I worked in Search and on Search-related problems for nearly 4 years so here's my 2c. For most users Google is "muscle memory". It is quite good for most Searc

I worked in Search and on Search-related problems for nearly 4 years so here's my 2c. For most users Google is "muscle memory". It is quite good for most Search needs. The angles of attack that tech-savvy folks think are amazing (category filters, privacy, fancy CoverFlow UX, bigger index - all real differentiators tried by real companies) - most normals don't give a rat's ass about. IMO the only real ways to compete are as follows: a) build something specialized that is several-x better in the specialty (Kayak) b) build something on top of data that Google doesn't have (Facebook Search, whenever it stops sucking) c) go head-to-head with nearly equivalent quality and buy your way in on the CapEx and the Ad/Mktg dollars required to change muscle memory (Bing) Beyond that, most companies have had trouble w/ direct search because Google + Wikipedia gets the job done, by and large. SearchMe? Dead. Kosmix Search? Pivoted successfully before acquisition. Cuil? Dead. Mahalo Search? Pivoted several times away from Search. Qwiki? Pivoted successfully. The sad thing is that Search is such a fun and challenging space with so much work to be done; I want the space to be more competitive but most consumers like the status quo just fine. Lastly - a salute to @yegg for taking this one on. He's a far braver than I, and most others on this thread critiquing his work.