Intelligent Enterprise

The appetite for Search continues to grow rapidly, and Fast Search and Transfer (FAST) has been one of the most aggressive players in the market. I've covered FAST's move to provide a contextual advertising alternative. Another company initiative boldly claims to revolutionize business intelligence. Yet FAST's recent difficulties, which appear to involve technical missteps and not just operational issues, should make us rethink the limits of search, particularly when it comes to extravagant claims about search-BI.

MarkLogic CEO and industry veteran Dave Kellogg calls the company's 2Q07 financial results a "train wreck." There's no arguing with the numbers: revenues down 31% from 1Q07 and down 11% from 2Q06. FAST characterizes issues as "internal operational and fixable" but is nonetheless "Undertaking strategic review of the business," whether motivated by prudence or because they foresee further trouble on the horizon.

I wonder if this review will put the brakes on FAST's claims, which now seem overstated, to have revolutionized BI by replacing traditional data warehousing with an Adaptive Information Warehouse (AIW). FAST says that AIW "provides ... access to all information through a familiar search and navigation interface regardless of the source, location, and structure of the data." Great stuff, but a search and navigation interface is not BI. Can you really do high-value, integrated BI against structured and "unstructured" data without a data warehouse? That ability would indeed be revolutionary.

I asked the rep at the FAST booth at May's Data Warehousing Institute Conference for an AIW demo. He showed me FAST Radar, a run-of-the-mill dashboard interface that FAST acquired last year, and he showed me a quite nice but standard-issue enterprise-search solution, FAST ESP. I wrote off his inability to show Radar-ESP/AIW integration to product newness and later asked FAST's PR folks to arrange a demo. Yet the "BI Built on Search" briefing I received and a subsequent, July webinar did nothing to show BI value from FAST's integration of "structured, unstructured, rich media content into a virtual intelligence library." I followed up with FAST marketing VP Davor Sutija.

It seems that I had misunderstood that FAST does BI on "unstructured" data, that is, on the volumes of documents that search products can index for retrieval. It does not. FAST Radar must still draw data from tables in a conventional database. Yes, FAST eliminates the data warehouse/data mart design step, but while ESP and AIW may (or may not) integrate structured and "unstructured" data, Radar can't tap the "unstructured" part. Restating: ESP indexes "unstructured" information for retrieval. It also indexes databases to create an adaptive information warehouse whose contents Radar works against. But FAST ESP, AIW, and Radar do NOT allow you to do BI "regardless ... of the structure of the data."

FAST VP Sutija did not respond to two follow-up questions, whether FAST's BI-DW initiatives would be affected by the "strategic review of the business" that FAST is undertaking per the Q2 investor presentation and whether BI is in the category of "replicable, higher margin business" for FAST.

FAST deserves their reputation as an enterprise-search leader. They've undertaken excellent initiatives to extend search technology in new directions, but they have work to do if they (still) hope to put BI solutions "on top of the search platform to integrate and orchestrate all of the information needed to make BI truly effective." FAST AIW and Radar do not make the grade.