Your comment about Stadia gave me this thought. What odds would I be willing to bet that Stadia would be around in 5 years. I wouldn't take a even money bet. Maybe 5:1 that they'd still be functioning in 2025?
(edit, made the odds clearer)
It's an interesting question, I think i'd give it less than 10% but above 1%. It's outside their core competency, it involves them negotiating with rightsholders, and there's a bunch of competition coming from existing companies in the space (e.g. Microsoft, NVidia, etc.).
As I can see, their only edge is that they own a global content delivery network which in the best case might be able to give them a few years' head-start in terms of performance vs. other players that don't own IP infrastructure. (It's a recent development that cloud-hosted gaming is even viable from a latency perspective). But that doesn't seem like the core long-term problem; it might give you a first-mover advantage but I don't see that being enough to build a defensible moat in this space.
It’s debatable whether many people outside HN actually hold a grudge; Google’s overall favorability ratings are crazily high. But I wouldn’t be surprised if Stadia had a rough launch because of Google’s reputation.
Shoes are a bad analogy because there is little switching cost; you just ties a different pair of laces and you’re done. If you had five years of playlists on G Music and now you have to move them to Spotify, that’s a major project. (Source: I did this, and it took me many hours).
Put differently, as a happy customer, you don’t want the company to pivot. It’s usually bad news for all of the existing happy customers. A successful pivot is great for the company, hopefully great for the new happy users, and very disruptive for the old customers. In the best cases the net utility is even positive.