Earlier this year I was working with a client on a project where we were aiming to radically simplify their website into an app. The website was a marketplace offering locations to be paid for and booked. Like most marketplaces it had a standard search that enabled users to sift through many options to find the result for them. This worked well for people who wanted to plan ahead and do lots of comparison but this isn’t for every one.
The idea for the app was to take away a lot of this work and get the user to a booking in the fewest taps possible. One of the key ways we could do this would be to remove the search step altogether and just present the user with their result.
The company had an algorithm that would crunch all the options based on your location and could present you with the most suitable result. There would also be a settings screen where you could specify your preferences to help inform the algorithm. The product manager was particularly keen that we stuck with this route, convinced users wanted the work done for them.
I made a prototype of this option and carried out an early round of guerrilla user testing. People were happy with most of the flow and comfortable using it. However there was some common feedback when they saw they only had one result: “Is this my only choice?” “Is there a cheaper option available?” “What if I don’t want to go in that direction?” etc.
Overwhelmingly the users weren’t fans of only having one choice to book. When you are allowing the booking of things with variable prices and locations, it’s going to be very hard to get that spot-on for each user in only one result.
However, during my design I saw this being an issue so I also put together an option with three results. The three-option design allowed the users to see the price of each item in the context of two others and give them a sense of whether each was a good deal. Equally for location, they now knew which was closer and if there was a trade-off in cost for that.
This design had an option pre-selected so if they were happy with that they could instantly book, meaning it didn’t necessarily add any extra steps to the booking flow. The users were much happier when shown this second version.
It turned out the product manager was wrong. People didn’t want to be told what to have, they wanted choice so they could feel that they were making an informed decision, even in a brief experience.
You’ve probably heard of the paradox of choice, which concludes that people get paralysed when presented with too much choice. The famous experiment that illustrated this specifically showed that people don’t want 24 options to choose from: it’s far too many to judge all the different variables. However, this also doesn’t mean it can be extrapolated to the point that just one is the right solution.
Three options seems like a pretty good sweet spot. It’s enough variations to be able to consider at once and decide what represents a good deal without being overwhelming. It’s useful to consider this when designing for ecommerce websites.
Nathan Barry writes about how you can increase book profits when you create two additional packages. This anchors the prices to the highest one, and also brings into play the centre-stage effect, where people tend towards the middle choice if they aren't sure what to pick.
You want to avoid giving your users one option, because that’s no choice. Well, it’s the choice of buying or leaving your site altogether. By offering three options, the user will now be thinking about which one is most suitable, rather than immediately ruling out buying.
Here are some other ways you could use this principle in different ecommerce settings:
Here's how I carry out an end-to-end guerrilla user test, including preparation time, running the test, and importantly analysing the results. In under four hours.
Making sure every page or screen has clarity of purpose will go a long way towards making your ecommerce site, or any product, more effective and easier to use.
A look at how Canopy achieve a design for a curated set of search results compared to the massed results of Amazon.