Pandia Post’s Fast Search “Snowball”
March 2, 2008
This sunny but chilly Sunday in central Illinois contained a “snowball” of tightly squashed comments and ideas about the Microsoft – Fast Search deal. Pandi Post does a good job of summarizing important news and information about the search engine scene. The story that conked me in the face this morning is “Many Fast Search & Transfer Developers Hate Microsoft.” You can read the full story January 12, 2008, here.
The points that linger with me are:
- A reference to a Fast Search manager and a trade magazine reference to some Fast developers “having a negative religious relationship with Microsoft”
- Microsoft “may” move its search R&D to Norway
- Microsoft may not be absorbed into the body of the Microsoft coding machine but continue as a separate unit
- It’s tough to relocate Norwegian families to California
- Fast will have a research role at Microsoft
- If Fast becomes part of unit, the acquisition could “fail”.
Since this story appeared, the Fast Search conference “FastForward” has come and gone. One of the highlights of this hyper-user group was Microsoft presentation reassuring the attendees that their investments in Fast ESP (enterprise search platform) were wise. You can listen to an interview here.
My thoughts this morning as I watch the harsh winter light on the abandoned farm buildings outside my window are these:
First, I am not certain that Microsoft has figured out what to do with Fast Search technology, people, or customers. The buzz I picked up in Seattle in the last four days indicates considerable fluidity about what to do with Fast ESP. The limited information I have indicates that the Office unit engineered the acquisition and other groups at Microsoft are not “up to speed” or “on the same page” as of February 28, 2008.
Second, as the US economy worsens, the likelihood of mass staff defections from staff makes for interesting rumor mongering. Rumors are not reallity. The “grass is always” greener argument sounds good, but wholesale staff losses are not likely to occur before the deal is concluded and when Microsoft’s intentions whatever those turn out to be are more clear.
Third, the ESP customers have too much skin in the game to jump ship now. Once an organization gets its ESP installed, debugged, tuned, and operational, licensees want to let the system earn its keep. As much as other vendors prod ESP customers to look at an alternative, the ESP customers will drag their heels. Who wants to embark on another rip and replace adventure when budgets are constrained and everyone is burned out?
To sum up, Fast Search’s search system is unlikely to make any “fast” changes of direction in the short- or mid-term. Like other high-profile vendors of behind-the-firewall search, Fast wants to keep customers as happy as possible. The customers want to make their Fast ESP system “work”. And like some other vendors, Fast is working over time to blur the edges of search and retrieval. The company is pushing hard to make search into a call center solution, a sentiment analysis tool, and other broader, higher perceived value functions. The name of the game in search is to move search from a commodity to a solid big ticket enterprise application that warrants a seven figure price tag.
We must not forget that Fast Search & Transfer faces similar challenges that its new owner faces. Fast Search, like Microsoft, has a customer approach that emphasizes software licenses and engineering services for revenue. I may be the only voice on the farm shouting “It’s a business model problem”, but the challenge is not technology. The challenge is the business model for on-premises, complex, fiddly, and huge software systems.
What I find intriguing is that the future of Fast Search and Microsoft may be outside the control of Microsoft and Fast management. On-premises software installations are the traditional approach — I almost typed dinosaur-like — to delivering information systems. The once-unchallenged assertions of better security and control are now being measured against the costs and comomplexity of on-premises installations. What used to work so reliably in years past no longer may be the “right” way for smart organizations to go. Would Toyota use Henry Ford’s approach to inventory control? Oviously not. Why would Toyota want to hobble its operations with an old business process and model?
Beyond Search takes a hard look at the sticky mud of on-premises value-added search systems. I have gathered almost 50 pages of information about what to do when your behind-the-firewall system won’t work the way you want it to. I didn’t plan too write this 250-page study. Customers of traditional behind-the-firewall search systems suggested I gather the information. Why? The traditional approach was fraught with thorny problems, and no one was offering guidance. The vendors are generally circumspect about these matters.
The problems, therefore, are not exclusive to Microsoft and Fast. The business and engineering model for enterprise applications that has been around since the 1960s is the core issue.
I will be watching as this acquisition’s unknowns are revealed in the glare of the marketplace.
Stephen Arnold, March 2, 2008
Comments
-
- Subscribe to Beyond Search
Feature archive
News archive
-
