A contractor whose website loaded slower than a phone call

Nobody calls about page speed. What the owner noticed was that people kept asking questions the website already answered, and that quote requests came in from customers who said they'd found the business somewhere else — a directory, a neighbor, a truck door.

Checked on the office desktop over office wifi, the site looked fine. That's the trap. The office is the one environment where a heavy site behaves. Nearly every real visitor was on a phone, on cell data, often with one bar, standing somewhere they needed a contractor immediately.

The homepage was pulling several megabytes of photos straight off a phone camera at full resolution — images sized for print, scaled down in the browser to a few hundred pixels wide. The browser still had to download every byte first.

The phone number in the header was part of a graphic. On a desktop that's merely inconvenient; on a phone it means the single most valuable action on the entire site — tapping to call — was impossible. Visitors had to memorize digits from an image and retype them into the dialer.

There were no service pages. Everything the business did lived in one long scrolling homepage. From a search engine's point of view there was exactly one page competing for a dozen unrelated queries, and it was specifically relevant to none of them.

Nothing was measured. No Search Console, no analytics, no way to know how many people arrived and left, or which searches were bringing them.

Images first, because that was the difference between a site people used and a site people abandoned. Every photo was resized to the dimensions it actually renders at, converted to a modern format, and everything below the first screen was set to load only when scrolled to.

Then the tap targets. Real text phone number, tel: link, reachable in one tap from the top of every page. A short quote request that asks for the minimum needed to call someone back rather than a form demanding a project budget from a stranger.

Then structure. Each service the business actually sells got its own page, with its own title, its own description, its own photos of that specific work, and internal links pointing at it from the pages where a reader would naturally want it.

Finally, measurement. Search Console verified, analytics installed, calls and form submissions tracked as conversions, and a baseline recorded before the new pages went live so that later improvement would be a number instead of an impression.

Contractor searches are urgent and mobile. Someone with water coming through a ceiling isn't comparison shopping across six tabs on a laptop — they're on a phone, and they're calling whoever answers first. A site that takes eight seconds to become usable has already lost that job to the next result.

It also compounds. Every service that doesn't have its own page is a category of work the business can't be found for, no matter how good it is at that work. Splitting one homepage into distinct service pages is often the single highest-return change available to a trades business online.

Open your own website on your phone, on cell data, with wifi turned off, and count the seconds before you can tap something useful. That number is your real website.

Then check whether each thing you sell has its own page. If it doesn't, that's where to start — before any spend on ads or link building.