Ok, but Google didn't just start a search engine to compete with Lycos. They struck oil with a new approach to search.
It's hard to say "any argument that proves it was a bad idea to start Excite" because, you know, Ferraris and hookers and coke and all that. A great idea.
But if you're taking on Microsoft, Google and Adobe, you'd better have an awesome idea, and you probably don't need the "this is a good market" encouragement.
Google didn't just "strike oil." The idea that they discovered a magical new algorithm that guaranteed their success is an urban legend. The main thing that made Google different from the portals was that they actually cared about doing search well. (When Google was started, their predecessors were trying hard to get people to stop calling them "search engines.") So there is plenty of room for a new startup to have the same sort of advantage over Google etc in web-based apps that Google did over preceding search engines.
My understanding is, Google didn't start as a "search engine that cared about doing search well". It started as a company to capitalize on Page and Brin's doctoral work. Their original plan was to be an arms dealer for other search engines. It was an idea first.
"search engine that cared about doing search well"
That quote is not a quote, is it?
Thomas, I'm done with this thread. I'm sure you mean well (at least, I'll give you the benefit of the doubt), but I've noticed before that arguing with you is exactly like arguing with a troll. The defining quality of both being finding oneself always saying "No, what I said was...."
You said "they struck oil with a new approach to search." I believe nobody, not even other search companies cared about the algorithm. The search engines did not want to buy PageRank (outright, not just license), not even for $1M. Not even during the Dot Com craze. Nobody wanted to improve search, in other words. That's what PG is writing about.
Had Larry and Sergey received $1M, then I would agree that they would have struck it rich, and probably continued their doctorate studies. But instead they decided to build a company around it even though no other company believed they had anything worthwhile to, in the order of increasing significance, 1) purchase, 2) invest in, 3) license, or, most insulting of all, 4) not even pay their own employees to copy (at least not until 4-6 years later.) And yet, they decided to drop out of the best Graduate Computer Science program and try to start a great company because existing portals were not doing search well, and nor did they care. Only a long time later did Yahoo! decide to license Google search (nobody else did) and, only even later did Microsoft and Yahoo! started paying their own employees to clone Google's search tactics (as well as look and feel.)
=
1998: You shouldn't try to build new search engines because Lycos, Excite, and Altavista are already doing it.
Any argument that proves it was a bad idea to start Google is probably a bad argument.