What a Slow Website Costs a Contractor on Google

Most of what Google does to your business is invisible. You can’t inspect the ranking math or watch the ad auction. Page speed you can actually measure. Google publishes a free test, anyone can run it against any website, and it prints a score you can hold in your hand. If you check one technical thing about your own website this year, check this one.

Running the test

Go to pagespeed.web.dev in any browser. Paste in your website address and run it. There’s no account and nothing to install. In about a minute, Google loads your site the way a visitor’s phone would and hands back scores for mobile and desktop.

Read the mobile score first and care about it most, because homeowners reach contractors from phones: standing at the breaker box, or in the kitchen with a towel wedged under the sink. The desktop score flatters almost every site. The mobile score is the one that matches how your next customer will experience you.

Then spend ten more minutes and run your market. Search your trade and your town, take the contractor websites that come up, and put each one through the same test. Write the scores down. You now have a measured table of your local competition, built with Google’s own instrument, and almost nobody in any trade bothers to build one.

Keep the table simple. One row per company, with the mobile score, the desktop score, the date, and anything odd you noticed while the page loaded. Pull the names the way a homeowner would, by searching the trade and the town on a phone and taking the companies that actually come up, because the list Google shows a homeowner and the list of rivals in your head usually differ, and the first one is where the calls come from. Ten companies covers most markets, and the whole exercise fits inside a lunch break.

What the number is telling you

The test measures the experience of arriving: how long the visitor stares at a blank screen, and how long before the page settles down and responds to a tap. That’s the whole story a homeowner lives through in the first seconds on your site, compressed into a score. You don’t need to understand the technical readouts underneath it to use it. Passing and failing are visible at a glance, and the gap between your score and a competitor’s score is a fact either of you could verify in a minute.

Fine on your desk, slow in a driveway

The most common objection is the owner loading his own site on his own phone and finding it fine. That check is rigged in your favor twice. The office wifi outruns the cell connection your customer stands on, and your phone has visited the site so many times it keeps most of the page stored, so it barely downloads anything. A stranger’s first visit pulls every photo and script over whatever signal reaches a driveway, and that’s the visit the test simulates, the one your own thumb can’t reproduce.

Which page you test matters as much as which phone. The homepage is what owners look at; a search for a specific problem lands on the specific service page. Run the test against the water heater page or the roof repair page, whichever pages match the searches that pay you, because a fast homepage sitting in front of slow service pages fails exactly where the customers arrive.

The two ways slow costs you

The first cost is visitors. A homeowner taps your site from the search results on whatever signal reaches their basement, and if the page hangs, they hit back and tap the next contractor down. Everything it took to earn that tap gets spent, and the visit goes to a competitor whose page loaded. You never see it happen; the caller just moves to the next name.

The second cost hits shops running Google Ads. Google scores the page an ad clicks through to, and the experience on that page feeds the ad’s quality scoring. A slow page drags that score down, and the score is part of how Google decides what you pay per click and where your ad sits. Two shops can put the same money on the same search, and the one with the faster site gets the better end of the auction.

Speed and where you show up

There’s a third cost, smaller and slower than the first two. Google’s public guidance includes page experience, with speed among its parts, in how results get ordered. Treat the direct effect as modest; pages that answer the search and a strong review record move placement more than any speed score will. The indirect effect is bigger and has nothing to do with the algorithm: every slow load hands a visit you already earned to a competitor, and that loss repeats on every search the page was built for. Passing is a floor rather than a boost, and a failing score drags on everything else the site was supposed to earn.

What a lost visit costs in your trade

The statistics decorating every page on this subject, the three-second abandonment figure, the conversion percentage lost per second, trace back to old studies of retail checkout pages and to vendor decks quoting each other. Nobody publishes a defensible number for what slowness costs a contractor site specifically, and this page won’t invent one. The arithmetic that matters is yours anyway. A contractor website exists to produce phone calls. You know your average ticket and about how many calls it takes to book one job, so the question comes down to how many calls a month the site produces and what one more would be worth.

The visitor being lost is also the best one you get. A homeowner usually taps through to your site having just read your review record, which means the visit comes from someone most of the way to a decision, and plenty of them are standing next to the problem when they tap. Losing that visitor to a spinning page wastes everything the reviews just did. And once the page does its job, the call it produces still has to get answered; when those calls arrive and who should be answering them is written up separately.

