[swiftcoder]

> Shipping stuff that turns out to just be useless clutter absolutely happens - the problem is it is hard/impossible to know the difference going in.

It's not that hard. Plenty of A/B testing is done along the way, and yet useless clutter still ships. Why? Partly because it's pretty easy to goose one or two metrics at the expense of a small global regression, partly because everyone needs shipped product at the end-of-year review to justify bonusses/raises/promotions, partly because there are 100+ individual teams working on the product and they all think their feature will benefit users more than the next one... a multitude of localised decisions will necessarily gradually erode the product experience as a whole.

You see this most starkly in fairly decentralised companies like FB, Google, or Valve, less so in companies with centralised command-and-control over product experience (Apple, Amazon's device org, etc). But even centralised command-and-control can only hold this is check for so long at massive scale. Amazon's larger and less focused divisions, post-Jobs Apple... all on the slippery slope

[tqi]

> Partly because it's pretty easy to goose one or two metrics at the expense of a small global regression

That is true, but again I think it's more complicated than that. AB tests are good for measuring incremental changes, but really bad at predicting the impact of 0 to 1 type products / features, because so much of the success or failure depends on public perception and network effects. Twitter couldn't AB test "Fleets" and have any confidence as to what the product adoption would look like. So companies have no choice but to launch and see.