Loading
Loading
What the test misses
The test asked whether your website fits on a phone. It never asked whether anybody could use it, and Google switched it off at the end of 2023.
About 9 minutes
The short version
Somewhere along the way you were told your website is mobile-friendly. Possibly by whoever built it, possibly by a plugin, possibly by a report that arrived in an email with a green tick on it. Whatever the source, the phrase came from one place originally, and it meant something quite specific and quite small.
Google ran a test with that name. It asked four things.
Is it set up for a small screen at all
One instruction inside the page, telling the phone not to pretend it is a computer. Every template written this decade has it.
Does the content fit across
Nothing sticking out past the edge, so there is nothing to drag sideways to read.
Is the text big enough to read
Big enough that nobody has to pinch and zoom to get through a sentence.
Are the things you tap far enough apart
Far enough that pressing one of them does not press the one next to it.
Read the left-hand column again. Every one of those is a question about whether the page fits. None of them is a question about whether a person got anywhere.
Passing meant your website is the right shape for a phone. It never meant somebody could use it, and it was never sold as though it did.
At the end of 2023 Google switched it off. The standalone test and the mobile usability report that went with it were both retired, and what is left in their place is a broader speed and experience report that measures different things again.
Which is worth knowing for one practical reason. If anybody is still showing you a screenshot of a passed mobile-friendly test as evidence that your website is fine, they are showing you a result from an instrument that has not existed for over two years.
Google itself was always fairly plain that these checks were a starting point rather than a verdict on the experience. That caveat travelled nowhere near as well as the green tick did.
This is not a complaint that the test was badly built. It was built well for what it did. The limitation is structural, and it is worth understanding because it applies to every automated check that has ever replaced it and every one that will.
A tool reads your page. It can measure anything written down in there. What it cannot do is hold the phone.
Can they get to it
Fitting on the screen and being within reach of the hand holding the screen are not the same problem, and only one of them was ever measured.
Can they get through it
Every check is of one page, on its own. Nobody came to look at one page. They came to finish something.
Will they wait for it
A page can be laid out perfectly and still lose somebody in the four seconds before it appears.
Can they decide
Whether the thing they came to find out is actually on there. No tool has an opinion about that, because no tool knows what you sell.
Those four are where a phone call is won or lost. Not one of them can be read off a page, which means not one of them has ever appeared in a report you were sent.
It is tempting to conclude the whole thing was a con. It was not. Fitting on the screen genuinely matters, and a site that fails these four checks is in real trouble. The problem is what a pass got treated as: a clean bill of health, handed over at the end of a job, closing a conversation that had not properly started.
How a site ends up in that position in the first place is its own piece: your website was designed by somebody sitting down.
If you have ever been shown a diagram about this, it was probably a phone with coloured bands across it. Green near the bottom where the thumb goes easily, amber through the middle, red across the top where it cannot reach. It is a tidy picture and it has been passed around design articles for well over a decade.
The field research that picture is drawn from does not actually say that. Somebody went out and watched a large number of people using phones in the street, in queues, on trains, which is a genuinely useful thing to have done. What they saw was messier than the diagram, and considerably more useful.
What the diagram says
What was actually watched
The comfortable area is not a region of the glass. It is wherever the hand happens to be, and the hand keeps moving.
The most useful single finding is the one that gets left out of every diagram. When somebody wants to tap something, they do not stretch towards it. They scroll it down into the middle of the screen and tap it there, without ever deciding to.
Which is why the middle of a phone screen is the most valuable space on it, and why an argument about the bottom corner was always only half the story.
It also explains something that otherwise looks contradictory. A control in the top corner is not unusable. It is just expensive, in small deliberate movements, every single time. What the biggest businesses learnt about that is the piece next to this one.
If the comfortable area were fixed, the job would be to move everything into it, once, and be finished. Because it is not, the job is different and rather easier: make sure that whatever somebody needs is near them at the point they need it, wherever on the page that happens to be.
A phone number at the top of a long page is not badly placed. It is well placed for the first screenful and absent for the other eight.
The advice most websites were built on was written when a phone was a small thing you could work the whole of with one thumb. On a phone that size, a menu in the top corner was a stretch. Annoying, but a stretch.
Phones are now most of the size of a paperback. The top of a modern screen is not a stretch from the bottom of it. It is a different grip, which means letting go slightly, sliding the phone down the palm, and taking hold again, usually while walking.
The same corner that used to cost a stretch now costs a small calculated risk of dropping the thing.
Meanwhile the checks did not move. A tool asking whether the text is legible and the buttons are far enough apart returns the same answer on a big phone as on a small one, because none of what changed is written down in the page. It is in the hand.
This is the part worth sitting with if your website has not been touched in a few years. Nothing about it changed. It still passes everything it passed the day it went live. The people holding it changed, and the thing in their hands got bigger, and a layout that was merely a bit awkward became genuinely hard work.
No report would ever have told you. There is nothing in the page to measure.
About ten minutes. Every one of these is chosen because it sits exactly where an automated check cannot go, which is the whole point of the last four sections. Do them standing up, on your own phone, with one hand. Nothing gets scored and there is nothing to send anybody.
No tool can tell you whether it was easy.
Only a person can.Do the stranger test, and do the hard part of it, which is keeping quiet. Ten seconds of watching somebody else's thumb hover over your home page will tell you more than any report you have ever been sent, and it is the one thing on this list you genuinely cannot do for yourself.
Worth saying plainly, having spent this long on the subject. The five things above do not add up to anything. There is no number at the end and no grade, and if there were, it would deserve exactly the scepticism the green tick did.
The fault was never that somebody measured something. Measuring is fine, and a check that tells you your text is too small to read is genuinely useful. The fault was measuring whether a page fits and then treating the answer as though it described what happened to a person.
There is only one result that settles this, and you already have access to it. If the phone is ringing, your website is doing its job, whatever any report says about it. If it is not, the reason is in one of those five tests, and no tool was ever going to hand it to you.
Next