Scene 1 :

Presentation to a prospective client.

Pardon me for the cliche - 'You don't sell products to a customer. You sell dreams'. To project that dream you need to build up a story around your product/service. You need to have a vision and importantly project it extremely well to the person controlling the cash.

Market Research helps you put fillers/justifiers or meat (if you will) to that story. But does it always play that role? Visualize that you wish to show a constant reduction in OPEX due to the usage of your product/service. A comprehensive market study is carried out to establish the need of the product/service for your client's market and elaborate projections are worked out. 

Data, after all is just data. Unless you actually put a trend around it and analyze the trend, data is not of much use. Suppose the trend analysis of your data yields a result that is contradictory to your story. 

Your gut/experience says that your client needs this service - Justification is becoming an issue as the mined data tells a different story. Which line would you tow? and Why?

Scene 2 :

A journalist's inferences questioned. 

"The dollar is weakening and hence the Oil prices are bound to shoot up. This is creating a second virtual bubble and may lead to a double dip recession." A good motherhood blanket statement. The gist of the rest of the article went on to prove that Dollar Value and Crude oil prices worked on an inverse relationship.

Well at a broad level i would agree. The purchasing power parity economic theory comes into play here. Consider this. 

Suppose the U.S dollar is trading for 4 Dirams in the Exchange rate market. Crude oil prices are at 80$ a barrel=> effectively 320 Dirams. Now dollar price weakens due the state of the U.S economy, reduction in spending due to stoppage of fiscal stimuli, dwindling exports - multifarious reasons. Essentially now the dollar is weaker and maybe worth less than 320 Dirams. 

Now, from a oil retailer's perspective - If he sells crude oil to his local market as well, the local customers would obviously buy the crude oil in Dirams at a fixed rate (This rate is effectively independent of the dollar value). So, for a customer in the local market, he would rather convert his Dirams into U.S.D and buy more quantity of oil. To make the situation equitable, the oil reseller/retailer jacks the price of crude oil. He jacks up the value of crude oil in the local market (i.e. even if you exchange your currency for the U.S.D , you still would get the same amount of oil) and in an obvious after-effect, the price increase spills over to the market outside the local market. 

Pardon me for the oversimplification but I have tried to touch upon a justification for that motherhood statement we discussed a while ago. 

Cut to August 2008 when the Oil prices were spiralling even when the Dollar was strengthening after the fiscal stimuli. A similar relation with respect to Gold prices and the BSE are contradicted when you actually analyze the data in the background.

The explanation given by the PPP theory is pretty intuitive and logical. My gut feel tells me that is by far the most probable reason. But once i evaluate the data/figures and see anomalies, I am in a quandary as to whether to go ahead and put forth my inference as a justification.

Scene 3

My question to you

This brings me to the purpose of this writeup. In such a case where the data (supposed to feed into your story) contradicts the story itself, what does one do? Does one stick to the data and kill one's story? or does one stick to the story/gut feel and completely/partially ignore the aberrations in data?

Mckinsey says - Data/information should be respected. They say , a recommendation is not as important as the data that supports that recommendation. "Respect and Accept data with all its anomalies" : Well I would but then what about my pretty experienced gut?