Information Age: News, analysis & insight for IT & business leaders

If the executives of Tibco, the integration software company, seemed to have a bit of swagger about them as they outlined their plans for their recently completed acquisition - the UK software company Staffware - then it is understandable. Things have not been at all easy of late.

Back in mid-1997, the CEO, Vivek Ranadivé, wrote a book called The Power of Now, in which he extolled the virtues of integrated, real-time computing. It was aptly titled: Ranadivé was one of the men of the moment, celebrated as a true visionary in a valley full of seers. He presided over a hot software company that had managed to win big orders on Wall Street as it outstripped its rivals in the emerging sector of integration software. By 2000, the company was valued at more than $30 billion, and a year later sales peaked at $322 million.

But by 2002/3, Ranadivé was definitely looking like yesterday's man. Many analysts began to write off Tibco as a fading star as sales slumped more than 20% and its market valuation crashed to little more than $1 billion. With five year losses adding up to $140 million, Tibco was in danger of becoming a penny stock.

Executives attributed Tibco's woes to the dire economy, and the fact that the company was still building business in a long-term strategic market. But critics could see all kinds of other issues.

First, the new technology of web services, many observers said, would wipe out the need for complicated integration hubs like Tibco's; second, they noted that young companies such as Sonic Software, touting disruptive, low-cost standards-based Enterprise Service Bus (ESB) alternatives, were threatening to undercut the company at every turn; and third, they could all see IBM, Microsoft and SAP all lining up to offer serious competition.

To make matters worse, Tibco was locked into a distribution agreement with its 49% shareholder, Reuters, that meant it had little control over sales into its most important market, financial services.

But what a difference eighteen months makes. On almost every front, Tibco's outlook has brightened. Restructuring, coupled with resurgent sales, means the company is now both profitable and growing - with sales forecast to reach a record $370 million for the year.

Web services, it turns out, will probably benefit rather than harm Tibco, helping to open up the integration market.

Even the low-cost ESB alternatives to Tibco's ActiveEnterprise are doing no harm at all, claims Ram Menon, the company's chief marketing officer and strategist. "In every case where we've had people say they want to buy an ESB, they've taken a whole Tibco product instead."

Meanwhile, Tibco has extricated itself from the Reuters resales agreement and is now attacking the financial services with vigour. Menon says that Tibco has "performed beyond our wildest expectations" in financial services.

Such is Tibco's progress, that Massimo Pezzini, in of Gartner's leading middleware analysts, positions the company as the most likely consolidator among the 'pure plays' in the overcrowded integration and web services market.

Menon, however, insists that this is not Tibco's strategy. "Consolidation is a strategy that focuses on other companies that have failed. From the technology and customer perspective, it is not something Tibco is interested in".

Certainly, the profitable UK business process management (BPM) company Staffware, acquired for $123 million, could not be classed as a failed company, even if the management were known to be concerned about the company's relatively small size in a market that is increasingly attracting heavyweight players.

But the acquisition has created some strategic dilemmas for Tibco. The question: How much should the integrator integrate? In the UK, Staffware has strong customer relationships, especially with business rather than IT managers, that Tibco is wary of disrupting.

There are also architectural considerations. BPM is an integrating technology, but it relies on creating step-by-step processes using a workflow approach. Some customers want to keep it at that level: "We are very conscious of keeping the process layer independent. That is the way a lot of companies think about BPM," says Menon.

But that is not the whole story. For two years, Tibco has had a strategy of building process management into its integration suite, and more integration makes sense.

Tibco's solution has been to do the fashionable thing - loosely-couple Staffware. The company, now operating as the Tibco BPM Group, will continue to sell its standalone Staffware Process Suite, focusing on human-centric processes. But at the same time, Tibco is also building new, standard-based links between the products - while insisting that both products can work well separately. These interlinked products are expected to be available in late 2004.

Further and deeper integration work is likely. As the service-oriented architecture takes hold, modern integration tools will require more integrated BPM functions - and vice versa. Furthermore, BPM technology could play a key role in the next generation of software models, the 'event-driven architecture', an approach Ranadivé outlined even before it became fashionable in yet another book, David Luckham's The Power of Events.