We Audited 40 Contractor Websites the Industry Calls the Best. Here Is What They Actually Get Wrong

Search for the best plumbing websites and you will find eight or nine articles listing them. Every one is a gallery. Screenshots, a paragraph of praise, on to the next. None of them measure anything, which means none of them can tell you whether the site being praised would actually get a homeowner to call.

So we took the companies those articles hold up as the best and ran them through the same checks, on a phone, the way a real customer arrives. Forty homepages across plumbing, HVAC and roofing. Eleven measurements each. No opinions, just what is on the page.

The headline result is that the industry has solved the obvious problem and left the hard ones alone. Almost every site makes it easy to tap a phone number. Almost none of them answer the questions a homeowner asks before deciding whose number to tap.

We pulled the company names from published roundups of the best plumbing, HVAC and roofing websites, then loaded each company's own homepage in a headless Chrome browser emulating an iPhone at 390 by 844 pixels with a mobile user agent. Every measurement is taken from the rendered page, not from source code, so anything injected by JavaScript counts.

We attempted 42 sites and analysed 40. Two returned nothing usable to an automated browser, which is its own finding but not one we can fairly fold into the percentages. For each remaining site we recorded whether a tap-to-call link existed at all, whether one sat within the first screen, whether the page named a price, a licence number, a review score, financing, or emergency availability, how many internal links it carried, whether it carried local business schema, and how many kilobytes it transferred to load.

Two caveats worth stating up front. This is a sample of celebrated sites, not a random sample, so if anything it flatters the industry. And a single homepage is not a whole website, so a company that keeps its licence number on an about page is marked absent here even though the information exists somewhere. We measured what a first-time visitor sees in the first ten seconds, because that is the window that decides whether there is a second visit.

Thirty-nine of the 40 sites had a tap-to-call link somewhere on the homepage. Thirty-five, or 88 percent, had one inside the first screen on a phone, reachable without a single scroll. The median site carried five separate tap-to-call links, and one HVAC company carried twelve.

This is worth saying clearly because it changes the advice. Ten years ago the standard recommendation to a contractor was to make the phone number tappable and put it at the top. That advice is now redundant. If you are still being sold a website rebuild on the promise of a clickable phone number, you are being sold something the entire industry already has.

The competitive advantage moved somewhere else, and the rest of this data shows where.

Thirteen of 40 homepages stated a licence number, a CSLB number, or an equivalent state registration anywhere on the page. The other 27 did not.

Homeowners are told, by insurers, by consumer protection agencies, and by every article about avoiding contractor scams, to verify a licence before letting anyone into the house. It is close to the first question asked and it is answered on fewer than a third of the sites the industry considers exemplary. A licence number is a string of characters. It costs nothing to display and it pre-empts the exact objection standing between a visitor and a phone call.

The pattern held across trades. Five of 12 plumbing sites, four of eight HVAC sites, and only four of 20 roofing sites named a licence on the homepage. Roofing being the weakest is the strangest result in the whole audit, because roofing is the trade where insurance work requires a licensed contractor and homeowners are explicitly advised to check.

Twelve of 40 homepages displayed a star rating, a review count, or a visible reviews widget. Seventy percent showed nothing.

This one is harder to explain away than the licence gap, because these companies have the reviews. A contractor with a decade of work has hundreds of them sitting on a Google Business Profile. Leaving them off the homepage means a visitor who arrived from a Google search has to leave the site and go back to the results to find out whether other people were happy, which is a moment you rarely get back.

If you are wondering what to do with the reviews once you have them, our guide to local SEO step by step covers how the profile and the website are supposed to corroborate each other rather than existing separately.

Thirteen of 40 sites put any dollar figure on the homepage. In roofing it was two of 20.

We understand the reasoning. Nobody can quote a re-roof or a system replacement from a web page, and contractors are wary of being compared on a number stripped of context. But the choice is not between publishing a price and publishing nothing. It is between publishing a range with the conditions attached, or letting the visitor assume the worst and go get three other quotes.

The sites in this sample that did publish were noticeably more specific about everything else too. Naming a number forces you to name what is included, what is excluded, and what changes it, and that specificity is what a nervous homeowner is actually shopping for. This is the single largest gap between what the market says it wants and what the industry provides, which is why we build every one of our trade websites around a published figure.

This was the most alarming number in the audit. The median homepage transferred 2,755 kilobytes on a mobile connection. Ninety percent were over one megabyte, 65 percent were over two, and 48 percent were over three. Seven sites were over five megabytes and the heaviest single homepage we measured came in just under 18 megabytes.

Roofing is the worst offender by a wide margin, with a median of 4,073 kilobytes against 2,447 for plumbing and 1,414 for HVAC. The cause is not mysterious. Roofing sells on photographs, roofing sites are built around galleries, and galleries are where uncompressed images go to live. The irony is that the gallery is usually the most persuasive thing on a roofing site and the reason nobody waits around to see it.

It is not that it cannot be done. The lightest site in the whole sample was a roofing company at 95 kilobytes, and three more came in under a megabyte. Google is explicit that loading performance is a ranking and experience signal in its Core Web Vitals documentation, and on a phone in a driveway with two bars, three megabytes is the difference between a lead and a back button. We went deeper on this in why page speed matters.

We counted unique internal links on each homepage as a rough proxy for how much site exists behind it. The median was 18. Fifteen sites, well over a third, had ten or fewer.

A contractor with ten internal links has a homepage, a contact page, and a handful of stubs. There is no page per service, no page per city, nothing for a search engine to rank against a specific query. That company can only ever compete for its own brand name and whatever the map pack sends it, which means it is renting its lead flow from Google Business Profile and owns nothing.

The fix is not glamorous and it is not fast, but it is the highest-leverage work available to most contractors: a real page for each service you want more of, and a real page naming each city you drive to. Our roofing website and plumbing website pages both lay out that structure in detail.

In order, based on effort against return. First, put your licence number and your Google rating in the first screen. That is an afternoon of work and it closes the two objections a homeowner has before they will call anybody. Sixty-eight percent of the best sites in this industry have not done it.

Second, compress the homepage. If you are over two megabytes you are losing visitors on cellular before they see anything, and the images are almost always the entire problem. Third, publish a starting price with the conditions attached for your two or three most common jobs. Fourth, build the pages. One per service, one per city.

Notice that none of this is a redesign. Every item is content and weight, not layout. The most common thing we see is a contractor spending five figures on a rebuild that looks better and answers exactly the same questions the old one did, which is to say almost none of them.

Forty sites is enough to see a pattern and not enough to be definitive. The sample is drawn from published roundups, so it is biased toward sites someone already thought were good, which makes the gaps more striking rather than less. We measured homepages only, on one device profile, on one day. A company may well answer all of this on an interior page.

What we would not walk back is the shape of the finding. The easy, visible, frequently-sold fix is done nearly everywhere. The unglamorous work of stating your credentials, showing your reviews, naming a number and keeping the page light is done almost nowhere. That gap is where the jobs are.

If you want to know how your own site measures against these eleven checks, send us the URL and we will run it and send the numbers back within 24 hours. No pitch attached, and you can take the list to whoever built it. You can also browse the trades sites we have built and check them against the same list, which is only fair.