Why Your Contractor Website Gets Traffic But No Calls (2026 Diagnostic)

You’ve already done the part most contractors never get to. You’re being found. Somebody is typing your trade and your city into a phone and landing on your site, hundreds of times a month, and the analytics dashboard looks like something is working. Then you look at the call log and the booked jobs and none of it lines up.

The advice you’ll get at this point is almost always “you need a redesign,” usually from someone who sells redesigns. Occasionally it’s “you need more traffic,” from someone who sells traffic. Both answers skip the diagnosis entirely, and both cost between five and twenty thousand dollars to be wrong.

Here’s the structural reason the gap exists. Ranking and converting are two different systems that happen to live on the same domain. A site can rank because it has pages, links and a Business Profile pointing at it — none of which require the site to load fast, work with one thumb, or tell a homeowner in the first three seconds that you handle their exact problem in their exact town. Nothing about getting found forces a site to be persuasive. So most contractor sites are competent at the first job and accidental at the second.

What follows is a diagnostic for website traffic that never turns into calls, not an education. Symptom, likely cause, how to check it yourself in ten minutes without buying anything, and what the fix actually is. Along the way we’ll retire the single most-repeated statistic in web design, because it’s misattributed and you’ve almost certainly been sold with it.

What’s in this guide

1. First: prove the calls are actually missing
2. The diagnostic table
3. Speed — and the fake statistic everyone quotes
4. Your site is a phone site, whatever it looks like on your desk
5. The first screen decides the call
6. Trust signals a homeowner actually checks
7. One action, repeated, with nothing in front of it
8. “Looks good” and “converts” are different jobs
9. Questions contractors actually ask

First: prove the calls are actually missing

Before you touch the website, rule out the possibility that the leads are arriving and dying somewhere else. This takes an afternoon and it saves a rebuild more often than anyone in our industry likes to admit.

Invoca’s 2026 home services benchmarks, built from more than 70 million analysed calls, found that only 52% of inbound home-services calls are answered by a person. Of the ones that are answered, 45% convert to a booked job on that call — and 55% of businesses never actually ask the caller to book. So roughly half the calls a site generates may never reach a human in your shop at all, and a good share of the ones that do end without anyone asking for the work.

Second check: look at what the traffic actually is. If most of your sessions come from one blog post about “how often to change an air filter,” you don’t have a conversion problem, you have an intent problem. Homeowners reading maintenance tips were never going to call. Sort your top landing pages by sessions and mark each one as buying, researching, or neither. Judge your conversion rate only against the buying pages.

The diagnostic table

Find your symptom. Most sites have two or three of these at once, and they compound — a slow site with a weak first screen doesn’t lose the sum of both problems, it loses almost everyone.

Symptom → cause → how to check → fix

What you’re seeing	Likely cause	How to check it yourself	The fix
High traffic, very short average session	Page is slow, or the first screen doesn’t match what they searched for	Run your top three pages through PageSpeed Insights on mobile. Then load the page on your own phone on cellular data, not shop wifi.	Cut image weight and third-party scripts first; they are almost always the bulk of it.
People reach the contact page and leave	Form friction — too many fields, or it asks for something they can’t answer yet	Count the required fields. Try to complete it yourself on a phone, one-handed, without autofill.	Five fields maximum. Name, phone, ZIP, service needed, free-text box.
Calls come in but they’re price shoppers and tyre-kickers	The site never qualifies — no service area, no minimum, no indication of the work you want	Read your homepage as a stranger. Can you tell what this company does not do?	Name your service area, your trades, and the job sizes you want. Repelling wrong-fit leads is a feature.
Desktop converts, mobile doesn’t	Tap targets, sticky headers eating the screen, forms that trigger the wrong keyboard	Complete your own quote form on an actual phone. Time it.	Phone number tappable in the header, form fields with the correct input types, nothing under 44px.
Good rankings, good traffic, almost no form fills	No trust signals above the fold — license, reviews, local proof	Screenshot the first screen on a phone. Show it to someone who doesn’t know you and ask who they think you are.	Star rating and review count, license number, city names, real job photos. Not stock imagery.
Form fills arrive but never answer the phone back	Response lag — you’re calling back hours later	Check timestamps: form submitted vs first outbound call.	Same-day is the floor; inside 15 minutes is where the booking rate lives.
Emergency searches don’t convert	No urgency path — nothing says you can be there today	Search your trade plus “emergency” plus your city on a phone. Land on your own page. Could you get help in one tap?	A sticky call bar, stated response window, and after-hours coverage that is actually staffed.
Traffic is up, calls flat, nothing else changed	The new traffic is informational, not commercial	Segment landing pages by intent before comparing anything.	Build and promote the money pages, not more tips posts.

