Fast Search Founder Slowed Down, Then Stopped
February 10, 2014
I read “Fastgründer John Markus Lervik dømt til fengsel.” Assuming the story is accurate, Dr. John Lervik, the founder of Fast Search & Transfer, will serve at least one year in prison. The issue is related to the financial reporting of Fast Search & Transfer.
In 2008, Microsoft purchased the company for about $1 billion, a deal compared to the price Hewlett Packard paid for Autonomy and about what Oracle paid for Endeca. Mr. Lervik will pay to pay legal fees. He will take appropriate legal steps to overturn the decision.
Enterprise search is a tough nut to crack technically and financially. The monetary challenges stem from the brutal costs of marketing and customer support. But these are at least as expensive as the cost of dealing adequately with technical challenges of enterprise search. For example:
- The time required to make a system deliver what the marketers assure customers are “ready to deploy” functions. Most large scale search solutions are not products. These are complex systems. Because each customer has specific requirements, the marketers do not understand that what they sold may take time to create, test, and deliver. Time is money. With an open ended problem, the cost is staggering.
- The problem of responding to crashes. When an enterprise search system flips over and dies, the cause may be the vendor, the reseller, or the client. Unfortunately the vendor takes the heat because many tech centric managers feel the “buck stops here.” Responding when a client is crazy mad is expensive. Failing to address the client’s need may delay payments or trigger legal action. Expensive stuff.
- The need to invest to keep pace with the information environment. Most of the mainstream search systems, including Fast Search and other older systems, focused on text. Handling different file types and different content types is an expensive operation for some vendors. The choice is stark: Spend and develop the components in house, spend money for third party solutions and then spend more to integrate those solutions into the core system, buy a company that has the people and the software needed, or ignore the client. There may be other options, but these four have big price tags. The cost of keeping up is brutal because information retrieval does not stand still.
- Figuring out why routine operations are slow or output unexpected results. Most search systems are far trickier to set up than licensees expect. With many knobs to turn, Fast Search could be tweaked so that results could boost certain content or address relevancy under specific circumstances. In a complex system, like Fast and many others, turning one knob and experimenting with threshold values could cause some darned exciting consequences. Rolling back those changes was an exciting operation in itself. When a Fast engineer had to figure out how to get the system back on track, the work was not trivial. What’s it cost to get an expert engineer to figure out what a licensee did? In many instances, a lot.
If you add up the costs of the technical work required for a complex search system, the need for money is significant. Dr. Lervik is not a financial expert; he is an expert in information retrieval. Not even ex-Googlers are adept managers. Witness the AOL goof related to “distressed babies.”
But a senior manager is expected to find solutions to difficult managerial, technical, and financial challenges. If the news story is true, it seems that Dr. Lervik was caught in a situation that set the stage for the unfortunate drama that has been playing out over the last five years.
The big question is:
Will other search and content processing vendors find themselves in a similar situation?
In my opinion, yes.
Warning signs are easy to spot. When search vendors that are seven or 12 years old continue to suck in venture funding, the warning flags are flying in my opinion. Search is essentially a zero license fee utility at this point. Firms that have yet to return a profit or show significant growth may find themselves taking financial short cuts.
The Xenky analyses make clear that financial stress is nothing new to search vendors. Check out the Convera, Delphes, and Fulcrum Technologies profiles. What’s different is that in today’s business environment, the consequences may be increasingly severe. You can find case studies of search vendors at www.xenky.com/vendor-profiles. There is no charge for these reports. Many describe enterprise search solutions that struggled financially and either shut down or sold out.
Enterprise search is a tough business. A sad quack for Dr. Lervik.
Stephen E Arnold, February 10, 2014
Comments
6 Responses to “Fast Search Founder Slowed Down, Then Stopped”
-
- Subscribe to Beyond Search
Feature archive
News archive
-
very interesting news, thanks.
Yes… as you say, “Enterprise search is a tough business.”
but fraud is fraud. and the buck has to stop somewhere.
[…] Fast Search Founder Slowed Down, Then Stopped, February 10, 2014 […]
[…] is, in my estimation, the real reason why some bad-egg companies such as bubble-era MicroStrategy, Fast Search & Transfer, or Autonomy (not yet settled) are tempted to inflate results. I think it’s less about […]
[…] is, in my estimation, the real reason why some bad-egg companies such as bubble-era MicroStrategy, Fast Search & Transfer, or Autonomy (not yet settled) are tempted to inflate results. I think it’s less about […]
[…] John Lervik, who it seems will serve at least a year in prison for financial reporting issues. See here for Arnold’s analysis on why enterprise search is so difficult to manage financially (and […]