Loading
Loading
The moment they leave
Somebody opened your website on a phone and left. You assumed it was the price, or the photos, or something you said. It was almost certainly none of those.
About 10 minutes
The short version
Everybody who has ever looked at a report about their own website has done this. People are arriving and the phone is not ringing, so you go looking for the sentence that put them off. The prices are too high. The photos are getting old. The wording on the services page has never sounded right.
Then you change one of those things, and nothing happens, because you were looking at the wrong end of the visit.
Here is what a real one looks like. Somebody taps your result on their phone. They wait. Something appears over the top of the page. They go to close it, miss, get it on the second go, start reading, and the page jumps under their thumb. They end up somewhere they did not choose. They go back. Now they are looking at the search results again, with the business under you sitting there ready.
Ask that person ten minutes later why they did not ring you and they will not be able to tell you. Nothing happened that was worth remembering.
The back button is not something you chose, or paid for, or could change if you wanted to. It belongs to the phone. Which is exactly why it beats everything you built.
It is in the same place on every website in the world, it has never once failed to work, and it needs no aim. On most phones it is not even a button any more. It is a flick of the thumb in from the edge of the screen, which is about the cheapest movement a hand can make.
Everything else on your site has to be found first, then understood, then hit accurately. Leaving has to be none of those things.
So it wins every close call. Not because your website is bad, but because leaving is free and everything else costs something.
As far as we can find, nobody has run a proper study of what makes somebody leave a small business website on a phone. What there is instead is a great many people saying so in public, at length and with feeling, and the same five things come up.
They are in the order they come up, which is not the same thing as measured, and we are not going to dress it up as though it were.
Something covered the page and the cross was too small to hit
The loudest of the five, and it is not really the pop-up they mind. It is the cross. Half under the header, off the edge of the screen, or about the size of a grain of rice.
The page moved while they were tapping
Something arrives late, everything below it shifts down, and the thumb already on its way lands on whatever slid into that spot. Twice is enough.
Back did not do what back does
They open one of your photos, swipe back to close it, and the whole website closes instead. It gets a section of its own below, because nobody sees it coming.
They could not find the thing they came for
A phone number. Whether you come out to their suburb. Roughly what it costs. Not on the first screen, not on the second, and there is a menu to open before anybody finds out.
It had obviously been built for a computer
Text they have to pinch to read, a page that slides sideways, boxes the width of a matchstick. People say it makes them trust the business less, which is not what it was meant to say.
Read down that list again. Four of those five are things that happened to somebody's hands. Only the fourth is about your business at all, and even that is not a judgement on what you charge. It is a judgement on whether anybody could find out.
Very few people leave because of your prices. Most of them leave before they get that far.
This is the part that is hard to see from the inside. Any one of them is survivable. Somebody will put up with one without noticing and with two while getting slightly annoyed. It is the third that finishes it, and the third is usually small.
Which is why the last thing they saw looks so innocent when you go back and check it yourself, and why owners so often conclude the visitor was never serious in the first place.
The last of the five has a piece of its own, because it is not really a design decision at all: your website was designed by somebody sitting down.
Of the five, the third is the one worth understanding properly. It is the only one most owners have never heard of, and the only one that happens on sites that otherwise look completely fine.
Here is the whole thing. It takes about four seconds.
Nobody did that on purpose and it is not a fault in your website builder. It is what happens when nothing has been said otherwise. The photo viewer, the menu that slides in from the side, the booking panel that rises up from the bottom: all three look like a new screen, so people treat them like one, and back is how you leave a screen.
If back has not been told that opening the photo counted as a step, it does the only other thing it knows. It leaves.
Open a photo on any of the enormous websites you use every week, the ones with more customers on phones than Australia has people, and swipe back. The photo closes. The page underneath is exactly where you left it, at the same point down the page, with the same things open. Swipe again and you go back a step further.
None of that is decoration. They worked out a long time ago that on a phone the back gesture is not really navigation at all. It is undo.
This is what makes it different from every other fault a website can have. Somebody who gives up on a slow page has lost some time. Somebody who gives up on a confusing one has lost some patience. When back takes them out of your site, they do not land on nothing.
They land on a tidy list of businesses that do exactly what you do, one under the other, with yours now the one they have already tried.
At some point somebody has shown you a bounce rate, and the number was high enough to be alarming. It gets explained as the share of people who took one look at your website and left.
Here are two visits to a plumber's website on the same evening.
The visit that failed
The visit that rang you
As far as any report is concerned those are the same visit. One page, eight seconds, gone. Both of them counted against you.
Worth knowing, because the definition changed a couple of years ago and the fear did not.
The old one was simple and unforgiving. One page and no second one was a bounce, whether they had been there four seconds or four minutes. That is where the folklore comes from, and it is the version most people are still frightened of.
The one in use now is narrower. A visit only counts against you if all three of these are true at once: it lasted less than ten seconds, nobody opened a second page, and nothing happened on it that you had set up as worth counting.
That third condition is doing all the work, and hardly anybody has set it up. Tapping a phone number can be counted, but only if somebody turned that on deliberately, and on most small business websites nobody ever did. So the visit that ended in a phone call is still filed under the visit that failed.
There is no way to separate those two visits inside the report. There is a way to work out roughly which one you have, on your own phone, in about five seconds. Open your home page and do not scroll. If your phone number is on that first screenful and it dials when you tap it, some share of those short visits are people ringing you.
If it is not there, none of them are.
About ten minutes, and nothing here needs anything but the phone in your hand. Do them cold, standing up, on mobile data, and not on the version of your website you have looked at four hundred times on a laptop.
In that second, your competition is not the business two doors down.
It is the back button.Do the escape test. It takes ten seconds, a great many sites fail it, and it is the only one on the list where the person you lose was actively trying to see more of your work. Everybody else on that list was on their way out anyway. That one was on their way in.
Worth saying at the end, because a piece like this can leave you treating every exit as a wound.
Somebody who wanted a service you do not offer should leave. Somebody who lives an hour outside where you go should leave, and the faster the better. A website that gets those two off the page in ten seconds has saved you a conversation you did not want, and the report will file both of them as a failure.
The exits worth caring about are much narrower than the number suggests. Somebody who wanted exactly what you sell, was ready to pay for it, and went anyway. Nothing above tells you how many of those you have had, and anybody who offers to is guessing.
What it does tell you is where to look if you have had any. Not at the prices, and not at the wording. At the four seconds before they gave up.
Next