How would you bootstrap such a company?

The solution as I see it is to forcefully break-up Google, such that these services can have their run as independent companies (Maps, flight search, product price comparison, etc.). This has been done before with big companies (e.g. break of Bell Systems into 7 companies: https://en.wikipedia.org/wiki/Breakup_of_the_Bell_System)

The way Google: ship it and make it work really, really well.

The product was so good that people noticed. They were able to take this and leverage this into being search for a really large property, like

Back in the 90s there were lots of search engines. People didn't pick Google because it was hard to switch, they converged on Google because the product was that good - there was Infoseek, HotBot, AskJeeves, Alta Vista, Lycos, Yahoo, Excite ... the list goes on.

It's a theoretical problem to have a significantly better search engine and not find users for it.

I'd argue that Google was able to win specifically because they had so much competition. They weren't fighting against one or two massive entrenched competitors in an established space. Search was still pretty new at that point and Google's approach to it was innovative.

Google also had lots of space to improve on its competition when it came to metrics like latency, which paid off. Nobody really has the option to beat them on latency now. That's not really Unfair in any fashion but it is an example of how it would be much harder to dethrone the ruling search engine in 2020.

> People didn't pick Google because it was hard to switch, they converged on Google because the product was that good

That's not how I remember things from that time. I started to use Google, because it was faster than dogpile and the other oddly named search engines.

It wasn't a quality reason. There was the expectation that a search for information would take longer than a single click, so a little digging wasn't an issue.

My experience was that the quality of the search results were much better.

I experienced that as well, but the reason we used Google at that time was the speed that was sorely lacking is all other offeri we used.

Most of those independent companies wouldn't have a good revenue stream no? I really doubt Google Maps is sustainable simply by selling API access.

It and especially the more niche google products seem to be provided as a value add to their other services (i.e search in this case).

Awww, you mean parts of the monolith would have to compete on merit with all the other companies that are having to compete on merit?

What a terrible shame...

It will be a terrible shame when I no longer get real-time traffic on the small streets, yes.

I use OSM multiple times per week, but I still start GMaps anytime I need to drive somewhere.

Google Maps is essentially TripAdvisor with better search, but fewer reviews. In addition, Google Maps is probably the most used GPS routing software these days.

They would do fine standalone, either following a path like TripAdvisor or Garmin or both.

And what good came out of the remains of Bell Systems afterwards?

DDG has an IRC channel on Freenode, I went one day asking about result quality when searching for Python stuff because it used to send me on documentation pages for old versions.

Something happened, because a few weeks later it appeared that DDG started getting me results for the current version of Python.

You might want to try to reach out to DDG if you have some improvement suggestions.

I just use both DDG and Google if needed via !G. That way i get more results from different sources.