On the CS side, I would add that if you mash POS data with marketing spend the results are often shocking. Consumer goods companies have terrific raw data but t

On the CS side, I would add that if you mash POS data with marketing spend the results are often shocking. Consumer goods companies have terrific raw data but they ignore a surprising amount of it to their peril. I built a little sales analysis tool for Accenture that they have since sold to several consumer goods clients. It pulls in raw marketing data--pricing, promotion types, locations, revenues--then calculates a few dozen metrics related to ROI, cannibalism etc. Client managers then drill into their marketing data from a couple layers of abstraction and from there they can see fairly clearly how their choices are affecting P&L. We built that because the leading consumer goods companies were essentially guessing at the repercussions a given sales promotion would have on the bottom line. Today the leaders are more sophisticated but the little guys continue to revere intuition and gut instincts over statistics. Our pilot client was Procter & Gamble. I pulled in their POS sales and marketing data e.g. endcap on Tide detergent at [x% discount] in [region y] during [phenomenon z]. I mashed that with private and public data sources, laying out the results for trending and drilldown. P&G bought as did L'Oreal and some others. I'm curious how things are going so if anyone reading has worked with these or similar tools, I would love to know how they've evolved. It's amazing how many expensive marketing decisions are done without regard for scientific analysis. A lot is decided on the basis of how the marketing director/committee feels. In the 90s that was acceptable technology but today it is called Not Learning From Your Mistakes. Compare Google's A/B testing with "throwing ideas at the wall to see what sticks" and you begin to see the problem. Trust me, there are big P&L opportunities in this stuff.