Would-be Google (Nasdaq: GOOG ) slayer Cuil -- pronounced "cool" -- had a rough debut. Outages and errors stalled the hopes of those looking for a search alternative to the Big G.
Yet it would be premature to write Cuil's obituary: 50 million queries touched the company's servers on day one, CEO Tom Costello told me in an interview yesterday. That's not much when compared to Google, which averaged 236.5 million search queries per day in June, or Yahoo! (Nasdaq: YHOO ) , which averaged more than 80 million, according to Web tracker comScore. But Cuil was likely better than Microsoft (Nasdaq: MSFT ) , which averaged roughly 35 million searchers per day last month, and IAC's (Nasdaq: IACI ) Ask Network, which yielded less than half that over the same period.
Costello points this out during our chat not as an excuse so much as an explanation -- Cuil wasn't ready for the buzz it had created.
Anatomy of an upstart search engine
Was the attention deserved? I'm not as sure, but I will admit to being impressed by the idea of Cuil, which (ahem) seeks to cull meaning by searching pages for multiple uses of entered search terms. Content, Costello says, is a more likely predictor of meaning.
"Google is good at doing one good thing," Costello says. "Giving you back results based on popularity." Logically, that should lead to results that contain far too much extraneous hoo-ha. And it does from time to time, in my experience.
Cuil has built an infrastructure aimed at limiting the hoo-ha. Instead of using thousands of machines to break apart the Web into the thousands of searchable chunks, and then searching them all concurrently, Cuil relies on 150 machines.
The trick, Costello says, is to make the index more intelligent by giving otherwise dumb servers areas of specialty. Search for "Tim Beyers" and "The Motley Fool" and you'll be routed to a specific cluster -- servers with my name on them. Now, that's cool.
What I'm wondering is if it's enough. Software can be brittle, as Cuil's opening day demonstrates. And while 150 machines, managed efficiently, appears to be more than enough right now, it seems absurdly low when compared to incumbents and insane in light of the Web's reputation for mocking sites with massive outages. Just ask Netflix (Nasdaq: NFLX ) , or any of the addicts who use Twitter. So frequent are hiccups in that service that an enterprising entrepreneur has created Twiddict, a digital ambulatory for the terminally connected.
Costello says he understands the risks but he's also a rebel. He says he wants his company to invent the Next Big Thing in search and make real the truths of tech investing -- the ones that say industry incumbents can never be safe.
"I understand why they haven't changed," Costello says, referring to Google and its 39% revenue growth in the second quarter. Why tinker with a business model that produces more than $1 billion in free cash flow every three months? Few firms short of Microsoft and Oracle (Nasdaq: ORCL ) achieve such heights routinely.
History provides the answer for Costello, who concludes our interview with a confession: He once worked for IBM (NYSE: IBM ) . His years of experience there and elsewhere in the industry appear to have shaped a worldview that sees the status quo as an enemy. "I never want to work on a mature product," Costello says.
Google? Mature? Perhaps. One thing's for sure: Cuil -- the upstart, the young'un -- is anything but, and that seems to be the way Costello and his cohorts prefer it.
Be cool and check out related Foolishness: