My big question is the whole HP/Autonomy spat is quite simple.  Did HP use Autonomy software when it did its due diligence?

It's impossible to say at this point who is right or wrong. When you get into 10 figure write-down amounts, you have to think that the truth or something close to it has to come out at some point.

Conducting due diligence

Back to the original question, though, and it's a serious point about how we conduct due diligence. Autonomy's software is supposed to find "meaning" in unstructured information, and is heavily touted in the e-discovery marketplace.  You would have thought that if HP were serious about getting into that marketplace, they would have smoked the dope before committing $11 billion.  Of course if they did, and the accusations prove to be true, they may want to be adding a warranty claim to the rest of the law suit.

Someone,somewhere is feeling nervous this morning.  Recent events on Twitter have shown that there can genuinely be not a single shred of fire, no matter how much smoke is in your eyes.  But what is for certain is that someone on one side (and possibly both sides) if this dispute is in the wrong. 

When I read the allegations I did scratch my head a little.  SOP 97-2 may not mean a lot to most people, but it struck fear into the hearts of software executives in the 90's, and made the accountants king in every US software company (much to the chagrin of the lawyers and salespeople alike - Not often you get them agreeing).

Standardising accounts

This AICPA Statement of Practice set out very clearly what were the rules for software "Revenue Recognition" (or Rev Rec, to give it its affectionate pet name).  No more channel stuffing, no more dodgy acceptance periods, and no more recognising support revenue up front.  In effect, it put in place a standardised and heavily exampled set of rules about how you could recognise software revenue for accounting purposes. If you worked with or for a US software company (or were listed on a US exchange) then you knew this inside out, and you will be familiar with the same principles today. 

Looking at Autonomy's financial statements pre-acquisition (footnote 2 to the Consolidated Financial Annual Statement of the 2010 accounts if you are that interested), those are the rules Autonomy followed.

Doing the legwork

If you haven't bumped up against "Rev Rec" before, trust me when I say there is a whole industry in the US and Europe that is devoted to this one subject.

Lesson One of a US company acquiring any software company, especially if it is non-US, is to check its true revenue position.  I've been involved in the acquisition of dozens of European tech companies by US firms, and that is the first thing you do. And it's not that difficult to be honest.

You pick a reasonable selection of deals, of different types, focusing particularly on channel-type deals.  You check the basics.  Has the software been delivered?  Were the payment terms usual (and were they met)?  Is the software part of a larger deal, and was a fair value allocated to the software element? Has the software maintenance revenue been taken over the life of the contract as it should be, or capitalised?  And if you are using resellers, you ask them for evidence of the actual end user sale.

It seems quite astonishing that with the use of reasonable forensic software and a rigorous approach to financial due diligence that a massive mis-statement would not have been picked up.  Undoubtedly there would be a few deals where both sides disagreed, but that is why you have de minimis clauses in the sale and purchase warranties, to allow a bit of give and take.

I'm still really interested if HP used Autonomy's software when they did their due diligence.  These days, I sell software that has found the "smoking gun" in unstructured data, so I'm probably only asking because HP didn't get me to look at Autonomy's books.  Or if Mike Lynch needs a hand in the upcoming litigation, my e-mail address is at the top.