Oracle’s acquisition of Thinking Machines’ (TM) assets provides further evidence that data mining tools vendors are a dying breed, soon to face extinction. Aside from a handful of hopefuls such as DataStage, NeoVista, NCR and DataMind, that are still managing to eke out a business in this market, there has been a general trend towards consolidation. Data mining tools are being subsumed into broader analytical applications suites – SPSS acquired Integral Solutions for its Clementine data mining products and Business Objects bought Business Miner from ISoft. The Thinking Machines acquisition by Oracle is the latest victim to fall prey to a contraction in the market place.
As an isolated market opportunity, data mining no longer has the lure it was once believed to have possessed. In 1998 the data mining market was only worth $40m, according to Herb Edleston at data mining analyst company, Two Crows. He predicts it will reach $70m at the very most in 1999. It is therefore hardly surprising that data mining companies are either shaping themselves up as acquisition targets or repositioning themselves by constructing suites such as customer relationship management applications that leverage mining tools a la DataMind.
Mining vendors are increasing embedding tools into specific applications. This won’t eliminate the need for broad-based tools but it will open up a new market opportunity, says Edleston. It is this new market opportunity that Oracle is going after with the acquisition of Darwin, TM’s massively parallel-based data mining tool.
The enterprise software giant plans to embed Darwin into its customer relationship management suite, in addition to business intelligence tools and enterprise resource planning software. The idea is to offer a deeper level of data analysis than is currently possible with its reporting and ad hoc analysis tool, Discover and its advanced calculations and forecasting product, Oracle Express.
Although Oracle has yet to outline a product roadmap, analysts are speculatively devising a blue print. Integration with the Oracle CRM applications will be the first priority, says Lynne Harvey at Patricia Seybold. There is a requirement in the market for tools to go beyond click stream data and market segmentation analysis provided by CRM applications to provide a cohesive and interactive backbone from which to analyze data from a number of sources. Darwin is strong in this area, she says.
However, Oracle faces a fairly significant integration challenge in enabling Darwin to interoperate with Oracle Express, the Oracle 8i database and the rest of its analyst tools, say analysts. Are these tools interoperable? No probably not. They are probably not even at the batch import/export level yet, says Lou Agosto at Giga Information Group.
That said, TM’s new parent still plans to develop Darwin as a standalone tool for applications such as database marketing in the financial services and telecommunications industries – a market in which it has an estimated 60 customers in either beta, implementation or roll-out phases. Aside from the Solaris and HP- UX based variants currently on the market, Oracle also plans to develop Darwin on NT later in the year.
Oracle already resells Darwin as part of its Warehouse Initiative scheme. The data mining application is based on parallel technology that enables users to extract information from very large databases. It also comes with a software developers’ kit, which includes wizards, to enable users to generate models so they can set the parameters for data they wish to search.
Once a high flyer in supercomputing hardware, TM came crashing to earth when the market for massively parallel machines failed to materialize into the lucrative revenue pool many expected. Four years after its brush with Chapter 11 proceedings, the company was still struggling to generate revenues from its reinvention as a data mining software concern. TM had invested heavily in the technology at the expense of marketing – a difficult business model for growth but nonetheless one that made it an attractive acquisition target, says Edleston.
Thinking Machines’ flawed sales and marketing strategy is something the company is quite open about. Oracle had been a business partner of ours for over a year and it came to us and said it would prefer to own Darwin rather than partner with us. It has 9,000 sales and field support staff [TM had three or four sales people] and so we took the opportunity to create new opportunities for the product, says Bob Deretti, president at TM.
That said, there is a suggestion that the majority shareholder in TM, New Valley Corp, had been exerting pressure on the management team to sell out. New Valley had been pushing for a merger between Red Brick and Thinking Machines and when Red Brick was bought by Informix, New Valley saw the opportunities to gain a return on its investment fading fast and therefore pressed to find a new buyer. Although Oracle is unwilling to disclose just how much it paid for privately-held Thinking Machines’ assets, the deal is widely believed to have been valued at $4.7m. In return it became the owner of the Darwin software, the company’s 33 staff mainly in engineering, its brand name, intellectual property and think.com URL in a deal that gained nods of approval from many investment analysts. We view this acquisition as a modestly positive for Oracle – more technology at a reasonable price, says Wendell Laidley at Credit Suisse First Boston.
