Auto-generated description: Rows of brick rooftops with chimneys are set against a backdrop of trees and distant hills.

I’m grateful to Tom Watson for sending me this link to Richard Pope’s blog, author of the book Platformland. Pope talks about specific examples from UK public services to make the point that starting with very specific use cases is problematic.

First, because expecting them to scale to other use cases is a form of magical thinking; second, because people change over time.

The first problem with use-cases is that, because they limit the chaos, when you scale you often discover that you have scaled a service that only works for those use-cases. You then have to confront the chaos at scale. Many of the hardest delivery problems or riskiest assumptions may go untested until it is too late.

[…]

The second problem with use-cases, is that complexity is not static. The needs of users change because people’s lives change, political priorities shift, technology evolves. Like levees on a river, use-cases create coherence in the landscape by holding back an inherently chaotic system. However, at some point, levees can get overwhelmed.

[…]

Most services with tightly defined use-cases experience this in some shape or form. There are always people who fall outside the bounds of what a service can handle. And if it’s a public service, not serving these users is not an option.

Source: Platformland

Image: Richard Bell