What the market actually looks like

We run this test in bulk. In every market we’ve measured, most contractor websites fail the mobile test. That includes plenty of sites with a monthly bill attached. What a contractor pays for a website and what the website scores turn out to be two separate facts. The common causes are unglamorous: photo files uploaded straight off a phone at full size, page-builder platforms hauling a stack of plugins, and sites nobody has opened since the day they launched.

That widespread failure is the useful part. If your site fails and every competitor’s site fails too, which is the usual finding, then passing is an open lane. A fast site in a market full of slow ones wins back the abandoned visits and buys its ad clicks on better terms, and none of them have ever run the test.

Run the test from the parking lot of a supply house sometime, on your phone’s connection, and just try to use your own site. Tap the phone number. If it places the call in one tap, good. If it’s printed text you’d have to copy by hand, that’s a fix worth requesting by name, because a visitor who has to retype a number will often just go back to the list. Find your service list and count the taps it takes. Feeling it once usually ends the debate.

Common questions

How fast should my website load?

Skip the seconds debate and use the grade. Google’s test reports whether your site passes on mobile, measured the way a customer’s phone experiences it, and passing that is the target. Chasing a perfect score past a solid pass buys little.

Does website speed affect my Google ranking?

It’s part of Google’s page-experience guidance, so yes, modestly. The bigger effect is indirect: slow pages send visitors back to the results, waste paid clicks, and produce fewer calls from the same traffic. Fix it for the calls and treat the placement help as a side effect.

How do I check my website’s speed?

Run pagespeed.web.dev against your address; it’s free and needs no account. Read the mobile score first, and test the service pages that match the searches that pay you, since those are where homeowners land. Ten more minutes gets the same scores for every competitor in your market.

What makes a contractor website slow?

Most often: photos uploaded at full size straight off a phone, page-builder platforms with a stack of plugins, bargain hosting, and scripts from bolted-on extras. Photos are the cause that keeps growing after launch, which is why sites that once passed drift back to failing.

Is speed worth paying to fix, or should I rebuild?

Oversized photos and stray plugins tune out quickly and cheaply; when the platform itself is the problem, tuning stalls below a passing score and the fix is a rebuild. A vendor whose optimizations keep stalling is showing you which case you’re in.

My site feels fast to me. Is the test wrong?

The test measures a different visit than yours. Your phone rides the office wifi and has most of the site stored from past visits; a first-time visitor downloads everything over a cell connection. The score models the stranger’s visit, and the stranger is the one deciding whether to call.

Using the number

If your site passes, screenshot the result with the date on screen and rerun it every few months. Sites drift as pages and plugins pile up, and a passing score from two years ago proves nothing about today.

The main thing that changes on a contractor site over time is photos, since job pictures keep getting added long after everything else is finished. Have whoever posts them resize the files before uploading rather than sending them up straight off the phone, and spot-check the score after any big batch goes live. That habit heads off the most common cause of drift.

If your site fails, you’re holding something better than a complaint. Take the dated score to the company you pay for your website and ask two questions: what specifically makes it slow, and when will it pass. Vague reassurance gets harder to hand you when Google itself issued the grade, and and the grade comes from the same outfit that decides where you show up when someone searches your trade when someone searches your trade in your town.

Put the two questions in an email so the answers land in writing. A real answer names causes you can picture, oversized photo files or a platform loaded with features the site never uses, and it comes with a date attached. If what comes back is reassurance about how the site looks, or the claim that the score doesn’t matter, run the test again the day after they call it fixed and let the number settle it. Keep each screenshot in one folder with the date in the file name, so the history of the site is something you can read at a glance.

There’s a fork to understand before that conversation. When the causes are oversized photos and a few stray plugins, tuning works, and the score moves inside weeks. When the platform itself is the weight, a page-builder stack dragging weight the site never needed, tuning stalls below a passing score no matter who does the tuning, and the honest quote is a rebuild on something faster. You can tell the cases apart from outside: fixes that move the score a few points and then stall are the signature of a platform problem. A vendor who says rebuild and can show why is handing you better information than one selling another round of tuning on a platform that can’t get there.

Either way, keep the competitor scores you wrote down. A market table like that turns every argument about whose website is better into a column of numbers. We publish that table for the markets we work in, every contractor site scored with the same test on the same day, on our proof page. This is the kind of thing we handle for the contractors we work with.

By Dave, The Contractor’s Contractor