From Computer Business Review, a sister publication.
Client/server buys prove problematic for US national ice hockey team the Hartford Whalers, the 1995-1996 season was another disappointing one. Its efforts were characterized by sportscasters as uneven, hard to figure, and lacking scoring punch. The team has not had a winning season for five years and ticket sales in its Connecticut home town are so poor that its owner is even threatening to move the team to Nashville. If Peter Karmanos does carry out his threat, he may find it hard to resist drawing uncomfortable parallels between the Whalers’ fortunes and parts of the other organization he heads – $650 million systems software and services firm Compuware. Since early 1993, 23-year old Detroit-based Compuware has been in transition, working to dilute its reliance on a traditional product line of mainframe-based software program testing tools. With a handful of strategic acquisitions, it has tried – and largely failed – to forge a leadership position in either applications development tools or systems management software – although a separate effort to ramp up its professional services business has proved more successful. Overall, the company is making money and growing. Its problem: all the upside is still all coming from the older product lines – Abend-Aid, CICS-AID, Playback and other tools that help mainframe programmers debug their code. The overdependence on aging product lines is not immediately obvious from the company’s financial results. Compuware’s first quarter came in ahead of Wall Street expectations, hinting that the company had bounced back after a weak fiscal 1996. In the quarter to 31 March, revenues grew 24% to $162.3 million, compared with a runrate of 15% for 1996. Operating margins also increased, up from 7.5% to 12.5%. But a break down of the numbers shows that Karmanos will not be walking away with many trophies – at least in distributed computing. They show mainframe licence revenue grew 31% quarter-on-quarter, yet client/server only crept up by a negligible 1%. The problems largely go back to its $280 million purchase of Dutch client/server application development software house Uniface in March 1994. Back then Uniface was one of the largest and most promising development tools companies with revenues of $57.6 million and a growth rate of 47%. While the tools market was then dominated by the database vendors, Uniface’s technology was well-regarded for supporting multiple database products and for its client/server development capabilities. On selling his company to Compuware, Bodo Douque claimed that the merger would allow Uniface’s tools to leverage Compuware’s existing presence in high-end glasshouse data centers, especially in North America. But any level of product and operational synergy has been illusory. Compuware originally intended to operate Uniface at arm’s length. As a wholly-owned subsidiary, it would continue to be headquartered in Amsterdam. But only nine months into the merger, it was folded into the parent operation and the Uniface power base shifted from Holland to Michigan. This coincided with the release of Uniface Six, a major upgrade of the core product. But over the next 18 months the unit seems to have simply stagnated, with Compuware failing to grow the existing business noticeably. Admitting that its strategy was faltering, at the beginning of this year Compuware redrafted all but regional sales staff back to Amsterdam. Compuware does not break out separate figures for Uniface, merely reporting revenues for software licenses, maintenance and professional services. For its fiscal 1996, ending in March, the company ended with revenues up 15% for the year to $614.4 million, making it the tenth largest software company worldwide. That revenue growth was well down on the 30%-40% reported in 1993-1995, evidence that Compuware had been unable to capitalize on its client/server acquisitions. That shows in the fact that overall software licenses for 1996 squeaked up by only 1.4% for the year, to $226.7 million. Net income also has been slipping. In the full year figures, profits slid from $62.1 million to $44.2 million (including restructuring charges of $10.7 million in the fourth quarter).
So, what went wrong?
Compuware corporate officers now blame two factors. We bought a company in the middle of a major product transition, says former president and now senior vice president of marketing Jim Prowse, referring to the move to Uniface Six. The other was a misalignment between the Uniface sales force and the marketplace, he argues. Sales people were concentrating on smaller, tactical deals, selling to technologists in companies, instead of strategic, larger scale projects aimed at CIO level. Now Compuware thinks it is back on track, especially after such a strong first quarter. With its ‘Compuware 2000’ reorganization, all its three core businesses have been geographically centralised – hence the closure of a Uniface outpost in California and t he pull back to the Netherlands. Uniface general manager Frank Slootman says there are now more Uniface developers (195) than there have ever been, and a refocused sales and marketing operation is pinning its revival hopes on release 7 of the produc t in the third quarter. Under the project name Quantum, this will componentise Uniface tools and allow the company to attack the middle ground between low-end desktop development tools like Delphi and PowerBuilder and the high-end now being colonize d by Forte, Dynasty, Nat Systems and Seer.
But how real is the makeover?
Karmanos admits that the overall transition of Compuware from a mainframe software to a mainframe and client/server software company has not yet finished but is now entering its final phase. Uniface has not been Compuware’s only push into client /server. Two acquisitions – EcoSystems in 1993 and Coronet Systems in 1995 – took it into the market for systems and applications management tools. But again, the underlying figures for client/server revenue do not suggest that sales of these Eco products have caught fire. Rather, the real success story at the company has been in services. Compuware has ramped up its headcount in professional services, first through the 1993 purchase of application development services company Techanalysis for $39 million, and last year though the takeover of 470-employee strong Computer People Unlimited for $29 million. In the most recent quarter, professional services revenues grew by 34% to $62 million, providing a third more revenue than software license fees which rose 36% to $50 million. On the product side, Karmanos now needs to start winning some games in the client/server arena. The effort is there, and will continue to be there, but if the result isn’t there, it isn’t enough. Those are the words of the captain of Karmanos’ hockey team; but they could just as easily be his.
Content from our partners
