F2F #67: Similarity kills innovation

First off, let me apologise for not hitting send last Friday. This post was left abandoned as a draft, so this belongs to last week! Let's get on with it!

I have been testing Comet, Perplexity's browser, and there's a subtle problem with it: it looks 99.9% like motherfucking Google Chrome.

While that should come out as a win to many, it certainly isn't for me. In practice, it might not be for a lot of people either.

There's a principle in UX that helps to explain adoption of new apps if they share commonalities with things the user is already familiar with.

The UX principle describing positive similarity between platforms or objects - reducing the need to learn new concepts when switching from one to another - is known as transfer-appropriate processing or more commonly, positive transfer in UX and cognitive psychology.

That is, over time, we've grown used to pulling down to refresh on mobile, or that the diskette icon was used to save your work, or that double tapping on something likes that thing. Most recently, we've learn how to use reactions everywhere, for instance.

The problem with Comet is that it looks too similar. In fact, too similar doesn't fully explain what I perceive. My brain still thinks I'm using Chrome, so I behave like I'm using Chrome. I don't reach for the assistant. I don't explore the new workflows. I default to old habits because I think I am still within Chrome.

This collides head-on with one of the classic rules of software adoption: make product B look like product A so users can transition from A to B with little to no friction.

The problem is that Comet follows this rule too well. The transition is so seamless that I don't register that anything has changed. If my brain doesn't notice change, I don't change habits.

Conversely, when I tried to move to Arc, it was too radical. I had to rewire my brain so much that I had a hard time understanding it. That, and because it was also filled with bugs, forced me to go back to Chrome.

Conclusion: if you're going to replace a software (or, at least, attempt to), make yours look like that software at a ratio of around 85%. It'll be very similar, so it'll help with the adoption and transition process, but it'll also have that 15% that will remind users that there's something new going on and that they have to re-train themselves even if only a little bit: familiar enough to feel safe, different enough to get a small boost of novelty and excitement.

That tension, between comfort and cognitive friction, is where adoption actually happens.