Search Turbocharging: A Boost for Search Company Valuations?

January 13, 2008

PCWorld’s January 12, 2008, story “Micrsoft’s FAST Bid Signals a Shift in Search.” The story is important because it puts “behind the firewall” search in the high beams.

A Bit of History

Fast Search & Tansfer got out of the online Web search and advertising business in early 2003. CNet covered the story thoroughly. Shortly after the deal either John Lervik or Bjorn Laukli, both Fast Search senior executives, told me, “Fast Search will become the principal provider of enterprise search.” In 2003, there was little reason to doubt this assertion. Fast Search was making progress with lucrative U.S. government contracts via its partner AT&T. Google’s behind-the-firewall search efforts were modest. Autonomy and Endeca each had specific functionality that generlly allowed the companies to compete in a gentlemanly way, often selling the same Forture 1000 company their search systems. Autonomy was automatic and able to process large volumes of unstructured content; Endeca at that time was more adept at handling structured information and work flow applications. Fast Search was betting that it could attack the enterprise market and win big.

Now slightly more than four years later, the bold bet on the enterprise market has created an interesting story. The decision to get out of Web search and advertising may prove to be one of the most interesting decisions in search and retrieval. Most of the coverage of the Microsoft offer to buy Fast Search focuses on the here and now, not the history. Fast Search suffered some financial set backs in 2006 and 2007, but the real setback from my point of view is in the broader enterprise market.

Some Rough Numbers for Context

Specifically, after four years of playing out its enterprise strategy, Fast Search has fallen behind Autonomy. That company’s revenues are likely to be about 30 percent higher than Fast Search’s on an annualized basis, roughly $300 million to $200 million over the last 12 months. (I’m rounding gross revenues for illustrative purposes.) Endeca is likely to hit the $90 to $100 million target in 2008, so these three companies generate collectively gross revenues of about $600 million. Now here’s the kicker. Google’s often maligned Google Search Appliance has more than 8,000 licensees. I estimate that the gross revenue from the GSA is about $350 million per year. Even if I am off in my estimates (Google refuses to answer my questions or acknowledge my existence), my research suggests that as of December 31, 2007, Google was the largest vendor of “behind the firewall” search. This estimate excludes the bundled search in the 65 million SharePoint installations and the inclusion of search in other enterprise applications.

One more data point, and again I am going to round off the numbers to make a larger point. Google’s GSA revenue is a fraction of Google’s projected $14 billion gross revenue in calendar 2007. Recall that at the time Fast Search got out of Web search and advertising, Google was generating somewhere in the $50 to $100 million per year and Fast Search was reporting revenue of about $40 million. Since 2003, Google has caught up with Fast Search and bypassed it in revenue generated revenue from the enterprise search market sector.

The Fast Search bet bought the high octane performance Microsoft bid. However, revenue issues, employee rationalization, and eroding days sales outstanding figures suggest that the Fast Search vehicle has some mechanical problems. Perhaps technology is the issue? Maybe management lacked the MBA skills to keep the pit crew working at its peak? Could the market itself changed in a fundamental way, looking for a something that was simpler and required less tinkering? I certainly don’t know.

What’s Important in a Search Acquisition?

Now back to the PCWorld story by IDG’s Chris Kanaracus. We learn that Microsoft got a deal at $1.2 billion and solid technology. Furthermore, various pundits and industry executives focus on the “importance” of search. One type of “importance” is financial because $1.2 billion for a company with $200 million in revenue translates to six times annual revenue. Another type of importance is environmental because the underperforming “behind the firewall” search sector got some much-needed publicity.

What we learn from this article is that “behind the firewall” search is still a highly uncertain. There’s nothing in the Micrsoft offer that clarifies the specifics of Micrsoft’s use of the Fast Search technology. The larger market remains equally murky. Search is not one thing. Search is key word indexing, text mining, classifying, and metatagging. Each of these components is complicated and tricky to set up and maintain. Furthermore, the vendors in the “behind the firewall” space can change their positioning as easily as a n F-1 team switches the decals on its race car.

Another factor is that no one outside of Google knows what Google, arguably the largest vendor of “behind the firewall” search will or will not do. Ignoring Google in the enterprise market is easy and convenient. A large number of “behind the firewall” search systems skirt Google or dismiss the company’s technology by commenting about it in an unflattering manner.

I think it’s a mistake. Before the pundits and the vendors start calculating their big paydays from Microsoft’s interest in Fast Search & Technology, Google cannot be ignored; otherwise, the dip in Microsoft shares cited in the PCWorld article might like a flashing engine warning light. Shifting into high gear is useless if the engine blows up.
Stephen E. Arnold
January 14, 2008


One Response to “Search Turbocharging: A Boost for Search Company Valuations?”

  1. Search Turbocharing: A Boost for Search Company Valuations? on January 19th, 2008 6:45 am

    […] Search Turbocharing: A Boost for Search Company Valuations? Google’s GSA revenue is a fraction of Google’s projected $14 billion gross revenue in calendar 2007. Recall that at the time Fast Search got out of Web search and advertising, Google was generating somewhere in the $50 to $100 million … […]

  • Archives

  • Recent Posts

  • Meta