At best, a properly integrated CRM program is complex, given the huge number of systems it touches, or at least should touch. But what happens when a $16 billion, 37-year-old chain wants to tackle CRM for the first time? Whole Foods is about to find out.
Whole Foods is hardly the only Fortune 500 retailer to have resisted CRM and loyalty programs. The biggest example would be Walmart. But Walmart and Whole Foods have resisted CRM for radically different reasons.
With Walmart, the reasons are price perception and price reality. The price reality is the incremental additional cost that CRM would add to operations, ultimately making it ever-so-slightly more difficult for Walmart to offer the lowest possible prices. But price perception is the bigger issue. If shoppers are told that loyalty cards bring lower prices, doesn't that signal that customers aren't already getting the lowest prices?
Whole Foods' reason for resisting CRM is privacy perception. Whole Foods embraces customer service as a differentiator in full-service grocery, much as does Trader Joe's in limited grocery and Nordstrom in apparel. And Whole Foods execs worried that a program that so obviously tracks customers closely might be off-putting.
But in 2017, with mobile tracking every shopper far more closely than any CRM system could, the concern may be moot. Put another way, if shoppers believe that privacy is already lost, they might at least get the discount.
This gets us into the IT mud. When a 37-year-old retail chain wants to suddenly embrace a loyalty program, should it leverage all of its existing data — scattered among various systems, such as basket analysis and item purchase histories and scant web and mobile customer activity — to get things started — or should it start anew so that the data is far more consistent?
Whole Foods is working with a vendor called Dunnhumby on category management, which will play a critical role in the chain’s CRM and loyalty efforts. According to David Ciancio, the company's senior customer strategist, "What is shifting is that data can be used in a benign way to create value for shoppers." What today's customers think, he said, is "I know you have my data. You better do something that is good for me."
But first, there are the technology obstacles. "The real barriers are POS systems in general" and "the number of different POS systems within each retailer," Ciancio said. Whether those different point-of-sale systems are there because of acquisitions or because the company is slowly rolling out a new version — or a new POS system entirely — across the chain makes little difference. At any one point, most chains simply do not have one consistent POS version across every store.
That POS integration touches heavily on terminal memory modules — "to handle the logic needed to activate a loyalty program" — plus integration of payment methods as well as mobile and web systems, Ciancio said.
Within payments, there is the ever-present debate of whether it's wise to piggyback loyalty atop specific payment mechanisms. For example, if the chain already associates a specific Visa card number with a specific shopper, why insist on a separate loyalty identifier? The card value still needs to be hashed and secured (you don't want PCI-protected sensitive data to be sitting in CRM files within marketing), but isn't that simpler?
The counter is that customers often have multiple payment cards (and they might use PayPal or some other alternative payment system online and with mobile) so you start to lose the consistent view of the customer by leveraging payments.
Although it doesn't impact Whole Foods at this time, the issues get more complicated with chains that have pharmacy departments, which ensnares them in even more restrictive HIPAA rules. A shopping cart with three prescription refills, two boxes of tissues and some over-the-counter cold pills must omit any prescription details from the CRM files.
As for what to start with, Ciancio simply suggests, "Start with what you've got. POS data alone, even without card IDs, is a really rich place to start." This gets into one of the better — and often overlooked — uses of CRM data: making strategic choices about which SKUs to renew.
This might be a case where a particular product isn't selling well at all, but the people who are buying it are among your highest-revenue regular shoppers. Before you alienate some of your best customers, maybe it's better for a category manager to just leave that broccoli-flavored toothpaste on the shelf.
Ciancio pointed also to associated purchases. He cited a small size of Frosted Flakes that may not move well, but it's a popular item with budget-conscious shoppers and it brings with it many other related purchases. Integrating CRM analysis can make those kinds of connections more obvious.
When CRM programs don't work well, Ciancio said, "most fail because they are seen as just a marketing tool" and are not integrated into other store systems.
There's actually a good reason to start anew and that is if management believes that its existing system data is flawed. For some, Ciancio said, "my data is good enough to make the decisions that I need to make" while others concede the opposite: "I know my data's a mess, and we can't do anything with it."