Sometimes, you need to design something really odd

I’m usually a little cautious when critiquing sites where I’m not familiar with the design rationale. I’ve often read a critique of a site that I’ve worked on, where the issues raised make sense from the point of view of that individual, but wouldn’t result in us changing anything as there was actually a good reason for the design they didn’t like. Usually, it related to the needs or behaviours of other personas that they hadn’t thought about, or an overriding commercial imperative.

Here’s an example.

In the checkout process on British Airways’, the customer is asked for a phone number. Here’s the relevant bit.

Provide phone number on

It was often pointed out to us that we should have put the radio buttons above the number input field. An alternative would have been to allow the customer to put in a number, and then let them tell us whether it was a landline or mobile number.

In this case, if you hit ‘No’, then you see this.

Result of saying ‘I don’t have a mobile’

The reason the fields are in this order is that it was quite likely that no use at all would be made of a landline number. In a real emergency, or if you were high value, then maybe.

Even if we’d suggested strongly that it was in the customer’s own best interest to provideĀ  mobile number, some people still wouldn’t, for their own reasons. By doing it this way, people would see the request for the mobile number, see it was mandatory (so far), and maybe be a little pissed off, but would give a number. Many wouldn’t notice the ‘No’ option. Even if they saw the ‘No’ after typing in a mobile number, people being lazy, they’d just leave it.

If there were customers who really didn’t have a mobile, they had put so much effort into getting to this point that they would hunt around for a way round this problem, and spot the next bit.

Whilst we never A/B tested this one, I’m sure that the end result was that we collected many more mobile numbers than we would otherwise have done, and that we didn’t lose significant revenue.