Elephant graveyard's new management

On the upside, IBM's acquisition of Informix Software promises to bring approximately 120,000 new customers to the Data Management Group, the home of the rising star DB2, as well as a few thousand talented database professionals and a number of useful pieces of technology.

On the downside, however, the acquisition also brings a plague of problems such as a history of business blunders by Informix management, thousands of ticked-off Informix investors, and--worst of all--several legacy database management systems (DBMSes) that entered Informix Software through its acquisition of Ardent Software.

Lest we forget, Ardent had evolved into a kind of elephant graveyard, where ancient and hoary IT beasts came to die--after disgorging considerable amounts of maintenance revenue, of course. As the final resting place of woolly mastodons like VMARK, O2 Technology and UniData, Ardent had enough maintenance revenue from legacy customer bases to fund a new data integration product called DataStage, as well as to fund acquisitions of Prism Solutions and Dovetail, which are now folded into DataStage.

Informix, too, had gone through its share of acquisitions, but it collected newer and hipper database vendors like Red Brick and Cloudscape. By acquiring Ardent, Informix entered the bone yard business. Informix executives felt they needed the data integration technology of DataStage to enable its burgeoning data warehouse business and its fledgling corporate portal effort.

To get this princely product, Informix was willing to accept the warted toads in Ardent's motley collection of legacy DBMSes. Yet, Informix soon found that no amount of kissing would turn these frogs into princes. So it embarked on an aggressive upgrade and migration campaign called Arrowhead.

But Ardent's crony customer base, steeped in 3-decade-old technologies like hierarchical databases and Pick systems, are, for the most part, either incapable of fathoming the cutting edge of object-relational databases or--more likely--do not see a return on investments from porting to a more modern platform. Hence, Informix Software found itself supporting seven code streams for DBMSes. Some insiders claim there are 11 code streams, with little or no hope of discontinuing any of them anytime soon.

That's a very heavy research and development millstone to wear, not to mention the confusion created by marketing--and trying to differentiate--seven DBMS products.

In an attempt to bring order to the bone yard, Informix restructured in late 2000 to create a holding company called Informix Corporation. It contains two subsidiaries: Informix Software (with a host of DBMSes) and Ascential Software (with Ardent's DataStage and Informix's Media360, iDecide, iSell, and whatever they call their portal today).

But this created its own points of confusion, because the Red Brick Server (for data marts) and XPS (for high-end data warehouses) ended up in Informix Software, while the complementary product DataStage--a data integration technology for warehousing--went with Ascential. In the end, the restructuring of Informix, which simply sorted elephant bones by type, burned up a lot of resources without any negligible benefit.

Big Blue's big gambit
All this cannot have gone unnoticed by IBM executives. In the course of due diligence, they must have opened a lot of closets, some bursting with pachyderm skeletons. It leaves me wondering: What's IBM's real goal, and is it worth assuming the management of an elephant graveyard? Before answering, let's look at the effect of the acquisition on competitor Oracle.

On the one hand, Oracle's overtly competitive corporate culture had most employees making nasty cracks immediately about the undignified death of Informix. But the passing of this once-great vendor doesn't mean it lost a meaningful competitor. After all, Oracle trounced Informix long ago.

On the other hand, Oracle's savvy executives most likely had a "Maalox moment" upon hearing of the acquisition. They're reaching for antacid because their most formidable foe--the data management division of Big Blue itself--is suddenly much larger and stronger.

When that great knife known as market share re-carves the DBMS pie chart, IBM's slice may soon exceed that of Oracle. And--in case you missed it--that's IBM's real goal with this acquisition. If it's a greater share of the DBMS market that IBM wants, then acquiring Informix Software is a fast and relatively inexpensive way to get it.

But let's hope that managing Informix's elephant graveyard doesn't prevent IBM from trouncing Oracle.