Microsoft Buys FAST; Last Year It Was BI, This Year It's Search

Microsoft announced today that it was buying Fast Search & Transfer, the Norwegian enterprise search firm, for approximately $1.2 billion. It looks like pretty much a done deal, in that FAST's Board of Directors is recommending the acquisition and the two largest shareholders are on board (per a ZDNet blog post).

FAST went into an operational meltdown last year (see Forbes article), with writedowns, layoffs, and the exiting of many longtime U.S.-based employees. This probably helped decrease the purchase price, and Microsoft seized the moment. While the operational wheels fell off, the FAST technology is strong at its core.

That FAST would be acquired is not surprising; Bjorn Olstad, the CTO, commented in a meeting I was at last year that the infrastructure players (IBM, Microsoft, Oracle) would increasingly encroach on FAST's space. In short, FAST was well aware of the challenges ahead, and sounded like it was amenable to being acquired. What surprised me was that Microsoft bought FAST; I always thought it would be Oracle, for a variety of reasons.

Last year, we saw the infrastructure players absorb business intelligence (Oracle bought Hyperion, SAP bought Business Objects, IBM bought Cognos); this year it will be search. At this point, Autonomy is the only large best-of-breed player left standing, and it will have a hard time going it alone.

This is a huge coup for Microsoft in the enterprise search space. After futzing around for years, Microsoft finally started to get serious with search in SharePoint 2007. It's not perfect--clients have started to tell me the boundary conditions they're running into--but it's a lot better than search was in SharePoint 2003. If you split the search market into three sectors: (1) cheap and OK, (2) relatively inexpensive and an 80% solution, and (3) expensive and sophisticated, Microsoft is targeting tier two with SharePoint Search. Microsoft Search Server 2008 Express is its answer to tier one (see my previous blog post here), and the FAST acquisition is its answer to tier three.

Strategically, Microsoft now has all the bases covered (and, as a nice side benefit, prevented IBM and Oracle from adding FAST to their arsenal). Now, of course, it has to execute, which is always easier said than done.

If you go down the list of competitors, they're now coming up short:

• Autonomy: Starting to look a bit stranded as the remaining, large, best-of-breed vendor, and strong only in tier three. Great for publishers and specialized niche search, but too expensive for general deployment.
• Oracle: Has a solid tier two offering with Secure Enterprise Search, but nothing for tiers one and three.
• IBM: A lot of strong technology, but oriented more for integrators (think IBM Global Services) than for an off-the-shelf purchase. It doesn't help that there are internal organizational walls to overcome. Search falls within the Information Management division at IBM, while collaboration is controlled by the IBM Lotus division.
• Google: A strong offering for tier one (Google Search Appliance), but nothing for tiers two and three.

To sum up, Microsoft just became a one-stop shop vendor for enterprise search.

Comments

[Tom Culler]

What about Endeca in the faceted navigation/BI search space?

[Guy Creese]

Yes, Endeca is certainly a player, but I would view them as a specialty player, not as a broad horizontal solution. Endeca is very good when you have a high metadata/content ratio, because that's how you can can filter results (by cost, color, weight, mutual fund type--it depends on the application). However, Endeca becomes less successful when the metadata/content ratio is low. While I wouldn't be surprised if Microsoft looked at Endeca, I think they made the right choice in that FAST is applicable to many more search scenarios.

[Peter]

You don't mention Vivisimo in this list either -- any thoughts on how they fit into this?

[johnny99]

Your comment about the third tier is right on track. FAST has superb core technology that can be manipulated and tune to a broad variety of needs. Microsoft is a mature company and it should be relatively easy for it to polish out the product and integrate it into already existing offerings.

I would disagree with you (and Forbes) on the meltdown. If you track the company's growth over the last several years, you will see that it was time to trim some fat.

The real question is: What will Microsoft do with the product. Will they can Linux support? What about SDAs written in Java? And what about the existing accounts that rely on the technologies that may become extinct?