Ramp Time for Web Killers: Google to Alta Vista, X to Google

May 26, 2009

Harry McCracken’s “How Long Did It Take for the World to Identify Google as an Alta Vista Killer?” here asks an interesting question. His write up provides some examples of early positive Google evaluations in trade and news publications. His conclusion was that no one figured out how good Google was until several years raced by. I agree with his concluding remark:

A Google killer may well be out there even as we speak. We may even be saying nice things about it. But it would amaze me if we’ve figured out yet that it’s going to kill Google…

Several ideas raced through my mind as I reviewed his chronological list of early Google references; namely:

  1. Google pushed into search at a time when the leading Web sites were becoming portals, an evolutionary arc that reached its zenith with Yahoo.com and the MSN.com Web sites in the mid 2000s. Both companies were in effect mini-AOLs with search relegated to a “search box” that wasn’t all that useful or interesting to me
  2. The leading Web search engines were running aground on two well known problems to those familiar with Web indexing: the cost of scaling to keep pace with the growing volume of new and changed content and the baked in problems of traditional server architecture. Google tackled input output, failure, and cheap scaling early in its history. The company did not reveal what it did until the job was done. This put the company several years ahead of its competition at the time of its 2004 IPO
  3. Existing search vendors were looking for exits from Web indexing. The most notable challenger after Hewlett Packard muffed the Alta Vista project was Fast Search & Transfer. At the time of 9-1-1, Fast Search had indexed breaking news before Google, and the Fast Search system was, in terms of Web indexing, the equal of Google. What did Fast Search do? It sold its advertising and Web search business to concentrate on enterprise search. A decision that cut a path to the financial quagmire in which Fast Search became stuck and the police action about which most people know nothing.
  4. Other search vendors ran out of cash, ran into index updating problems similar to those encountered by Excite and Lycos, or changed business direction.

Google’s emergence, as I have written in my Google trilogy here, was a combination of several factors: luck, technical acumen, talent availability from the Alta Vista effort, and business savvy on the part of Google’s investors. Killing Google, therefore, will take more than a simple technical innovation. A specific moment in time combined with other ingredients will be needed.

For some of the big players today, time has run out. A Google killer may be in someone’s garage, but until the other chemicals are mixed together, the GOOG has won. Every time I make this statement, I get howls of outrage from conference organizers, venture firms, and pundits. I stand by my claim that Web search is not effectively in Google’s paws. Let me excite some readers on a related front: Google is poised to pull the same 70 percent market share trick in other business sectors. Digital goodies from Yahoo and the Microsoft Bing Kumo play notwithstanding, embrace Googzilla or stay out of its way.

Stephen Arnold, May 26, 2009


4 Responses to “Ramp Time for Web Killers: Google to Alta Vista, X to Google”

  1. FoxJudsf on June 5th, 2009 6:20 pm

    Good, interesting article, but where took information?

  2. Stephen E. Arnold on June 6th, 2009 9:45 pm


    Yikes, looks like I made a link error. I will try to locate the original source.

    Stephen Arnold, June 6, 2009

  3. Reynaldo Mau on November 17th, 2010 9:46 pm

    One other very robust and powerful post. I’ve been reading through a few of your earlier posts and finally determined to drop a touch upon this one. I signed up to your publication, so please sustain the informative posts!

  4. Stephen E. Arnold on November 20th, 2010 10:59 am

    Reynaldo Mau,

    If I write something coherent, that’s probably an accident.

    Stephen E Arnold, November 20, 2010

  • Archives

  • Recent Posts

  • Meta