Speed — and the fake statistic everyone quotes

You have been told, probably more than once, that “53% of mobile users abandon a site that takes more than three seconds to load.” It’s in nearly every web design pitch deck in the country. It is also wrong, and knowing why gives you a useful filter for everyone who pitches you.

The 53% figure comes from Google’s own mobile page speed research — but in the original it refers to something else entirely. The number that gets reported as an abandonment-after-three-seconds statistic traces back to a separate 2016 DoubleClick study, a decade old now, and the “53%” that people cite from Think with Google’s mobile speed benchmarks is attached to page weight, not load time. Two different studies, welded together by repetition until nobody checks.

Here’s what that same Google research actually says, which is more useful anyway:

As page load time goes from one second to 10 seconds, the probability of a mobile site visitor bouncing increases 123%.
Think with Google — Mobile Page Speed Benchmarks

That’s a curve, not a cliff. There is no magic three-second threshold where customers evaporate; there is a steady, compounding penalty, and every second you shave pays you back. The same research found the average mobile landing page takes about 15 seconds to fully load, and that 70% of the pages analysed took more than five seconds just to show the content above the fold. Your competitors are slow too. That’s the opportunity.

Bounce probability rises with every second

Indexed to a one-second load. Google reports a 123% increase in bounce probability between one and ten seconds; the intermediate shape is illustrative.

The other number from that research deserves a place on your wall, because it explains why “just add a slider and a chat widget and a review carousel” is a bad idea. As the number of elements on a page — text, titles, images — goes from 400 to 6,000, the probability of conversion drops 95%. Every plugin someone talks you into is weight, and weight is measurable in booked jobs.

Numbers you’ll hear in a website pitch, graded

The claim	Status	What’s actually true
“53% of mobile users abandon sites that take over 3 seconds”	Misattributed	Conflates a 2016 DoubleClick study with a Google figure about page size. Use the 123% bounce-probability curve instead.
“94% of first impressions come from website design”	Unsourced	Traces to garbled retellings of old Stanford and Northumbria credibility research. No paper says this.
“75% of people judge credibility by your website”	Unsourced	No primary source. Design does affect trust — see Nielsen Norman Group — but not with that number.
“Over 80% of local home-service searches are on mobile”	No attribution	Mobile was 51.48% of global web traffic in Q2 2026 per Statista. Local trade queries skew higher, but nobody has published the number.
“Local service websites convert at 3.1%”	Secondhand	Circulated without a methodology. Benchmark against your own buying-intent pages instead.
“Bounce probability rises 123% from 1 to 10 seconds”	Verified	Published directly by Think with Google in its mobile page speed benchmarks.

Practically: measure with Core Web Vitals, on mobile, on the pages that actually receive traffic. Largest Contentful Paint is the one that correlates with people leaving. Uncompressed hero images and a stack of tracking scripts cause most of it, and both are cheap to fix compared with a rebuild.

Your site is a phone site, whatever it looks like on your desk

Statista put mobile devices, excluding tablets, at 51.48% of global website traffic in Q2 2026. That’s the honest global figure and it’s the one we’ll use, because the “80% of home service searches are mobile” line that circulates in contractor marketing has no source behind it at all. What we can say with confidence is that emergency trade searches skew heavily toward phones for an obvious reason: nobody walks to a desktop computer when a water heater lets go.

Housecall Pro’s 2026 spending report, based on 1,100-plus US homeowners, found that 58% contact a pro because of a sudden problem demanding immediate attention. That’s your traffic. Someone standing in a utility room in Katy, phone in one hand, flashlight in the other. Design for that person or design for nobody.

Three mobile failures account for most lost calls on otherwise decent sites. A phone number that isn’t a tappable tel: link, so it does nothing when pressed. A form that opens a full alphabetic keyboard for a phone number field. And a header that stacks a logo, a menu, an announcement bar and a cookie banner until the actual content starts below the fold. Every one of those is a half-hour fix that no one has been asked to do.

The first screen decides the call

A homeowner arriving from a search result has one question: can these people fix my thing, where I live, soon? Your first screen either answers that or it doesn’t. Most contractor sites open with a wide photo of a truck, the word “Excellence,” and a menu.

Most contractors stop at “our homepage looks professional.” That is not analysis. The test is whether a stranger can answer four things inside three seconds without scrolling: what trade you are, which towns you cover, how fast you come out, and how to reach you right now.

What earns space on the first screen of a phone:

• Trade and location in the headline. “Emergency AC repair in New Braunfels and Comal County,” not “Comfort You Can Count On.”
• One primary action, as a tappable phone button, with a secondary text or form option under it.
• A response promise you keep. “Answering now, most calls dispatched same day” beats “24/7” if 24/7 isn’t literally true.
• Your star rating and review count, as text, near the button.
• Your Texas license number if your trade is licensed.

