Borland: into the fast lane.

At first glance, the Borland-Tate merger looks like an exercise in simple addition, a combined enterprise whose 1991 revenues could top $600 million (roughly Wordperfect's size). The deal also puts an aggressive, market-savvy company in charge of the top two PC database languages, dbase and Paradox, and brings one of the industry's productivity laggards--Tate's 1990 sales per employee ratio was $144,086, only 65% of Borland's $218,047--under the sharp knife of a leaner, meaner management.

But numbers like these don't measure what we suspect will be the merger's biggest impact on the database marketplace: a dramatic speedup in the rate of technological change.

The fact is, Ashton-Tate has never shown much interest in being a technology leader. Most of the company's products were created by outside developers; long-term database strategy consisted of little more than a trendy notion of dbase as a front end to various SQL servers (a model that seems to have attracted more press attention than paying customers). Given enough time, Tate's new management might have created a more innovative and capable development organization--but the board of directors instead decided to cash in its chips.

Over the years, Ashton-Tate's lethargy opened the doors to a whole sub-industry of companies that provided dbase customers with products Tate couldn't get out the door, often with slicker features and superior performance. (These aftermarket companies have become so powerful that they even managed to give the dbase language a new name: "X-base.") But technical innovation always remained outside the mainstream; in fact, largely because of Ashton-Tate's conservative approach to product evolution, database technology became a sleepy backwater where innovators rarely achieved much commercial success.

Clearly, Borland--with its "barbarian" culture--doesn't expect to run the database market the same way. Philippe Kahn and Rob Dickerson, who heads Borland's database effort, already have begun to sketch out ambitious plans for next-generation database technologies. The pieces include an object-oriented language standard, a "scalable architecture" that will encompass both single- and multi-processor systems, graphical front ends like ObjectVision for higher-level applications development, a rich networking environment (based on shared technology with Novell), and server-based systems for "complex transaction processing" tasks that demand more power than current PC platforms can deliver.

When Borland was a fringe player, this stuff was technically interesting but not especially urgent. Now, however, Borland has the market clout to impose its development agenda on the rest of the database market. It's too early to predict exactly how fast Kahn and Dickerson will move to implement this agenda. But sooner or later we'll see a far more dynamic marketplace--and a shakeout of companies that can't keep up with Borland's pace.