While it’s easy to blame the weather or consumers’ thrifty ways, a recent Retail Systems Research report sponsored by strategy and analytics firm Precima found that only a third of retailers believed their company had a strategy in place to manage prices and promotions effectively across all channels.
“Never before has there been a greater need for analytics to inform a data-driven pricing strategy,” Brian Ross, president of Precima, said in a statement at the time. “Retailers with customer data have a leg-up in the market and can leverage their competitive advantage through identifying key customers and then pinpointing what is important to them and where they care about price.”
Apparel retail may be a beast unto itself, but the grocery business is just as complex. Food is perishable and poor logistics management can lead to a much worse outcome than an order of trendy tops hitting shelves a day or two late. Supermarkets and their suppliers have been slowly deploying analytics and forecasting tools to tweak their stock levels for years (predictive analytics, in some cases, are still a far-off dream), and apparel retailers looking to dip a toe into big data for the first time would be wise to look at how the grocery industry has managed it.
Data alone does not drive insight
“In Europe, point-of-sale data—and by that I mean purely till data, next-day data—is something that has been used tactically for probably 15 years,” said Thomas Beetschen, global information solutions director at confectionary company Mondelez International, speaking last week at an industry event in New York City hosted by customer relationship management (CRM) solutions provider StayinFront.
“And when I say using it, it’s getting the data from every single shop,” he continued. “It’s understanding an uplift, an out-of-stock, a promotion that’s not showing success, and sending the rep the next day to that very shop to solve the problem.”
But while that next-day point-of-sale data is accurate, frequent and granular, it’s not very rich, nor does it replace the need to send a rep to the store to see what’s happening in real time. Not to mention, it needs to be organized as it comes in.
“As opposed to, ‘Now we have this data—what are we going to do with it?’ We really have to come to develop the procedures,” said Ed Lynch, vice president of retail sales at Michael Foods, a food processor and distributor.
Lynch described the trouble with receiving raw, granular data from former customer Target a few years back, noting, “You needed to either work with a third party to get it down to a more useful level but even from there you’d need to get it down into which store—it could be geography, it could be demographic, it could be store size—but you needed to get it down to a manageable level so you could start to analyze what was really going on.”
The reason being, if an item isn’t selling well, a rep needs to visit that store as quickly as possible in order to fix the problem. Having access to the data first could lead to a solution sooner.
“We do have to make sure that we give the reps an opportunity to send information back to us so that we can analyze that data,” Lynch said. “Because if they do find a situation where that Target store has had zero sales for the last seven days but it looks perfectly fine, and you look again seven days later and you’ve sold one unit, maybe it’s not the problem of the rep, maybe it’s not the problem of the store operations, now we have another issue that we need to address.”
Data: the what, the so what and the now what
That’s why data management has to be set up in a logical manner so that as the information is coming in, it’s automatically going into the right buckets.
“Because if you just sit back and watch the data wash over you, you’ll certainly drown pretty quickly,” Lynch said.
Mike Sweeney, senior director of sales at Kellogg Company, furthered that sentiment.
“We have a person on our team that says data is the what, the so what and the now what,” he said. “So if the data answers those three things, we’re able to take action and drive insight.”
But back to Beetschen’s initial point: data alone does not drive insight.
“We are trying to marry that type of information with some rich data, basket data and things like that which we get from the likes of Nielsen, and then try to drive what we would call shopper insight which we would then try to put into an actionable set of capabilities,” he explained. “The challenge for us then is to cut it into little bits and find tools that enable us to put the right selling story in front of the right rep at the right time in the right retail environment.”
To that end, it’s worth remembering that while analytics provided by such companies as Edited or trend forecaster WGSN’s InStock tool can help apparel retailers track their competition and refine their own product planning and pricing, the key to success is organizing that information as it comes in and combining it with real-time customer feedback.