“Most of our users are on Desktop”

Whenever someone says “our analytics tell us most users are on desktop”, I always think of this diagram.

Back in WW2 they wanted to work out why planes were being shot down, so when planes returned, they noted where the bullet holes were most present, with a view to further protecting those areas on the plane.

On the surface, that checks out right? But what they missed was these were the planes that returned. Planes that got shot in the engines (notice how the engines have no dots on the diagram) never returned. They missed out this crucial pattern because they were looking at the data all wrong.

It’s classic survivorship bias.

Anyway, how is this related to “most of our users are on Desktop”?

Yes, your analytics may skew towards desktop usage, but is that actually accurate? Is your site super slow on mobile devices? Does it even work on mobile devices? Do mobile users land on your non-optimised site and bounce out to a competitor before your analytics load?

Sure, some sites are desktop only, but while mobile usage continues to grow and desktop continues to fall, it makes sense to at least dig deeper to make sure you know why you have more desktop usage.