Recruiting and retaining customers is a priority for any retailer – but not all of those customers will add to the profits and some may prove loss-making. Penelope Ody investigates why understanding customer profitability is a major challenge in today’s omnichannel world.
SOME WILL only buy when prices are heavily discounted, others are serial returners, while a few love complaining loud and long on twitter: anyone who has ever worked in a shop is only too well aware that some customers really can be more trouble than they are worth.
Long ago when it was only “the shop” experienced staff quickly identified both time-wasters and top spenders: the “good” who buy regularly with no complaints and the “bad” who find fault, demand discounts and inevitably bring things back. In an omnichannel world it is rather different: “free delivery and return” very quickly wipes out any profit for those customers who order three dresses in different sizes and inevitably return two. Then there are the ones who constantly opt for click-and-collect and then decide not to buy, or for whom those store-to-store transfers have wiped out any profit on low margin lines. As profit margins continue to erode what can retailers do to limit the loss-making shoppers?
Calculating the profit on every product including its cost to serve and then integrating that with an individual customer’s purchasing pattern and history is theoretically possible; applying that information in real time across all channels and millions of shoppers to accept or deny an order or activate differential pricing is another matter.
“It is something organisations may want to do but most don’t even have the profit per product data,” says Craig Sears-Black, UK MD at Manhattan Associates. “Getting down to individual customers would be challenging, but you can do profitability by characteristics looking at customer type.”
As Sears-Black explains, with a rules-based order management system it is quite easy to set up the website so that when that customer orders three different sizes of the same style, she is not offered free delivery or return. It may mean that such offers cannot be promoted on the home page, but it allows the system to automatically put in an appropriate charge at the checkout based on the likelihood that two items will be returned. Not displaying delivery charges up front is a common customer bugbear, but only displaying them after an individual shopper has logged in, been identified and personal preferences, behaviour and recommendations noted, may make the difference between profit or loss so could be worth any flak about transparency.
While personalised offers, made both in-store and online, tend to be the preserve of luxury retailers where clienteling is the norm, today’s millennials – familiar as they are with mobile apps, targeted geolocational promotions, and Amazon’s stream of recommendations based on previous purchases – expect it every time. It’s an area IBM is addressing with its “cognitive commerce” development. “Personalisation is a common theme,” says James Lovell, Retail Commerce Solutions Executive, IBM. “The step change is the ability to hyper-personalise the customer experience and actually learn from individual interactions. Cognitive systems learn and develop expertise as they consume data.”
Dan Murphy, Partner with consultants Kurt Salmon, is more cautious: “One-to-one relationships and marketing will certainly be the future at some point,” he says, “but a major problem is that retailers are currently trying to develop strategies for the 21st century using 20th century IT platforms.”
As Murphy points out, while every retail CEO would claim that “customers are at the heart of our business” only a handful at the luxury end, along with some of the pure plays, would be able to name their top 10 profitable customers. “Systems and reports are built around products, sales figures and SKUs,” he says, “and no-one wants to reconfigure all of that for the unknown ROI of customer centricity. Some are starting to look at the cost to serve – but even when systems are in place to identify loss-making activities they can still fall foul of human error on the shop floor.”
Murphy cites top-end consumer electronics where the profit margin is minimal and has to be helped by add-on accessories – such as sound bars, TV stands, warranties or installation services. “You could have a system which identifies items that would become loss-making with a store-to-store transfer for click-and-collect,” he says, “but in-store will that inexperienced staff member faced with a customer asking for such items remember to check a restricted transfer list or recognise that this particular customer is a top-spending regular deserving additional service?”
Online it could be more straightforward with rules-based tactics to limit the availability of low stock items to certain geographies to avoid those expensive inter-branch transfers, restrict click-and-collect options to store-based fulfilment capabilities or only offer particular products to certain customer categories based on previous history or loyalty group classification. Loyalty cards may be very 20th century, but loyalty apps are proliferating as an ideal means of understanding customer behaviour. “That is one of the biggest paybacks of loyalty schemes,” adds Sears-Black, “although promotional targeting still tends to be by purchasing habits rather than customer profitability.”
“Promotional targeting still tends to be by purchasing habits rather than customer profitability”
Using loyalty card data to classify customers is not new: after Boots launched its Advantage card in 1997 it very quickly identified a number of customer types. Among them were “the deal seekers”, who only ever bought promotional lines, and “the stock pilers”, who bought in bulk when goods were on offer and then didn’t visit the store for weeks. In those days Boots used the data to assess the true cost of its promotions including loss of full-price sales as the “deal seekers” and “stock pilers” opted for discounted lines instead.
Today, loyalty apps play a similar role with technology driving rather more sophisticated targeting as a result. “We are seeing retailers look at differential pricing based on loyalty card behaviour,” says Jason Shorrock, VP of Retail Industry Strategy EMEA, at JDA. “For example, customers can use loyalty apps to give them a special price when they scan an item in-store.” Online, high spending loyalty customers can also be identified with some retailers using different card types – black, gold, diamond, etc – to trigger specific offers. “Diamond customers may be given free delivery, for example,” says Shorrock. “It’s a very broad brush approach but it is starting to happen especially at the luxury end and with some clothing retailers.”
At NRF in January, JDA launched its Retail.me platform which uses cloud-based technology to enable assortments and promotions to be personalised to customer segments. The company is also working on the next generation of its fulfilment engine which will include various parameters nudging retailers towards that elusive cost to serve figure. It will answer those questions around the cost to service a particular order or satisfy an individual customer’s demands,” says Shorrock.
Big data, cloud-based technology, clever analytics to extract and model data from multiple sources – the technology is certainly there to enable cost to serve and customer profitability calculations to a highly granular level. But will retailers actually start doing this in the near future? Dan Murphy adds a reality check: “In theory it’s possible. The data and systems now exist. But we all know that most large retailers buy too many slow movers, end up with huge markdowns and do promotions with such a broad brush that they have no idea which customers are affected. To believe that they will start measuring customer spending patterns on the molecular level would be crazy. For a start, they don’t have any data scientists working for them – and if they did the buyers would never take any notice of them.”
Until that happens rules-based fulfilment may be the preferred option for dealing with those loss-making shoppers.