And the thing that earns its place more than any of the above: one real photo of your actual crew or truck, in a recognisable local setting. Stock photography of a smiling man in a hard hat is read, correctly, as “this company might not exist.” The Nielsen Norman Group’s work on trustworthiness in web design comes back to the same principle repeatedly: credibility is built from evidence of a real operation, not from polish.

Trust signals a homeowner actually checks

Reviews are the first thing they check, and they check them off your site as often as on it. BrightLocal’s 2026 Local Consumer Review Survey of 1,002 US adults found 97% read reviews for local businesses and 41% always do. The bar has risen sharply: 31% will only use a business rated 4.5 stars or higher — up from 17% the year before — 68% require at least four stars, and 47% won’t use a business with fewer than 20 reviews.

Recency matters as much as the average. 74% want to see reviews from the last three months, and 89% expect owners to respond, with 80% saying they’re more likely to use a business that replies to all of them. A 4.9-star average whose most recent review is eighteen months old reads as a business that stopped. If that’s you, the fix is a steady review programme rather than anything on the website at all.

Two more signals that move a decision more than most contractors expect. Housecall Pro found 62% of homeowners are more likely to move forward when financing or payment plans are available, and 77% say rising costs are pushing them to delay or shrink a project. If you offer financing and it isn’t visible until the invoice, you’re losing jobs to someone who put it on the page. And 72% would pay more to have an emergency resolved within 24 hours — which means a stated response window is a pricing advantage, not just a courtesy.

One action, repeated, with nothing in front of it

The most common conversion problem we find on contractor sites is not a weak call to action. It’s four competing ones: call, book online, request a quote, chat, join the maintenance club, download the guide. Every added choice costs you a decision, and a stressed homeowner resolves a hard decision by leaving.

Pick the action you actually want. For most trades in most Texas markets, that’s a phone call, because Invoca’s benchmark data puts the on-call booking rate at 45% — no form can do that. Make the phone button the primary action on every page, keep a five-field form as the secondary path for people who can’t talk, and delete the rest.

What belongs in a contractor quote form — and what doesn’t

Field	Keep it?	Why
Name	Keep	One field, not first and last separately.
Phone	Keep	Set the numeric input type so the right keyboard opens.
ZIP code	Keep	Qualifies the service area instantly and is faster to type than a full address.
What’s going on	Keep	Free text. This is where the real qualifying information comes from.
Service type	Keep, optional	A short dropdown helps routing. Don’t make it required.
Email address	Cut	You’re going to call them. Asking for email adds friction and invites spam.
Full street address	Cut	Ask on the phone. Nobody types their address into a stranger’s form on a first contact.
Preferred appointment window	Cut	You can’t honour it until you’ve qualified the job. It creates an expectation you’ll break.
“How did you hear about us?”	Cut	Answer it with tracking, not with the customer’s patience.
Budget range	Cut	They don’t know, and being asked makes them feel sold to before anyone has helped them.

“Looks good” and “converts” are different jobs

A beautiful contractor site and a high-converting contractor site overlap less than you’d hope. Design awards go to restraint, whitespace and atmosphere. Bookings come from specificity, proof and a fast obvious action. When those conflict, most agencies quietly resolve it in favour of the portfolio piece.

You can tell which one you bought by looking at what the site is confident about. A converting site names towns, states response times, shows the licence, prices what can be priced, and repeats the phone number. A portfolio site talks about craftsmanship and integrity and makes you scroll to find out where they work.

Neither of these is about taste. Google’s own SEO starter guide is blunt that useful, specific content is the durable asset — and useful, specific content is also what a homeowner needs to choose you. The two goals only look opposed when the content is thin.

If you’re weighing a rebuild, work out what it should contain before you shop for one. Our breakdown of what a contractor website actually costs in Texas covers the tiers honestly, and the questions in do you actually own your website will save you from the proprietary-platform trap that catches a surprising number of HVAC and roofing companies. If the rankings themselves are the weak link rather than the conversion, that’s a different programme with a different timeline.

Questions contractors actually ask

What to do with this on Monday

Take the diagnostic table, run the four self-checks that apply to you, and write down what you find. Almost every site we look at has one dominant leak and two minor ones, and the dominant one is usually speed, the first screen, or a form asking for things the customer can’t give yet.

Fix the cheap ones first: make the phone number tappable, cut the form to five fields, put the star rating and licence number where a thumb can see them, and compress the hero image. That’s a day of work and it’s free of risk.

Then measure for two weeks before you touch anything else. If the calls move, you’ve saved yourself a rebuild. If they don’t, you now have evidence about which part of the system is actually broken — and that’s a far better brief to hand a website team than “it looks dated.”