We all know that user testing can be a hassle. It takes time and planning, can disrupt an agreed direction and for smaller projects, it can be a significant portion of the budget. For every company and agency that swears by it, there’s another that dismisses it or struggles to sell it to their clients.
I’m not going to address the business and budget aspects of user testing today - except to say that we can help you plan and run effective tests within your budget - but I do want to talk about the unspoken objection that lies beneath those concerns: that it’s not good value.
I hear it a lot and I’ve been guilty of it myself. We’ve got a great design team, our client is on board, there’s nothing new or out of the ordinary and we’ve stuck to best practice. Why do we need to test? If people can use the internet normally, they’ll be just fine on our site. And if not, we’ll catch it in analytics and change it then. Let’s just crack on and get it done.
OK. Fair enough. It’s true that when done properly, web usability is largely a solved problem (on most sites, most of the time). And if that level of ‘good enough’ pragmatism suits you and your client, if you’re not pushing to innovate and, well, you’ve just done 5 other sites like this before and you’re old hands at it, then yes, it can be hard to justify the cost and effort of usability testing. But here’s the rub: user testing doesn’t just have to be usability testing.
Yes, you want to make sure your site is usable. But there’s also more broadly, your brand positioning, the range of products you sell, the motivations of people buying (or not) those products, what your competitors are up to, customer service successes and failures, and potentially conversations around your high street offering to be had. A usable site is one of the last steps in encouraging people to buy from you. There’s a lot that goes on before then, and your participants will have insights in those areas too.
So don’t restrict your users to just usability. I promise it’s not just messing around with semantics. Just that since you’ve gone to the effort of recruiting 5-10 participants which match your client demographic, got them into a room with you (or on video conference) and are paying them for an hour of your time - since they’re here, you might as well stretch the brief beyond core usability of your project and see what other insights and validation you can discover. Here are some themes we’ve covered in the past that work really well.
This is a usability classic, but always worth doing. Keep the task as broad as possible: e.g. ‘Please show me how you would shop for a christmas present for your mother.’ Don’t direct to specific pages or places initially, but watch your participant complete a whole visit to the site without interruption. Ask them to think aloud then review it with them afterwards. It’s not just usability issues you’re after, but any barrier to purchase - is the copy convincing? Do the images feel right? Do the reviews seem genuine and trustworthy? How do they feel about the delivery process?
One of the disadvantages of standard user testing is that the site information architecture rarely gets a look in. The usual model of asking users to complete related tasks means that at best, two or three sections of the site structure might get looked at in detail (and the problem is amplified when using prototypes and often only a defined path works at all).
The solution to this is tree testing. This is often done remotely and unmoderated, with a cohort of 50+ users to give statistical significance. But that doesn’t mean there aren’t advantages to bringing it into a qualitative environment like user testing. Dedicating 10 minutes of your session to asking users to explore the site menus rather than the pages themselves can give all kinds of insight into the way people search and shop - and helps ensure they can find your products in the first place.
To start with tree testing, make a list of 7-10 diverse products on the site. Ask users to use the menu to show where they would expect to find them, but don’t dwell on the product itself. If they can, great. If not, what’s causing them to go wrong?
How does the site match user expectations of your company? If they don’t know you, ask what impression the site gives. Is it the impression you want? Consider things like use of imagery and positioning text, and any hero products are on display. How does it compare to any retail outlets you might have?
Ask for the users’ general impression but don’t listen too carefully to any specific ideas they might have. Branding is a tricky business and people rarely want what they say they want when they're put on the spot.
Before you rush in and make changes to your site based on your competitors, ask your participants what they think too. All of the user testing staples and the ideas we’ve discussed here apply. Can users easily purchase? Is the site structure right? What do they think of the brand?
All of these aspects can give you both inspiration and warnings. And, as with all user testing, they can be folded back into your designs, prototypes and A/B tests so that you can continually optimise your site and hopefully your conversion rate too.
Kathryn has been working closely with customers and stakeholders for over a decade, helping them to define their information architecture, design and user research strategy. She has worked across a variety of industries and clients, specialising in secure systems for government before moving into the commercial sector with clients such as Vodafone, Fullers, British Gas and National Rail.