Many column inches have been written in the last week picking over HP’s writedown of its Autonomy investment, and its allegations of accounting irregularities by Autonomy prior to the acquisition.

But it seems to us that something that is rather overlooked in all this is HP’s overall software strategy. Although much opprobrium has been heaped on Léo Apotheker for his tenure at HP, in one respect at least we think he was right: HP needed a much better software strategy. And, given Apotheker’s background at SAP, we guess the HP Board thought so too, when hiring him. Other than through acquisition, and despite being one of the most important players in the B2B software market, HP Software had shown little or no growth for some years.

And even though the company is (very) good in some software areas, it is still widely perceived as a computer systems vendor. HP’s DNA consists of hardware for the most part. This is why Apotheker was prepared to spend so much to try to change things.

Unfortunately, we don’t think – we never thought – Autonomy is the answer to that problem.

Pre-Autonomy, HP had essentially three loosely connected software product lines – network and systems management, software testing products, and document and information management products. Products in the latter category had been bought rather than built, principally it appeared to us because they’d drive sales of more HP kit. Buying Autonomy bolstered this area of HP’s business but it was hardly going to be transformational. It didn’t solve the problem of the portfolio silos, nor was it big enough (at under $1bn in revenues) to change the balance of the software revenues or customer base.

What HP really needs is something different – enterprise apps or mobile development, say. Or a web commerce platform. It needs to be able to offer something big and important to many companies. Such an acquisition should be fast-growing in a hot and sexy market area, or really big, or both.

Given HP’s background, something in infrastructure or information management is easier, but business applications might work too – eg in contact center/service desk management. If it is going to buy big, then it really needs something that buys it market-gorilla status. Ironically, the best bet might have been SAP, but unfortunately SAP’s just too expensive now – if HP could have taken the whole company, not just its CEO, things would have looked very different today.

So in the meantime, while the lawyers sharpen their pens, HP Software languishes. What we’d encourage HP to do is to get over the Autonomy saga as quickly as possible and come up with a new, and more plausible route forward for HP Software.

As cloud computing changes the software landscape, blurring the boundaries between software and services (and reducing companies’ reliance on buying their own hardware) then surely now is the time to strike.

New classes of computing are emerging and taking over. HP made a potentially smart move with its Vertica Big Data acquisition (for rather less a tenth the price of Autonomy, according to press reports), and has backed that up with its Vertica SaaS offering in the HP public cloud. We’d like to see more that kind of thinking in the future and suggest HP builds entire solutions, eg by offering Twitter and Facebook data access and analysis as part of the offering.