Analytics is the in thing. Retailers who used to call their budgeting department simply finance, are beginning to call them analytics or strategic insights or central insights or by such other names. Research and analysis folks housed under marketing are now the elite in a separate unit for market intelligence.
Marketing and advertising agencies who used to present targeting and advertisement planning in their meetings find the job being also done by specialists, calling themselves marketing analysts.
Product development departments have always dealt with feedback, customer panels, marketing research and statistical validation in test marketing new ideas, etc. But even there, the speed at which social data can be used for early feedback on the product plus the science behind analysis of social data – with its negative bias and millennial segment bias, are all taking the product development area by storm.
But, fortunately, this wave of analytics that we see, is not at all old wine in a new bottle. Humankind has started doing analysis at a scale and complexity not tried earlier. And the retail and consumer goods area is no exception.
Retailers now don’t just use dashboards that, for example, show actuals vs budgets. They now have interactive dashboards on which one can drill down in the areas where numbers seem to be slipping, slice and dice across dates, regions, segments and arrive at a region and period that seem to have slipped the most from budget, if there is such a pattern. Correlating that region and dates with other events, for example store closures, warehouse events, competitive store events and launches, out of stocks, can lead to potential causes being identified.
Imagine a retailer’s Master Data Management problems, due to data coming from multiple sources, vendors and other partners, matching and merging that data has been a problem area for some time now. Now, we have machine learning algorithms that can look through (big) data of how humans did the match and merge, learn from that and replicate that, for MDM match and merge services to be automated. Similarly, image analytics has allowed us to read price labels on shelves and catch an exception where the price in the system is different from the price on the label, leading to better compliance.
Weather data is now routinely used to correlate with demand of key categories. If in the US, rain is a good time to promote ice cream – ostensibly staying at home and watching movies may correlate with ice cream consumption, similarly a Russian book retailer is finding a similar correlation between reading and rain – again one can imagine staying indoors leading to more books being picked up.
In Japan, similarly rain seems to drive more vending machine sales – again limited options for purchase while trapped indoors in offices/ malls, may be the driver. We could’ve hypothesized these relations but with big data and analytics around, we know. That is what data allows us to do, scientifically analyze and be definitive rather than only guess. As Conan Doyle’s most famous fictional detective once said “I never guess.
It is a shocking habit — destructive to the logical faculty.” Big Data and analytics will make retail decision making even more scientific and logical. Those that do not adapt will find competitors smarter.