 11,552,668 patient calls in 2025, attributes 90% of attributable patient calls to the Google ecosystem, with the Google Business Profile alone at 54% and organic plus Google Ads at another 36%. Meta and print combined came to under 3%. That is dental, it is the customer base of a call intelligence vendor, and it therefore skews toward more sophisticated practices. Directionally it is still the clearest picture available of where the phone rings from.

Read those two findings together. Most of your calls come from Google surfaces, and the fastest call path on those surfaces is quietly narrowing. The tappable number on your own page is not a nice extra. It is the part of the path you still control.

Where should the phone number appear?

More places than most sites use, and always in the same format.

Top of every page, visible without scrolling. This is the one that matters most, because a visitor who arrived ready to call should never have to hunt.

Beside every contact form. Some people will never type into a form and are waiting for permission to call instead. Give them both options at the same height and let them choose.

At the end of long content. If someone read a whole service page, that is a strong buying signal, and they should not have to scroll back up. This is one of the practical reasons long pages can outperform short ones when they are built with repeated exits.

In the footer, because that is where people look by habit.

And on any page you send paid traffic to. A dedicated page built for a campaign should carry the number prominently, since a landing page has a different job from your homepage and its job is a single action.

Use the same number format everywhere, and make the button text say what happens. A call to action people act on says "Call (555) 555-0100" rather than "Contact."

How do I know if mine is broken?

Check it on an actual phone, not a resized browser window. This takes four minutes.

Open your site on your own phone. Go to the homepage, a service page, the contact page, and a blog post if you have one. Tap every visible phone number. Each one should open the dialer with the correct number already filled in.

Then check what happens on desktop. A tel link on a laptop may open a calling app or do nothing at all, and that is fine. It should not produce an error page.

Watch for two subtle failures. A number that dials the wrong extension, which happens after an office move and nobody notices for months. And a tracking number that stopped working when a call tracking subscription lapsed, which silently sends every caller to a dead line.

That second one is the case for a real maintenance habit, since website maintenance is mostly catching things that broke quietly rather than anything visible.

Should I track calls as conversions?

Yes, and track two different things, because they are not the same.

The tap is a website event. You can fire an analytics event when someone taps the number, which tells you the page is doing its job.

The connected call is the business outcome. A tap that rings out unanswered is not a lead. If you only measure taps, your site will look like it is performing while your inquiry count stays flat.

The gap between those two numbers is not small, and it varies enormously by trade. CallRail's benchmark analysis of 1.1 million de-identified conversations, published January 2025, put missed call rates at 32% in health care, 28% in legal, 14% in home services and 9% in real estate. CallRail defines a missed call as an unanswered inbound call that goes to voicemail or is detected as abandoned. It sells call tracking, and its sample is businesses that already bought call tracking, so if anything those numbers flatter the average.

Call tracking numbers give you the second half, along with which source produced the call. Be careful with them for local search reasons, and keep your main published number consistent with what appears on your Google Business Profile.

If you are running paid campaigns, this matters more, because call volume rises and falls with spend. When budget pacing runs the account dry mid-month, calls stop, and without call tracking that looks like a website problem rather than a budget problem.

What is a working tel link actually worth?

Run it once with your own numbers. Five lines, and every input comes from a report you already have.

Start with the tap count. Say your analytics shows 54 taps on the number last month across all pages.

Compare it to connected calls. Your phone system or call tracking says 41 of those connected. Thirteen rang out.

Apply a booking rate. ServiceTitan's platform data for June 2022 puts the typical home services shop at a 42% call booking rate, with fewer than five technicians at 24% and 25 or more technicians at 59%. Use the band that matches your size. At 42%, those 41 connected calls booked about 17 jobs.

Price the 13 that rang out. At the same 42%, they were worth about five and a half jobs. At a $480 average job that is roughly $2,600 a month, or about $31,000 a year, lost after the visitor had already decided to call you.

Then price the fix. ServiceTitan's own analysis of the same data set puts every 5% improvement in booking rate at roughly $100,000 in additional revenue for a shop with 5 to 14 technicians, and notes it is achievable with less than one additional booked call per weekday.

Two honest notes on that arithmetic. ServiceTitan sells field service software and the data is its own platform, so it is vendor research, well sampled but not neutral. And the 42% is a booking rate on answered calls, not an answer rate; vendors routinely mislabel one as the other, and a 42% booking rate is not a 58% miss rate.

The gradient is the real finding. A shop under five technicians books 24% of calls and a shop over 25 books 59%. That is not a discipline gap. It is a capacity gap, because the small operator is under a sink when the phone rings.

One number to refuse if it appears in a pitch. Vendors selling call answering advertise booking rates of 90% or 92%. The 92% figure most often quoted is footnoted on the seller's own site as having been provided by a customer and presented as received, without modification. That is one self-reported number from one business, not a benchmark, and it is not a rate you should be measured against.

What to do this week

Open your site on a phone and tap every number you can find. Fix any that do not dial. Most of these are a two minute change in whatever page builder you use.

Add the number to the top of every page if it is only on the contact page. That single change is often the largest conversion gain available to a service business.

Check that the number on the site matches the number on your Google Business Profile and your invoices. Mismatches confuse customers and cost you elsewhere.

Run the five line calculation above, even roughly. It converts an invisible problem into a monthly dollar figure, which is the only version anyone acts on.

Then answer the honest question underneath all of this. If more people start calling, who picks up? Getting the call and taking the call are different problems, and fixing the first without the second just moves the loss one step later.

Questions about your phone setup?

Email me at eric@seod.com with your website address. I will open it on a phone, tap every number on your main pages, and send you back a short list of exactly which ones dial and which ones do not, page by page.

It takes me about five minutes and there is no call attached. If everything works, the reply will say so in one line and I will point at what I would look at next instead.

Or keep reading the conversion library.