IBM thinks they can overhaul the retail analytics industry with largest private team of mathematicians

Today, IBM says the analytic techniques that provide the most value are historic trend analysis and standardized reporting. Tomorrow, IBM believes that organizations want to see insights more clearly. Analytic techniques like data visualization, Watson-like simulations and scenario development (probability predicting), analytics applied within business processes, and regression analysis, discrete choice modeling and mathematical optimization will all provide more value than the traditional latter. They say a new kind of intelligence is needed to create what has become their mantra of the past few Watson-esque years, a Smarter Planet. 

IBM's Stephan Jou, from the office of the CTO of Business Analytics at the Ryerson Ted Rogers School of Management last week said that they plan to change the face of retail marketing research as they have assembled together the largest private team of mathematicians in the world to help counter the fact that 80% of new data growth is unstructured. They say information workers spend 9.6 hours a week on average searching data and 9.5 hours a week analyzing data, accounting for almost half of the work week. In fact, 15 petabytes of new data is generated a day- which is eight times more than all the information in United States libraries. 

IBM wants to increase your "analytics quotient" as the CIO's number one concern has become business analytics at 83%. They say that about 18-24% of client organizations are using analytics currently and the largest obstacle at 38% is a lack of understanding of how to use analytics to improve the business. IBM believes through automation and visualisation it can be done. 

That should come as no surprise considering Steve Levy of Ipsos Reid at the Canadian Marketing Association's Digital Day 2011 on Monday said that they are now measuring 17 facets of digital, something that has increased in every year since the conference began six years ago. 

It was further noted during a panel later that day on metrics that every new platform represents different metrics.

Therefore I don't believe organizations should standardize metrics as the Interactive Advertising Bureau and Mobile Marketing Association are attempting to do in mobile. As digital and mobile continue to change rapidly, so will the types of metrics being used, and so will the types of data organizations want to obtain.

Dave Fleet of Edelman Toronto, one of the world's largest and most respected PR firms says in this blog post last month: "Please – let’s stop creating fake numbers and take a long, hard look instead at how we can tie our activities back to business objectives, and measure against that". 

Too often information officers get burned by measuring the wrong thing- but if the measurement is done right the results can be tremendous. Real-time measurement tied in with location is certainly helping that cause, no doubt. 

Still, perhaps information workers should guard against putting more stock in data than inituition and experience unless you can be highly confident about it. Understanding the entire history of an industry can also help in being innovative once more as you may see new disruptive revenue models.

After all, Duncan Stewart, the Director of Innovation at Deloitte said at the Small Business Summit yesterday when an older gentleman asked him what was going to happen in the next two years that a combination of trends predict the future. Which trends do you measure though? 

Data is certainly complicated, and I wish IBM's brilliant mathematicians a good day, and good luck. I guess if they could create Watson, I wouldn't put it past them to succeed at this as well given their brilliant history, though.