Pritpal S Tamber: The fallacy of user interfaces and big data

I’m lucky enough to spend my workdays around the kinetic kids in Google Campus, London. From what I can tell, they sit about hacking code to see what they might make possible. Occasionally they have flat whites, crack jokes, and look around them, but, in general, hacking is what they do.

For reasons that are not always clear to me they’re repeatedly taken with the idea of creating consumer health apps. A few of them know that I’m a doctor so I get the occasional question about medical “facts.” I always answer with my usual rant about not hard-coding knowledge, and then point to NICE guidance or something. They’re usually fascinated with the idea of soft coding but less interested in NICE’s 160 page PDFs.

It may just be where I work, but I can’t help feeling there is a gold rush in the consumer health apps market right now. The apps tend to focus on the predictable areas of pregnancy, food, and exercise, but the more I download and play with them the less impressed I get (although, to be honest, I haven’t tried many pregnancy apps).

Ultimately, consumer facing health apps are about informing their users with a view to influencing their behaviour. While it’s too early to truly assess most apps, the general consensus on Campus is that most apps are here today, gone tomorrow. I don’t find this remotely surprising; most health apps are poorly thought through. For instance, it’s largely known that the single most important thing in behaviour change is one’s network—if your friends are fat, you’re likely to be fat. And yet so few apps truly consider this in how they’re designed and built.

That doesn’t stop the hype each time a new app is released. Reviewers gush about their promise, consumers swipe, tap, and download, eager to give away their personal data in the hope of a digital magic wand. And, of course, after a small blip in our universe we all get back to eating what we know we shouldn’t eat, not getting enough exercise, and popping pills of dubious value.

We should know better. The diet industry exists—and is lucrative—precisely because people fall off the wagon again and again. An inexhaustible pool of customers getting more and more desperate as time goes on, ever more willing to part with their cash. Just becoming “digital” doesn’t change a thing except how we consume the news that we’re as rubbish as we were last month.

The hype appears to be centred on two things—amazing design-led user interfaces and the messianic promise of “big data.” Both are too shallow, too poorly thought through, to make any real difference.

I’m a big fan of well thought through user interfaces, my current favourite being the shopping app, Svpply. But no matter how pretty the shop window, there’s no point going in if they don’t have it in black. In tech speak, the back-end has to have the right kind of logic to ensure it works and can scale.

With regards to big data, I’ve expressed my reservations before but it’s not hard to grasp that if the app isn’t collecting the right kind of data it’s unlikely it will do any interesting analyses—rubbish in, rubbish out. Too many app developers believe that having user data will be the core of their business and yet few think through what health related data should be collected and how it might be analysed to yield useful, action-orientated knowledge.

Healthcare is probably the one industry growing despite the global recession. Perhaps this is what’s fuelling the outside interest in creating digital solutions. But app developers need to spend more time understanding the problems before they design interfaces or start capturing data. A little less time hacking on Campus and more time understanding the real and messy world of people and their habits.

Pritpal S Tamber is the director of Optimising Clinical Knowledge, a consultancy that helps different types of organisations in and around healthcare design and deliver clinically credible strategies that have a measurable impact on performance and care.

Competing interests: I do not have any competing interests in relation to this post.