Do you also not buy products from big companies? Because they also kill their own products or let them wither on the vine out of apathy.

Sure, but we're talking about probability here. Established enterprise products from large companies usually have a published roadmap and guaranteed support period years in advance. It may be a mediocre product, but a business knows Oracle isn't going to kill off their ERP system any time soon, and also knows the minimum sunset date for all versions and can plan updates & upgrades accordingly.

If Oracle ERP is your standard it might be hard to get anything modern. Thought experiment, for which product are you more confident about its longevity: Google Voice or Basecamp?

I'm just using Oracle ERP as an easy example. It's a spectrum though, Oracle ERP just happens to be on the far end of one side, while a < 2 year old startup or any new(ish) Google product is on the opposite end.

(Though even more established Google products that are not Search/Ads still never make it far past the middle)

How many people are buying Oracle ERP as a new purchase these days? Isn't that pretty much a legacy techology?

For a startup product to be "safe", the startup has to not go under (and many do) and also either not get acquired, or get acquired and have the product not change too drastically.

There's an awful lot higher chance of a startup product not being a "safe" buy compared to an established company to kill a somewhat popular product and you can always look at their track record before making a decision (something a startup, by definition, typically doesn't have).

No evidence this is "usually" true. It is sometimes true. It is also sometimes true that the acquired company thrives.

The deadpool of acquired companies runs very, very deep.

While it might not necessarily be true of 50%+ of companies that are acquired go into the deadpool, it's common enough that an acquisition of a company's product that you use should cause you at least to make sure you have some sort of strategy to deal with a potential deadpool situation. Maybe it's not a guaranteed probability, but it is indeed very likely.

Sure. Agree completely. However it's worth noting that the customer-centric view, while valid, is rarely a good one. Companies are acquired b/c they are about to go out of business, because the IP is good, because the team is good, because the customer list is good, etc. Pissing off early customers is often the cost of growing the thing to its fullest potential.

This gets said a lot, but it's not internally consistent. If you are forced to piss off the people who love you, perhaps you shouldn't call it "growing". Call it "a pivot". You're taking a flyer for the _possibility_ of future growth.