Every one from politicians to the IT industry is trying to raise their profile by establishing their green credentials. This causes lots confusion as they I’m greener than you arguments fly across companies, parties and even countries.  I think a few solid BI principles would be good here:

What you measure is what you get. The UK government’s figures on carbon emissions don’t include air and sea travel and the fact that many of our goods are imported form countries with less clean energy plants. So we can all feel good that we are hitting carbon targets as we all fly and use stuff from other countries. However that’s like trying to diet while ignoring your alcohol intake!

Similarly there’s loads of spin around e.g. “use our stuff and save the planet”, and Microsoft are no exception here. For example Vista power saving is great for the environment BUT not if you have to buy a new laptop to run it as the impact of making that laptop (I have seen estimates 11 barrels of oil per laptop) are significantly greater than the energy savings you would make over the life of  running that laptop. Then there is the disposal cost, and wouldn’t it be fun if there was a cost built into new hardware to reflect that cost like there are for new cars.  So again it’s important to be clear on what you are measuring as what you measure is what you get to quote my friend James.

Impact.  There’s also loads of buzz about reducing server room running costs but this only accounts for 2% of  energy usage, and then something like windows server 2008 will take 15% off of that figure, which isn’t a lot. Virtualisation might cut that in half , but that’s still only 1%.

Of course any saving is good but if you want to buy software to save the planet then that’s not going to be your top choice. My candidates for this would actually be collaborative software like Forefront, SharePoint and Office Communications Server to allow your workforce to work from home or at other sites to reduce travel, and the size of offices you have.

Location Intelligence. Where things happen is important and actually high flying aircraft bizarrely help rebuild the ozone layer because they large quantities of nitrous dioxide they chuck out is part of the ozone formation process. Ozone at low altitudes is very bad for humans and plants as it’s toxic.  Another example is that an acre of rain forest fixes much more carbon than an acre of forest in the UK.  Putting power consumption next to power generation is a good example of doing the right thing such as Microsoft is putting server farms near to hydro electric plants.

History. Business Intelligence data usually has rich data embedded in it, so trends can be analysed.  Historical climate is very open to interpretation, because it has to be derived form circumstantial sources such as tree ring growth, pollen , ice and soil cores and so on.  The variations being analysed are very small and the margin for error is quite variable as techniques get refined.

These issues make analysis of the environment very difficult and my point is that if you don’t address this sort of stuff in your BI project, you’ll end up having the same kinds of debate as there are for the environment :

  • Do we trust the figures?
  • Are we doing as well/badly as we think we are?
  • My figures are different form yours?