NICK DOUGLAS — When I hear someone saying they're the next Google, I wonder: Does this shit happen in other industries? Does Bob sit around Bob's Boise Brewery and say "I'm gonna make the next Bud Light! Yep! Bob Light, baby!" Actually, that probably happens. But that doesn't help the odds of these wannabe Google-killers. The following sites aren't just grad-school projects that wisely focused on a niche. They all think they're the next big thing in search, and they're all wrong.
ChaCha: Needs lessons
Twist: "A smart search engine that 'learns.'" Harnesses power of humans (somehow), which is odd, because I thought we invented computers to do just this sort of thing.
Hubris: "A smart search engine that 'learns;'" refers to "first generation search engines." You know, like crusty old Google.
Results for "house": Decent results including the TV show, the legislative body, and real estate. (So does Google.)
Status: At least it's live, and the site's pretty.
Chance of success: 200/1. Not revolutionary, not scalable, and rough Alexa stats show it's not catching on. Hell, even Valleywag's in better shape.
Search Wikia (née Wikiasari): Today Britannica, tomorrow the world
Twist: Open-source search engine by Wikia, the company of Wikipedia founder Jimmy Wales, contrasts with the "black boxes" of Google and Yahoo.
Hubris: CEO says Wikia wants 5% of the search market. Wales: "The idea that Google has some edge because they've got super-duper rocket scientists may be a little antiquated now." Also: "Search...is broken."
Results for "house": Well we have to build a search first, no?
Status: Much-hyped, but the actual site is just a placeholder-style wiki page.
Chance of success: Five percent share? 30/1, if Wikia actually sticks with the project. Look at this this way: director Quentin Tarantino talks about a million projects before he finally makes one. Of course that one kicks ass, but don't get fooled into thinking the others will ever see daylight.
Tall Street: Will break when popular
Twist: Users can rate search results on a 1-5 scale. Through a stock market, they can also bet play-money that a site will get clicked when it's in certain search results.
Hubris: The intro page says Tall Street wants to "help out the little guy;" philosophical statements; needs a dedicated community.
Results for "house":
Status: Still the little guy. Hadn't heard about it until the founder e-mailed us. If it gets popular, won't armies of users from massive sites easily game it?
Chance of success: 5000/1. Maybe it'll show up in a sidebar in Wired: "Eight search engines that could be the next Google."
Powerset: Pay no attention to the man behind the curtain
Twist: Natural language search: Like what Ask tried to do, but with a computer parsing the questions instead of humans hard-coding answers.
Hubris: Creators brag that everyone loves the demo. Raised $12.5 million last year.
Results for "house": Powerset isn't live. But if it were, it'd probably want a question like "When is House on?" or "How much is a house in Tampa?" Then, we're told, it'd give an answer, not a list of sites with the phrase "When is House on" in them. (Yep, that's all Google gives.)
Status: In demo. Testers say it's great? Then put it in beta.
Chance of success: 5/1. So much could go wrong, but Powerset is the one engine with a good shot at making search feel good. The two tricks: getting the system to work without constant embarrassing mistakes, and getting people to ask questions like humans again. (The latter assumes we know what we want to find.)
Snap: Honey I blew up the widget
Twist: Visual search shows previews of sites. Of course, this slows down the search site, which also features an annoying drop-down that guesses your search term as you type (oh boy, I saved 0.6 seconds typing and wasted 2 seconds using the drop-down). Snap also makes uses these link previews for the annoying bubbles that pop up when you hover over links on some widget-happy blog. Did you really care what the site looks like when shrunk to 300x500 pixels?
Hubris: Slogan is "The other way to search;" About page calls Snap "the next great search engine;" led by big-swinger, big-misser Bill Gross and his incubator Idealab.
Results for "house": Result #4 is "sponsored." Read: spam. Overintegrating paid results is a throwback to the terrible pre-Google days; site previews are often outdated or missing, and the whole page runs slow as hell. Plus it's apparently running Ask's search.
Status: Fully deployed in its bloated (but slickly wrapped) glory.
Chance of success: 100/1. Snap has enough business momentum to wallow around until someone buys it to harvest the pieces. The preview feature could be useful, but it sure ain't a Google killer. (Disclosure: Snap recently sponsored a party for which I helped build the guest list. Sorry dudes, prose before hoes.)
Eurekster: Others searched for "WTF?"
Twist: Community-powered search. Hold on to your butts, this is gonna get bumpy: Eurekster powers "swickis" that live on different sites. You can find a swicki by visiting a site that has one, or by searching Eurekster's list. Then search for a term within that swicki. Or pick a term from the "buzzcloud," which shows what others have searched — not that you need a suggested term if you're already looking for a particular term.
Hubris: Eurekster's front page has four award badges, seven press quotes, and no search box.
Results for "house": First result in the swicki list is a substance abuse search engine by Aloha House. Okay, let's go there and search "pills." Decent results! Of course, there are all these mysterious buttons, bubbles, and paragraphs on either side...
Status: Over 50 thousand swickis deployed, including spam and weirdly specific topic areas.
Chance of success: 400/1. Again, some useful technology, but confusing implementation that makes for an awkward search experience. The goal isn't to make search tougher, guys.
Hakia: Yeah she ugly, but she sure can cook!
Twist: Clusters results into categories, guessing what a search means.
Hubris: It's "the Web's new 'meaning-based' search engine." Except for, you know, the other ones. The CEO asks, "Are we the Web's equivalent of the first color TV?"
Results for "house": Impressive! Headlines about the HR and White House, then a string of categorized links to "House, M.D." information: show profile, episode guides, fan sites, and more. A shame the ads drew a different conclusion and showed me real estate ads. But what's this little face at the top asking bot-like questions? And does every result page really need an "e-mail this page" link? "Gee, you know my friend Sarah would sure enjoy this search result list!"
Status: "Beta 13." At public revision 13, maybe you can give it a real version number. In any case,
Chance of success: 50/1. Hakia is actually pretty neat, and it could eventually form part of a larger, slicker search engine. You know, like Powerset.
Freebase: At least the verb is fun
Twist: It's actually a collaboratively edited database. Judging from screencaps, it feels a bit like Wikipedia but bigger.
Hubris: Actually reasonable, even if it claims the goal of turning the whole internet into its database. All the hype's coming from outside sources like think-tank owner Tim O'Reilly.
Results for "house": I'm not on the exclusive user list, so I'll just have to wait to find out. Guess that screws up my search-turnaround-time stats.
Status: Private alpha.
Chance of success: Define "success." Will it go as big as Google? 60/1. Will it prove useful? 1.2/1.
(What will replace Google? Maybe a retooled search from someone like Ask.com, or a hacked-up tool from three guys in Russia that no one's heard of. But my money's on the Balkanization of search, in which users check Wikipedia, Yelp, or Flickr for specific types of searches. No wonder Google made sure to corner the two prime search niches of maps and video.)