When Expedia, the online travel platform, started, they were working on making the experience as snappy and convenient for their users as possible.
One of the suggestions to make the experience faster was:
The moment a traveler enters their origin and destination for flights, run an API call in the background to fetch the results, even before the person has clicked on the Search button. So when they click on Search, the results are already there, without any wait time. This would result in a much snappier experience for the users.
But what Expedia ended up finding out is that the drop-off rates actually increased after they implemented this design hack — something that they clearly didn’t foresee coming.
Naturally, the Expedia design team wanted to find out what was causing this increase in users dropping off from the website instead of going ahead with their bookings. And what they found out is that users were thinking that the system was unintelligent: that it wasn’t doing any work to fetch the best results. The users expected that such an activity requiring intelligent curation would take more time. And since the system was serving results almost instantaneously, they felt that the system was unintelligent and wasn’t doing the work it was supposed to do.
So much so that the Expedia team had to add an artificial delay of two seconds to make them trust the Expedia application and fit their expectations better! They had to add complexity to improve the UX.
A lot of UX in the real world has to do with psychological conceptions and underlying motivations and models of the world than it has to do with straightforward optimizations.
Rory Sutherland puts this across brilliantly via an example:
“The question was given to a bunch of engineers: how do we make the journey to Paris better? They came up with a very good solution: spend £6 billion building completely new tracks from London to the coast, knocking 40 minutes off a 3½ hour journey. But it strikes me as an unimaginative way of improving a train journey merely to make it shorter. What you should in fact do is employ all the world’s top male and female supermodels, pay them to walk the length of the train handing out free Château Petrus for the duration of the journey. You’ll still have about £3 billion pounds left and people will ask for the trains to be slowed down.”
Simple is not always the fastest or the most optimal. Simple is what works flawlessly, while fitting the expectations of the user.
For example, say a university came up with an intelligent algorithm to filter admissions based on merit, without the students having to write long essays or give entrance tests.
My hypothesis here would be that people would trust such a system far less, because their existing mental model tells them that to enter any good school worth entering, you need to put in a lot of effort. And the playground on which students are competing against each other to get admitted should be transparent.
If someone is taking you in that easily without you having to put in a lot of effort, is the school even worth going to?
Here’s something for you to think about:
Fintech apps today disburse loans with minimum effort on the borrower’s end. Getting a loan has become a simple two- to three-click process. How does this affect customer perception around loans and entities giving such loans? Does it significantly change human behaviour around how they see debt in the long run?
People aren’t used to getting credit this easily in the history of civilization. What happens when such a thing is made incredibly quick and effortless? What are some second- and third-order effects?
Write about it on your LinkedIn or Twitter and tag us. If we like your response, we’ll feature your take in an upcoming newsletter.