However, teams and products that do manage to make a name for themselves potentially end up leaving more unhappy people in their wake when they're 'acquihired'. It's not quite 'fail fast' from the point of view of the users who've been around for a while. (NB I'm not suggesting that the Posterous team could/should have done anything differently.)
Edit: FWIW each acquihire-type story nudges me slightly more to consider open-source alternatives, even if they involve some pain to set up/manage. Purely to feel a little more insulated from scenarios like this.
Right. While it was certainly unfortunate that Posterous was acquihired the impact from a saas/paas closing would be magnitudes more disruptive.
Everyone comes across as a winner and doing well with no chance of going under until it happens. Imagine building something using mailgun,appfog, engine yard or (pick whatever you want) and then having to scramble (even over the course of several months) to try and replace the functionality.
A good analogy might be the crappy customer service provided by AT&T, Comcast and [insert most utility cos] over the past 20 years. Most consumers now expect poor service and I'd wager that has, to some degree, had an influence on purchasing decisions.
It just is risky to do business with startups.
I think this is the main differentiating point between 'failing' and 'acquihires' (from a user's perspective).
If a company is simply shutting down, there's some sympathy for the team and it's known that the service will stop. With a talent acquisition, nobody really knows what's going to happen to the service (maybe even the founders don't know) and that extra level of uncertainty is off-putting. [1]
Eventually, that extra level of uncertainty transfers over to the List Of Reasons Not to Work With Startups(TM).
[1] This is also true of straight acquisitions but there's a reasonable expectation that the acquirer is purchasing the product as well as the team.
Or in other words, I would sacrifice many many startups to get EtherPad back again.
In semiconductor startups, you sometimes need to secure "second source" agreements with a bigger business (that could be your eventual acquirer ) that assures they'll takeover manufacture your existing silicon if they go down.
In our case we had a second-source partner until our acquisition that had "rebadged" parts available. It was annoying, but it's sometimes the only way for a startup to sell to big multi-million dollar customers.
It's even worse if you're selling to enterprises, and was worse long before anyone ever thought of the term "acquihire". You can trounce your competitors on every axis, come in at half the price, charge a flat-rate site license instead of a metered seat license, and buy the purchasing team 52-day aged ribeyes every week for 6 months and still lose if you're the one who looks like you might be out of business 2 years from now.