Actual usability VS perception of usability
How redesigning the worlds most profitable menu taught me that users sometimes prefer *less* usability.
Some backstory
2015: Richard III is reburied after being discovered under a carpark in Leicester. The 50 Shades of Grey film is released. Kobe retires. And I had just joined the worlds biggest online food brand as a UX Designer.
It was exciting working on a product that handles over £2.5 billion pounds worth of orders. But, I must admit, it was also quite daunting: How do you make an impact to a brand that already rules every market it’s in?
I was even more apprehensive and quite honoured when I was asked to redesign the menu. The menu is the main thing after all!
Don’t make me scroll
After running some user testing, one of the most common complaints were ‘the menus are too long’.
Of course they need to be fairly lengthy to incorporate all the dishes, and participants didn’t want any reduction of choice: They just wanted less scrolling.
I produced a few prototypes to test different patterns to combat this. Perhaps we should separate each section into pages? Or we could use accordions to minimise the sections that they weren’t interested in? This approach frustrated users with ‘more clicks’ and obfuscating choices. It also led to far smaller order sizes as, with the long page, people often discovered something tasty whilst scrolling.
It was pointed out by several participants that the menu wasn’t using the screen space very well on wider screens, being as it was, just a column in the middle of the screen.
My instinct was that having multiple columns would make the page shorter but also items harder to find as you would have to look left and right, as well as up and down. In testing this was proven, with participants taking far longer and in many cases outright failing to complete the task of adding a list of items to their cart.
Measure the love
At the time our stated company mission was to ‘empower customers to love their takeaway experience’. Despite this, ‘love’ was one of the few things we didn’t measure. In response I had introduced a couple of initiatives to deal with this: The relevant one here was ‘JESUS’ or Just Eat System Usability Score based on the well known SUS scoring method, but rewritten to be more appropriate to our users. The aim was to have a standardised way to measure improvements of usability AND satisfaction across all departments.
Interestingly, though users were struggling far more with the multicolumn approach the satisfaction was higher — especially for a variant that used modules where we had grouped dishes together into blocks. Even participants that had failed to complete the task were reporting that as their favourite with comments gushing about how ‘easy to use it is’!
So what do we do? Knowingly release an objectively worse product that customers love? Or keep the more usable but disliked approach?
This was a conundrum and one that I hadn’t faced in my still fairly new UX career. I spoke to several people with more experience but for the most part they weren’t sure what to do either. The general consensus was the long page meant people were able to find the dishes they wanted and led to bigger order values for the company. It was less satisfying to use but it would be hard to justify a drop in order value unless we could prove it led to greater retention longterm — which I wasn’t sure we could.
In less of a flash of inspiration and more of a hail-mary, I tested a version that retained the modular look that was so popular but just stacked on top of each other. This was just as usable as the long page, and proved just as satisfying as the multicolumn modular approach.
It turned out we could have our pizza and eat it.
Though I noticed — with some irony — that the additional UI chrome of the winning variant actually made the page longer!
So what did we learn?
It’s often said ‘we eat with our eyes’ and I believe this is true for websites and other designed products as well. Things can look and feel usable even if they are objectively worse.
SUVs are super popular. Especially crossovers that are just hatchbacks on stilts making them heavier, less economical, less safe, dynamically compromised, and more expensive. But people love them. That comes from the heart, not the brain.
I remember when Uber first started becoming popular, their app was held up as a bastion of usability, but I had some serious reservations about the UI. In fact, I remember discussing this with a colleague over a beer. He very much fell into the opposite camp of me and argued passionately about how perfect it was. Shortly after leaving to jump in his Uber, he returned red faced after it turned out he had ordered his taxi to pick him up from his house not the bar.
More recently we, along with many families this Christmas, bought a Yoto. A cleverly designed little media player that allows kids to listen to stories and music through the insertion of cards. All the parents are raving about how easy and brilliant to use it is. And for the most part, I’d have to agree… yet, with unlabelled controls and hidden buttons, I don’t know a single person who has worked out how to use the ‘radio’ function and everyone has had issues with the app and setting up family accounts.
In short, the look and feel of a thing contributes massively to the experience. And can often cover the cracks where the product fails.
As product designers we should be aware of this and can lean in to it to make more satisfying products that people enjoy using. MVPs can be a useful way to get something out quickly, but often ‘viable’ requires those that use it to *want* to engage. If your competitors have already got market share, or a bigger advertising budget, or simply a better product, one way to win is by delivering a product that is loveable.
Building products that work should be a given — but if we want them to be really successful they need they appeal to the heart too.
One last thing
It takes data and learnings from many different sources to be able to understand what customer want, what they need, and how to deliver that. But it can be difficult to bring many sources together in a way that is actually useful and helps innovation happen.
One way you can solve this is with an atomic insights repository like Glean.ly — this is a product I’ve been building for the last five years and is helping some of the worlds biggest and best UX teams share and connect knowledge from across their organisation and beyond. Give it a go with a 30 day free trial. I hope to see you on a demo or onboarding call soon!