> It failed because it wasn’t small enough.

I've never seen a preference like this, in real life. Usually the thing closest to what you want is the preferred option. You're suggesting there's a hump in the preference curve, pushing people away from their preference, buying a larger phone than the smallest, when they "want" a smaller one.

I have trouble believing this is true. Do you have any other example of this type of preference curve? I suppose the "uncanny valley" may be one, but that seems more understandable.

I'm not sure I'd call it a curve.

Small phone vs. larger phone is a very simple tradeoffs calculation.

Large phone: good screen, bad ergonomics Small phone: small (thus worse) screen, best ergonomics

I'm willing to pick the second option above.

Unfortunately the Mini is somewhere in the middle: smaller screen than the larger phone - thus worse in that aspect -, combined with worse ergonomics than an actually small phone. It's the worst of both worlds.

I don't know about other things, but ever since the iPhone 5 I've been wanting another model that I could use with a single hand. The Mini was never that, so why would I sacrifice a good feature (larger screen) for... nothing in return?

Also, increasingly shrinking the screen places increasing demands on apps and app developers to support those dimensions.

It's not like enlarging the screen where you can at least generalize it by scaling everything up and it's still useable.

With shrinking screens, you have to decide on tap target and content size minimums. It's quite an undertaking that needs to pay in the market.