When designing for edge cases, let's call a spade a spade.

A good design is geared towards its target audience; it is built assuming that audience has some base of knowledge.

A colleague recently sent me a story about an elderly gentleman in the North of England, who had been experiencing postbox UX fail for around 2 years. Every time he went to post his mail, he did so (unbeknownst to him) into the local dog poop box. This happened because the man was near-sighted, and the poop box was around the same colour, size and shape as a U.K mailbox.

This got me thinking — in cases like this, do we phone and complain to the poop box design team, or scratch this up as an impossible case to anticipate? When designing something, does trying to anticipate all possible edge cases cripple its usefulness to the majority?

If we take the poop box as an example, we could - in an imaginary design review - say: “hey wait a sec, what if somebody hard of sight accidentally mistakes this for a mailbox? Let’s put a proximity sensor on it and, when somebody comes close have a looped audio recording that loudly states it is a poop box.”

Okay, this would avoid the case of the elderly gentlemen, but would throw up a number of other issues, like: it would need regular maintenance to check the sensor is working; it would need electricity to work; it would be noise pollutive in a quiet area.

To conclude.

A good design is geared towards its target audience; it is built assuming that audience has some base of knowledge. Trying to anticipate all edge cases risks making a design over-complex, to the point of uselessness.

You wouldn’t expect to buy a spade and have it adorned with stickers exclaiming: “hold this end”, “put this end into the ground” or “recommended for earth and other soft materials”. If marketing it to visiting aliens, maybe we’d need to dig the stickers out, but a spade’s target audience (most humans) know what a spade is, so to adorn it with instruction would confuse its users.