Data is key to personalization-based marketing—but you may be doing it wrong. That is, while the rich data that retailers have at their disposal is enabling new levels of personalization, the marketing may be going in the wrong direction.
“What we’ve noticed, and what has been true, is that retailers establish the marketing campaign and then find the people to target,” says Anil Kaul, CEO and co-founder of analytics company Absolutdata. “While what I believe should be happening is to look at the customers you’re trying to reach, literally create a campaign for each customer, and then aggregate that.”
While that may sound like a revolution and a risky bet, Kaul told Retail Dive that retailers are increasingly finding the task to be smooth and, better yet, highly effective.
The downside of past-purchase data
One of the most confounding uses of past-purchase data, experts told Retail Dive, is that customers end up getting messages or even coupons for items that they’ve already obtained.
Getting a coupon for an item that was bought at full price can be irritating, but, at best, it’s useless because the customer already has the thing in question. Even getting a coupon for a product that you might buy again is often mis-targeted because point-of-sales systems routinely spit out a different brand.
“Supermarkets have been handing out coupons at the checkout for a decade. They do a terrible job,” says Brett Wickard, president and founder of lean retail company Fieldstack. “If you buy organic yogurt, you don’t want the Dannon coupon you get. They should hand you a coupon for another relatively healthy product. I assume that this is because the couponing is done to satisfy the vendors, and not for customer benefit.”
Stop ignoring the past
Similarly, Kaul says that customers who buy a shirt should get suggestions and coupons for other merchandise that reflects the style, price range, and perhaps the color line of the original shirt. And he has statistics that show that such “micro-campaigns” are highly effective at translating into sales.
“With this approach we see a 30% increase in conversation rate,” Kaul says. “Look at customers' history: If I buy a shirt, maybe I don't want another shirt, but I’m likely to buy a different thing.”
Yet, says Wickard, purchase history is ignored, or used with what he calls “brute force retargeting."
“You buy an air filter for your generator and then you start seeing ads for generators,” he told Retail Dive. “But your purchase history is being ignored, because your filter purchase proves you already own a generator. As a consumer, you want the opposite to happen: you buy a generator this year and next fall you get an ad for the air filter.”
Both Wicker and Kaul say that retailers have the data they need to properly communicate with customers.
“You can fake this with tighter list segmentation,” Wickard says. “Big data can classify hardware store shoppers as boaters, hunters, people into pinewood derby, knitters, woodworkers, plumbers, and so on. What’s missing is an awareness of what project somebody is working on today. The person who bought a faucet last weekend probably doesn’t need silicone tape today. She would have bought that when she bought the faucet if she needed it.”
Timing as a tool
Both Kaul and Wickard see timing as another tool for this kind of micro-marketing. As Wickard notes, there are logical future recommendations based on customer purchases, whether it’s for an item that needs to be bought again, like those air filters, or for a spring collection from a certain designer based on what was purchased in the fall.
And for the holidays, Kaul says, it’s not too much to ask retailers to pay attention to when their customers shop.
“You don’t need to send holiday emails to everyone on Dec. 1,” he says. “Based on past history, you can find out which customers plan ahead and which are last-minute. If I have 20% off deals going to customers at the beginning of December, I will have missed the ones who shop at the end of December.”
Although search is generally used with a mass approach, it can be another tool that yields a huge amount of data about customers’ intention, Kaul says.
“But you have to think in terms of micro-segments,” he says. “In this world you can’t go with five or 10 segments. People have 1600 micro-segments. If you go to Netflix, for example, they will show you ‘movies that have funny endings’ and all kinds of other categories from which to choose. As a retailer, you have to be ready with your data.”
Change the marketing mindset
And that takes a change in mindset, perhaps more even than a change in digital tools. “Retailers already have CRM systems that already keep record of these things,” Kaul says. “Even from a technology point of view, it’s an opportunity that’s in front of you — just do it.”
Wickard says that retailers can aim for the kind of personalization that a customer might find in an old-fashioned brick-and-mortar shop, the kind run by people who know you and what you might want. And that, he says, isn’t necessarily just better “marketing” but also improved “customer service.”
“We need to… think about improving the shopping experience,” he says. “We need to provide better customer service. Marketing is ‘Do you want fries with that?’ Customer service is ‘We have a new gluten-free dessert. Would you like to try it?’”
If you’re asking because your customer has demonstrated a preference for gluten-free foods, the answer could easily be “yes.”