Saul Hansell's latest article in today's New York Times offers a rare glimpse into the "Search Quality" team at Google, who oversee the main algorithmic engine that makes Google, Google (Techmeme link here).
Housed in building 43, the secretive group headed by ex-Yahoo! and Amazon search guru Udi Manber, the group constantly tweaks, tests, and tortures the knotty, growing glob that make up the "Ranking Algorithm".
Here's how the article introduces us to this group:
"...Amit Singhal and hundreds of other Google engineers are constantly tweaking the company’s search engine in an elusive quest to close the gap between often and always.
Mr. Singhal is the master of what Google calls its “ranking algorithm” — the formulas that decide which Web pages best answer each user’s question.
It is a crucial part of Google’s inner sanctum, a department called “search quality” that the company treats like a state secret.
Google rarely allows outsiders to visit the unit, and it has been cautious about allowing Mr. Singhal to speak with the news media about the magical, mathematical brew inside the millions of black boxes that power its search engine."
But unlike that secret formula, which maybe gets tweaked and tortured once in a while to give us concoctions like Diet Black Cherry Vanilla Coke, Google's formula is "improved", constantly, in near real-time (image source).
(As an aside, this similarity is one reason alone why Warren Buffett's Berkshire Hathaway should consider adding Google to his Coke position, despite the notion that he doesn't "get technology"...I know, I know, his friend Bill might not like that very much, but this is BUSINESS, not personal)*.
The piece is a great glimpse into how Google is inventing it's future around one of it's core competencies. (As an aside again, the other of course is it's algorithmic genius around MONETIZING the "ranking algorithm", which could use a whole separate article by the New York Times).
Especially at a time, when there are so many startups trying to out-do Google in it's core competency.
The latest VC-backed, Search effort is an interesting case in point.
Here's how Mahalo's FAQ page describes the service:
"Mahalo is the world's first human-powered search engine powered by an enthusiastic and energetic group of Guides. Our Guides spend their days searching, filtering out spam, and hand-crafting the best search results possible.
If they haven't yet built a search result, you can request that search result. You can also suggest links for any of our search results."
Asked about Mahalo in his interview at the D Conference, Google CEO Eric Schmidt replied:
"...human editors can be approximated with algorithms."
As the New York Times article illustrates, the company's approach to improving it's search algorithms centers around hundreds of Search Quality specialists testing endless number of variables that drive the "best" search results.
It's an effort that contrasts sharply with those of startups like Mahalo, that tries to define the "perfect search" BEFORE the search.
Google's approach focuses on endlessly improving on the "perfect search" AFTER countless searches on millions of items that hopefully give results that are "good enough" initially, and get better over time.
These are entirely different approaches.
And that's what gives us such an interesting horse race.
(*DISCLOSURE: I own shares in both Microsoft and Google).