My takeaway from the current Brightpoint/Emptoris lawsuit is the same that I had years ago when I was first involved in the spend analysis market. If a client has inaccurate expectations with respect to the commodity mapping process -- for example, if the client believes that there is a magical error-free automated or other mapping process -- then this will lead to unhappiness. This is because any commodity mapping process requires extensive manual review and correction, and the result is inevitably quite far from perfection. Contrariwise, if the client is educated regarding the commodity mapping process, and furthermore is provided with the ability to easily correct mapping errors themselves, then everyone has their eyes open from the start of the process, there is no unhappiness, and there are no unreasonable expectations.
I remember a client meeting with a user of spend analysis years ago. A gentleman at the back of the room kept drilling into the weeds and finding mapping errors (this is *always* possible, and it's *easy* to do). At that time, nearly all the mapping was done offline in most solutions, and our solution was no exception. So all the presentation team could do was to grit their teeth and move on, with this guy at the back (who was opposed to the whole initiative) sniping away, along the lines of, "if you can't map THIS, why should we believe ANYTHING you've presented?"
I remember the presentation team coming back to the mapping team and ranting and yelling about how "embarrassed" they were, and "what about all your claims about how good your mapping is," and "how can we sell this if your process allows such stupid errors," and so on.
All of which is just pointless. Our mistake was that somebody on our team allowed the client to believe that our mapping was more accurate than it could possibly be. In fact, our mapping WAS extremely accurate, for the top N vendors and top N GL codes, which is all that's typically necessary in most cases. The rest of the spend was mapped heuristically. And, of course, vendor mapping alone is inadequate (the Polaris example Jason cites in his article from earlier today on the suit is an illustrative example). Even supposing that vendor mapping is sufficient, which it is not, industry-leading vendor databases have a best-case coverage rate very far from 100%. And that's assuming that it is possible to match vendor names accurately, which in many cases it is not. The best possible grade is therefore very far from perfection.
So I can understand a client being upset, especially if they started out believing (for whatever reason) that the process was perfect, or that it was much better than it actually is. The downside of a lawsuit on the issue, though, is that there are a host of companies who desperately need to perform a spend analysis. Now they may be disincented to do so, because of the publicity surrounding this lawsuit. The sad part is that even the simplest spend analysis -- whether it is done by hand, with in-house tools, or with a third-party commercial tool -- will provide significant value if it hasn't been done before.
Spend Matters would like to thank Anonymous for sharing his thoughts on the topic.