Imagine for a moment that you are 14th century apple farmer and on your orchard you produce six different varieties of apple. Imagine further that there are three towns within easy travel distance, each of which you visit once per season to sell your crops.
Effective merchants know their customers, and they take steps to uniquely engage those customers, and so continuing in our example; suppose that 2 of your towns are bilingual, each with different languages. Further suppose that in town 1, historically your sales of apple variety ‘A’ are your bestseller and apple variety ‘F’ almost never sell at all. The signs that you hang over your apple cart now need to appear in three or four different languages, and you’re starting to think that perhaps you need a different apple cart to give more space to variety A – at least for visits to town 1. Easy math shows us that depending on which apples are ripe at any given time you might have as many as 72 (6 types of apple x 3 towns x 4 seasons) different sets of apples that you sell.
If we translate this scenario into modern language, you, a 14th century merchant, could potentially have 288 different planograms, that encompass 72 different assortments in 4 different languages, and we haven’t even gotten to margin or shrinkage. Hopefully you sell enough apples to justify purchasing parchment and a lead stylus, because there’s no way you’re going to keep all this in your head.
The point of this example is to illustrate that retail merchandising is the original big-data problem; it predates sentiment and social media, it does not care about channels or shopping habits – just keeping the store running, in a customer centric way, demands reams of data that start with careful planning and end with sales performance.
Retailers need to approach merchandising with a big-data mentality. Start with the assumptions that you can measure reset-lift across thousands of locations, that you can compare historical performance in set completion, that you can plan a budget for in-store activities, and that you can expect to measure a return on those investments. There’s a saying that goes “there are no new stories, there are only new audiences” and using big-data to drive merchandising is an old-story in search of a new audience.
Written by: Nick Downey, CEO MerchLogix