Nobody Waits for a Slow Website, and Google Stopped Waiting Too

A visitor decides whether your business looks competent before a single word of your copy has loaded.

Here is the scenario that costs the most money, and almost nobody tests for it. Someone taps your link on a phone, on a patchy signal, standing outside somewhere. Your homepage begins to assemble itself. A hero image arrives late and shoves the text down. They go to tap a button and it moves. They go back and tap the next result instead.

No bounce message, no complaint, no way of knowing it happened. Just a job that went to somebody else.

The three numbers Google measures

Site speed used to be a vague argument between developers. It is not any more. Google publishes three specific metrics, called Core Web Vitals, and it judges your pages against defined thresholds.

INP replaced the older responsiveness metric in 2024, and it is the one that catches sites out, because a page can paint quickly and still feel like treacle the moment somebody touches it.

• 2.5sLargest Contentful Paint target
• 200msInteraction to Next Paint target
• 0.1Cumulative Layout Shift target

You are judged on your worst visits, not your average ones

Google assesses these at the 75th percentile of page loads, split between mobile and desktop. So a quarter of your visitors can be having a bad time while your own testing (on office wifi, on a laptop, on a warm cache) looks perfectly fine. Test on a phone, on mobile data, in a private window.

What actually makes a small business site slow

It is rarely anything exotic. In practice it is the same three things.

Images nobody resized

A photographer sends 4000-pixel files, somebody uploads them straight into the page, and every visitor now downloads a print-quality image to look at a 400-pixel box on a phone. This is the single most common cause and the easiest to fix.

Scripts you forgot you added

A chat widget, two analytics tags, a heatmap tool, a review carousel, a font service, a booking embed, three tracking pixels from campaigns that ended last year. Each one felt free at the time. Together they are the reason the page will not respond when somebody taps it.

Hosting chosen on price alone

Shared hosting at a few dollars a month means your site is queuing behind hundreds of others. It is genuinely fine for a brochure nobody visits. It is a poor decision for a site that is supposed to bring in work.

There is a fourth cause worth naming, because it is the one that creeps back. A site is built fast, then over two years somebody adds a slider here, a popup there, a new font for a campaign, a plugin to fix something a previous plugin broke. Nobody removes anything. Performance is not a job you finish; it is a thing that degrades unless somebody occasionally looks.

Wavefront Studio built our entire website ecosystem from scratch - over 12 connected sites, all running smoothly. Their attention to detail and deep understanding of business was unlike any other agency we’ve worked with.

(Tony Dyke, CEO of ResinRock)

What to do, in order

1. Run your homepage and your busiest service page through Google’s PageSpeed Insights and write down the three numbers for mobile.
2. Fix the images first. Correct dimensions, modern formats, lazy loading below the fold. This alone resolves a surprising share of failures.
3. Audit every third-party script. Remove anything you cannot name a current reason for. Defer what survives.
4. Reserve space for images, embeds and banners so nothing shifts as the page loads. That is your CLS fixed.
5. Only then argue about hosting and caching.

Do it in that order. Teams routinely start with the expensive infrastructure conversation and never get round to the 6MB header image that was causing most of the problem.

What speed will not fix

A fast website is not a good website. If the copy does not say what you do, if there is no obvious way to get in touch, or if the offer is unclear, shaving a second off the load time changes nothing. Speed removes a reason to leave; it does not supply a reason to stay.

Treat it as a floor rather than a strategy. Below the floor you are losing people you never hear about. Above it, the rest of your marketing finally gets a fair hearing. It is also worth being realistic about the ceiling: once you are comfortably inside the thresholds, further milliseconds stop being where your money is best spent, and the conversation should move to what the page actually says.

Start with the phone test. Your own site, your own thumb, mobile data, private window. If you find yourself waiting, so is everybody else.