Somebody heard about you. Maybe a friend mentioned your name, maybe they drove past your sign, maybe your work showed up in a search. Whatever it was, it worked. They are interested. They pull out their phone and they look you up.
That moment is the most valuable thing that happens to your business all day, and it is almost entirely out of your hands. You are not in the room. You cannot explain anything, you cannot show them the good stuff first, you cannot say “hang on, that photo is old.” All you get is whatever loads on their screen.
And what loads, for a lot of small businesses, is a white rectangle.
The part nobody watches
Here is an experiment worth doing before you read any further. Pick up your own phone, turn off wifi so you are on cellular like a real customer would be, and load your own website. Do not tap anything. Just watch it.
Most business owners have never done this. They look at their site on a laptop, on office wifi, with every file already cached in the browser from the last time they looked. On that machine, the site is instant. It has been instant for years. That is not the experience anyone else is having.
On a phone, on a cell connection, from a cold start, you are watching the thing your customer watches. And what you are looking for is not whether the site is pretty. It is a much simpler question: how long is the screen blank or half-formed before the actual thing you sell appears?
What Google actually measures
Google has a metric for this and it is worth knowing by name, because it turns a vague feeling into a number you can check.
It is called Largest Contentful Paint, usually shortened to LCP. It measures how long it takes for the biggest piece of content on the screen to finish rendering. Usually that is your hero image, or your headline, or the photo of the thing you make. In other words, LCP measures the moment the page stops being a suggestion and starts being a page.
Google’s published threshold is that LCP should happen within 2.5 seconds of the page starting to load. That is the line between “good” and “needs improvement.” Google also specifies that you should measure this at the 75th percentile of your page loads, which is a technical way of saying it has to be fast for most visitors, not just the ones on good connections.
I want to be careful here about what that 2.5 seconds means and what it does not. It is not a cliff. Nobody’s business collapses at 2.6 seconds. Google chose that number by combining research on how long a person will wait before their attention wanders, which lands somewhere in the range of one to three seconds, with data on what real websites are actually capable of achieving. It is a threshold built to be demanding but reachable.
The reason it matters to you is that it is a published, checkable standard. You do not have to take my word for whether your site is slow. You do not have to take your web person’s word either. You can run your own address through Google’s PageSpeed Insights right now and get a number that Google itself stands behind.
Why four seconds and not 2.5
I titled this piece four seconds and then told you the standard is 2.5. That is deliberate, and here is the difference.
The 2.5 seconds is a technical measurement of when pixels appear. The four seconds is the human experience of the whole thing, which includes some parts no stopwatch catches. Your customer taps the link. The page starts loading. Something appears. They read it, or fail to read it. They decide whether this looks like a real business.
That last part is not measured by any tool, and it is the part that decides whether you get the job.
Speed buys you the chance to make an impression. It is not the impression itself. A site that loads in one second and looks like it was built in 2009 has simply arrived at its own funeral on time. But a site that never loads at all never gets to make any impression, good or bad, and that is the more common failure by a wide margin.
What is actually slowing you down
In my experience auditing small business sites, the same handful of causes come up over and over. None of them are exotic.
Enormous images. This is the number one cause and it is not close. Somebody exports a photo at full camera resolution, uploads it, and the browser scales it down to fit a phone screen. So a visitor on cellular downloads a file sized for a billboard in order to look at something the size of a playing card. The picture looks identical either way. The load time does not.
The fix is genuinely boring: export your images at roughly twice the size they actually display, and save them in a modern format. This is a twenty minute job for most sites and it is frequently the single biggest improvement available.
Everything loading at once. Some sites are built so that nothing renders until everything has arrived. Every script, every font, every tracking pixel, every chat widget. If one of those is slow, the whole page waits. There is a tell for this: the page sits completely blank and then appears all at once, fully formed. That is not a fast site. That is a site that refuses to show you anything until it has everything.
Plugins and widgets nobody remembers adding. Chat bubbles, popup builders, review carousels, social feeds, three separate analytics tools. Each one is a small tax. Ten of them is not a small tax. I regularly find sites carrying widgets for services the owner cancelled years ago, still loading, still costing time.
Fonts. A custom font is a file that has to download before text can render in it. Handled well, this costs nothing because the browser shows a fallback font first and swaps it in. Handled badly, your text is invisible until the font arrives.
The part where I do not give you a number
You will find a lot of articles about page speed that tell you exactly what a slow site costs. A specific percentage of conversions lost per second of delay, multiplied by your revenue, producing an alarming figure in bold.
I am not going to do that, because I do not know your numbers and neither does anyone else writing those articles. The studies behind those percentages were run on large e-commerce sites with enormous traffic, and the results do not transfer cleanly to a plumbing company in Fall River or a shop selling handmade goods. Anyone who tells you your slow site costs you a precise dollar amount is guessing and dressing the guess up as arithmetic.
What I can tell you honestly is the shape of it. People leave slow sites. The slower the site, the more of them leave. The ones who leave were, by definition, interested enough to click, which makes them the most expensive visitors to lose. Everything you did to earn that click still cost you whatever it cost.
If you want a real number for your own business, the honest way to get one is to fix the speed and watch what changes. That is the only version of this calculation that is actually about you.
What to do this week
You do not need a rebuild to find out where you stand. You need about twenty minutes.
Run your homepage through Google PageSpeed Insights and read the mobile score, not the desktop one. Desktop scores are flattering and irrelevant, because your customer is on a phone. Look specifically at the LCP figure and compare it to 2.5 seconds.
Then do the phone test. Cellular, not wifi. Watch the screen.
Then look at what the tool tells you is causing the delay. It will name specific files. Nine times out of ten there will be an image near the top of that list, and nine times out of ten that image can be a fraction of its current size with no visible difference.
That is the whole assignment. Not a redesign, not a new platform, not a strategy. Just find out whether the front door opens before your customer walks away.
The uncomfortable part
Here is the thing I find myself saying to business owners more than anything else, and I mean it kindly.
The work is usually fine. Frequently it is better than fine. The people I audit are often genuinely good at what they do, with real customers who like them, and a reputation earned over years. That is not the problem.
The problem is that none of it is visible in the four seconds where the decision gets made. A customer standing on a sidewalk looking at a phone does not know about your twenty years of experience. They know that the page is blank and the other place’s page was not.
That is not fair. It is also not going away. The good news, and it is genuinely good news, is that this particular problem is one of the cheapest ones you will ever fix. Nobody has to fall in love with a new brand. Nobody has to rewrite anything. Somebody just has to make the files smaller and stop the page from waiting on things it does not need.
Twenty minutes to find out. Usually about a week to fix.
Every number in this article is either measured with free public tools or cited to a named source. Google’s Core Web Vitals thresholds, including the 2.5 second LCP target and the 75th percentile measurement standard, are published at web.dev.