Microsoft put out a fascinating white paper last week. It’s a research report entitled “Impact of Unlicensed Software on Mid-Market Companies.”

It was actually produced by the Harrison Group, but the study was commissioned and paid for by Microsoft. Its stated aim was to find out whether there’s a correlation between “the presence of unlicensed software in mid-market companies and IT related problems those companies might face.”

Obviously Microsoft will have been hoping the report would say there was, and indeed it does. On its very first page it asserts that “companies less committed to using genuine software are much more likely to have system failures, leading to lost critical data and employee downtime.” And there’s more: “businesses using fully licensed software report they’re growing faster, are more profitable, and are more likely to be leaders in their market.”

Jumping to conclusions

Microsoft must be delighted with these validations of its WGA message, but personally I find them a little fishy. The clear implication is that running unlicensed software leads to system failures and poor business performance. “The rest of the company can’t perform at the highest levels,” explains the report, “if underlying IT systems are not operating with those same high standards.”

But “unlicensed software” covers a multitude of sins. If you run your office on dodgy ISOs from the Pirate Bay then yes, you might run into problems. But I don’t believe many companies do that – at least, not in the developed western world. A business is far more likely to be unlicensed because, deliberately or not, it’s simply installed Office or Windows from the original discs on a few too many computers.

In this scenario, “going genuine” merely means wiring a ton of cash to Microsoft and getting a big pack of certificates back. It doesn’t affect a company’s software or working practices one jot. It might mean you receive software updates that would otherwise have been withheld – that’s an admitted benefit. But it’s quite a stretch to imagine that it will actually cause your company to grow more quickly, or to become significantly more profitable.

And indeed, there’s no reason to believe it will. As clearly stated in the report, the actual research only looked for correlations; it didn’t investigate cause and effect at all. So there’s no grounds for suggesting that licence compliance gives companies a competitive edge. It’s just as possible that good licence management is merely a symptom of good overall management. Which, personally, I find rather more plausible.

Madness in the method

I’m also not wholly convinced by the methodology that was chosen for this research. The report’s findings are based on interviews of IT professionals, and we’re told that, in order to achieve “a strong cross representation of large markets”, the data combines interviews from the UK, US, Brazil and China.

But of course, China and Brazil have far higher rates of software piracy than the UK or US. So when the report makes its broad-brush generalisations about “companies that use unlicensed software”, it’s referring to a group that’s predominantly Chinese and Brazilian, with a much smaller proportion of British and American companies thrown in. This puts a rather different complexion onto declarations such as “members of this group are less likely to be market leaders.”

The report does claim that its findings were “consistent” across the different markets, but there’s no explanation of what that means. Nor is there a geographical breakdown of responses. So ultimately this report tells us nothing about which hazards are genuinely connected to software licensing and which are simply prevalent in emerging markets.

It’s not right, but it’s OK

Given the express purpose of the research, that’s a pretty big failing, wouldn’t you say? It basically invalidates the entire study. And yet, conveniently enough, this flawed methodology yields the best possible result for Microsoft: apparent proof that using unlicensed software is inherently dangerous. It’s based on airy presumptions and bad maths, but then maybe that’s just as well.

Because what would a more scientific investigation find? Almost certainly it would confirm that a slapdash approach to licensing is likely to go hand in hand with poor management in other areas. And it would necessarily point out that neglecting your licensing obligations puts you at risk of legal action.

But I’m sure it would also recognise that exceeding one’s licence allocation won’t somehow magically cause a company’s computers to crash – nor its share price. For the average western business, the biggest concerns are likely to be a lack of access to updates and the possibility of software being remotely disabled via WGA.

Because while Microsoft is happily pushing the idea that unlicensed software is inherently dangerous, the truth is that it’s largely dangerous only because Microsoft makes it so.