After months of anticipation, the engine launched at 12:01 a.m. this morning. There’s already considerable discussion on Techmeme. Danny Sullivan has written a comprehensive post at Search Engine Land. The early consensus is that Cuil is not a “Google Killer” (let’s banish that construction: Google Killer, iPhone Killer, Craigslist Killer).
The servers have crashed and so I only got a couple minutes to play with it before writing this (they’re back up now apparently). I was briefed with the rest of the Search Engine Land crew last Wednesday but nobody had a chance to conduct searches until this morning.
If there are three key variables in search — relevance, index and UI — Cuil has done interesting things in each category. There’s a different methodology for ranking being used (explained in some detail in the Danny Sullivan post). The Cuil index claims to be much bigger than Google’s; although Google got wind of it and posted this to preempt that claim. The UI offers a two or three column format at the election of the user, as well as some other interesting features. Anna Patterson, Cuil co-founder and president, likened it to a magazine.
Relevance is a very hard thing to measure, especially given how many sites and pages are online now. One could certainly do systematic testing of a large number of queries and come to conclusions about Cuil’s relevance vs. other engines. But most users won’t. Thus the UI becomes very important in getting attention from users and maybe getting them to come back. Having said that, Ask’s very innovative “3D” interface has apparently done little to improve its position in the market.
I was impressed by Patterson and her attitude about search and the broader search marketplace. She is a search veteran, along with other members of her team; and they don’t seem to be in it for the money. They’re taking a long-term view and understand that Google wasn’t built in a day. This attitude and the patience of their investors, which I specifically asked about, will serve them well.
Powerset had some impressive features and capabilities but no way to fund its expansion to the entire Internet; hence the Microsoft acquisition. Cuil has indexed a huge portion of the Internet and now has to work on refining features and the presentation of results. (Some argue relevance too.)
In this world of instant analysis that we’re now living in, people are quick to pronounce winners and even quicker to pronounce losers. Let’s give Cuil a year or two to see what happens. Then it will be fair to predict whether or not Cuil will potentially ascend to the top echelon of search or become another also-ran.