The acquisition of software management vendor Mercury Interactive by technology giant Hewlett-Packard is something of a curious one.

The bare bones of the deal are straightforward: HP will pay $4.5 billion in cash for Mercury. Ostensibly, this will help HP bolster its management software portfolio, helping “CIOs reduce IT costs while increasing the effectiveness of IT spend,” says CEO Mark Hurd.

In acquiring Mercury, HP will, at a stroke, double the size of its software business to $2 billion – and broaden its focus beyond server management software. This is a significant change in strategy for a company whose software division has been something of the poor relative ever since it mothballed its middleware operations in 2002. In spite of calls by both financial and industry analysts for HP to strengthen and broaden its offerings in this key strategic area, OpenView, its longstanding and highly regarded network and systems management suite has been its only significant software offering since then.

Yet the deal has still attracted some criticism and raised some concerns among HP investors. Hurd was brought in to stabilise the embattled tech giant, which under the previous CEO Carly Fiorina struggled for nearly five years to deal with the impact of its mammoth Compaq acquisition. And while Hurd was adamant that his arrival did not spell the end of buying companies, he had promised that in future any deal would have to be both “manageable and digestible”. And there’s the rub: this deal has the potential to be neither.

One issue is the price. At $52 per share, HP is paying a 33% premium on Mercury’s pre-deal stock price. That is a very healthy mark-up for a company undergoing the trauma of a US Security and Exchanges Commission probe into its finances. HP may point out, though, that Mercury has been growing steadily in recent years and its profits are rising.

Then there is also the issue of the technology. Through its acquisition of Mercury, HP will gain a comprehensive software testing platform, but as Paul Herzlich, principal analyst at market watchers Ovum points out: “HP is nothing in the testing world, and it is not credible to believe that it has suddenly got the religion.”

So what else has Hurd seen in Mercury? Some believe that Hurd is most interested in the service-oriented architecture (SOA) governance and directory tools from Systinet, which Mercury bought in January 2006.

Systinet is the “gem of the deal”, says Jason Bloomberg, a senior analyst with SOA advisory group ZapThink. If HP can successfully integrate its acquired technology with its OpenView platform, Hurd could have found a way to resurrect its software division. The problem with this analysis is that Mercury paid just $105 million for Systinet in January – a fraction of the $4.5 billion HP is now paying.

Hurd’s rationale is that the combination of OpenView’s strength in systems network and IT service management with Mercury’s application management, application delivery and IT governance capabilities will allow HP to become an “end-to-end leader in enterprise IT management for the entire IT lifecycle.”

And Hurd baulks at suggestions that he has paid too much: “Let me make this clear: the absolute, riveted focus we have on costs in this company will not stop. We had a strategic opportunity that makes tremendous sense for us.”

“It’s now a battle of the Titans, HP versus CA and IBM Tivoli, to see who can put together the most complete, service-oriented enterprise IT management (EITM) story for the enterprise,” says ZapThink’s Bloomberg.

But for HP to compete with IBM “on a more level footing” would take further acquisitions, says Dennis Gaughan, of AMR Research. BEA Systems, Sybase or Tibco are all possibilities. But it remains to be seen whether Hurd has the appetite for a deal of that magnitude.