Iíve had quite a few comments back from my BPI missing millions post so decided to add some further thoughts.

Firstly, just to emphasize a couple of previous points: the impact of file sharing on future music buying patterns cannot be underestimated. At Jupiter we have written a number of reports indicating this and regular readers of this blog will probably have heard my line on this too many times! But, the current impact is overplayed: five percent of the total population can only have so much impact, particularly when you consider that 40% of them are under 25, half are not working and a quarter classify themselves as dependents.

Secondly, for argumentís sake letís take the BPI numbers at face value and assume that file sharers did indeed spend £414m less in 2005. That would leave UK revenues at £1.5 billion in 2005, which is £334 below the actual 2005 market value of £1.8 billion. So to make up the difference, some other consumers would have to be spending £334 million more. Letís assume it is a similar size number of people that spent more on music as currently file share and (for the purposes of this hypothetical argument) spent £415 million less. If so that would mean an average increased spend of £101 per buyer for this new group of buyers, which in anyoneís books has got to be quite a result, but unfortunately isnít particularly probable either.

The problem is that the extrapolation from the survey panel simply doesnít tally up with the actual market numbers unless you have another segment of the population buying that much more to compensate. At Jupiter we are always very cautious about using consumer survey data to inform spend levels at a micro level, let alone a macro level. We always use a combination of supply and demand side metrics to balance the equation. When you ask a consumer about how much they have spent it will be a directional input at best. If I asked you how much money you spent over the last 12 months are you really going to be able to give me an accurate number? The BPI data pulls upon a spending measurement panel that uses note books, but that methdology is far from watertight, as illustrated by the concerns over Rajar’s accuracy.

I fully understand why the BPI is trying to hype file sharing so much. It needs to protect itís future and it knows full well that it is going to be hard pushed to do anything meaningful against individual file sharers with the next generation of anonymous networks. So it needs to get its punches in now. As a music fan my heart is with them, but as a forecasting analyst my brain is uncomfortable with the use of a single sourced methodology in this context.