The numbers crept up on a lot of people. By the start of this year a good number of retailers were seeing a fifth or more of their traffic arrive from phones, where eighteen months earlier it was a rounding error. Nobody sat down and decided to court mobile shoppers; the shoppers just turned up on phones, and the site was never built for them.
If you sell things online, do the thing nobody enjoys. Get hold of an iPhone and a cheap Android handset, sit down with both, and go all the way through a purchase on your own site. Add to basket, log in, enter a card, complete the order. I’ve done this with a few sites now and it is a chastening half hour every time. Tap targets you can’t hit with a thumb. Forms that assume a full keyboard. Menus that only open on a mouse hover, which a phone doesn’t have. Product images that take an age to arrive over 3G.
There are two ways people are solving this at the moment, and I don’t think either is obviously right yet.
Some are building a separate mobile site, the m.dot approach, where a phone gets redirected to a stripped-down version. It’s quick to stand up and you can tune it hard for small screens. The cost is that you now have two sites to keep in step, two sets of URLs, and a redirect layer that goes wrong in interesting ways when someone shares a link.
The other camp builds one site that reshapes itself to whatever screen it lands on; the responsive approach Ethan Marcotte wrote about a couple of years back. One set of URLs, one codebase, no redirect. It’s more work up front and the performance can suffer if you’re not careful, because the phone still downloads a great deal of what the desktop does. For a content site I’d already lean responsive. For a big transactional shop with a fiddly checkout, I can see why people are still nervous about it.
Wherever you land on that, the checkout is where the money actually drains away. You can have a lovely responsive homepage and still lose the sale at a payment form that wants a sixteen-digit card number typed into a field the size of a postage stamp.
If you’ve only budget for one thing this quarter, I’d put it into the phone checkout ahead of the homepage. Fewer form fields, bigger ones, and guest checkout so nobody has to invent a password on a train. It’s the least interesting part of the site to work on, which may be why it’s the part still costing you orders.
