What I think the parent poster is saying is that in hindsight, you can easily say which customers you were right to pivot for and which your were wrong to pivot

What I think the parent poster is saying is that in hindsight, you can easily say which customers you were right to pivot for and which your were wrong to pivot for, but going in from the front no one, and I mean NO ONE can actually, really tell which customers are weird customization cases and which customers are revenue-bridges that will lead you to better market fit. There are always obvious cases in this mind of thing, but the larger point is that we could easily have been looking at an alternative history where Uber was on the scrap heap, and we would all stand here, wagging our beards, berating them for "moving from niche use-case to niche use-case" instead of doubling down and building something of long-term use and value, of executing on their vision. Instead, we're saying that Uber was an instance of a GOOD pivot, but at Uber, at the time it was happening, anyone could've made the argument, "Hey, aren't we just filling a really niche use-case here, of regular people who are willing to trust their lives to unlicensed strangers for a ride?"