Five psychology principles that make an app feel better

You open an app and it just feels good. You open a competitor with the same features and something is slightly wrong, but you can't say what. I used to put that down to design taste. After watching Wyatt Feaster's video on app psychology, I think a lot of it comes from how well the product matches the way people's brains already work.

He covers five principles. Here they are, with what I'd change in a product because of each one.

Goal gradient effect

People speed up as they get close to finishing something. A runner who wants to quit at kilometre 35 finds extra energy once the finish line comes into view.

Apps use this with progress bars, streaks and onboarding checklists. "Complete this task" gives you nothing to aim at. "5 of 7 tasks done" or "3 minutes until your onboarding is complete" tells you the end is close, and that pulls you forward. Feaster points out that onboarding flows with a progress bar the whole way through tend to have higher completion rates than the same flows without one.

For any flow with more than two steps, I show the real number of steps and where the user is. The count has to be honest. A bar that jumps to 80% and then stalls does more damage than no bar at all.

Peak-end rule

We don't remember an experience evenly. We remember the best moment, the worst moment and how it ended. You remember the steakhouse where the steak and the service were both outstanding, and you remember the night the food was bad and the waiter was rude. The perfectly fine dinners in between are gone.

Airbnb and Duolingo put a lot of effort into success screens and small rewards for this reason. It works in reverse too. If a long flow breaks halfway, people remember the frustration long after they've forgotten the parts that worked.

So I spend design time unevenly. The screen after a user finishes something important gets real attention, and the worst point of friction in the product gets fixed before anything gets polished.

Labor illusion

We value things more when we can see effort going into them. It's part of why a marathon or a meal that took three hours feels worth more than something quick.

Feaster's example is identity checks. Some apps show a short "verifying your licence" or "running background check" step even when the check finishes instantly. Sensitive tasks that finish instantly can feel careless, and people want to see that something is actually happening before they trust the result.

I only use this where trust is at stake: identity checks, payments and permissions. In those places I show what the system is doing while it does it. Adding delay to a search box or a settings page just makes the app slower.

Von Restorff effect

The item that looks different is the one people remember. In an ice cream shop with thirty flavours, you remember the strange one and forget the chocolate and vanilla.

In an interface, a long list where every item has the same weight gives the user nothing to hold on to. Good apps pick the few things that matter and make them look different, with an icon, an illustration or more space around them.

When I lay out a screen, I pick one primary action and make it clearly stronger than everything else. When a screen has four equally bold buttons, the user has to work out for themselves which one matters, and many won't.

Choice overload

A menu with a hundred dishes makes ordering harder. You feel a bit of tension, you can't decide, and you might end up ordering the thing you always order. Apps do the same to people with crowded navigation bars and settings pages that expose every option at once. The user feels overwhelmed before they've tried anything.

The fixes Feaster lists are search, filters, grouping things into categories, and removing items altogether.

I start with the fewest choices a screen needs and reveal more only when the user asks for them. If a screen can't be used without search, I take that as a sign the structure behind it is too flat.

Where to start

If your app feels off and you can't say why, go through it with these five in mind. Check whether people can see how close they are to finishing, what the last screen of each important flow looks like, whether sensitive steps show their work, whether each screen has one obvious action, and how many choices each screen asks for. In my experience at least one of those points straight at the problem.

Source

The principles and examples come from Wyatt Feaster's video The psychology trick that makes any app feel 10x better. The notes on what I'd do with each one are mine.