Since the beginning of grocery retail, those running the store have wanted to be inside their customers’ heads.
What are customers looking for when they come here? How do they shop? What is the best way to configure the store? How hard is it for them to find items? How can they be provided with the best service? What are the best ways to build loyalty to encourage return trips?
As e-commerce options grow, retailers have more ability to understand their customers, their motivation and their shopping options. Online shopping captures useful data, which retailers have been using to improve their stores — both online and brick-and-mortar. Several people who work on the data side of e-commerce described where that data comes from and how it is used to improve omnichannel experiences for shoppers — and the bottom lines for retailers.
Many retailers have had one form of shopper data collection for years: the loyalty card. In exchange for discounts and perks for consumers, loyalty cards have given retailers a look at who shops, what they buy, when they shop and how often they come to the store. But e-commerce can unlock even more shopper data.
“Purchase data, looked at in the online card first, and then in combination with others, gives you a progressively larger base of understanding where the customer is,” Bill Bishop, chief architect and co-founder of grocery analyst firm Brick Meets Click, told Food Dive. “That allows the retailer to do a better job of planning how to serve that individual customer.”
Unlocking the data
How can retailers start to unlock all of this data? And how do they start getting there?
“It begins with recognizing the future of grocery is digital,” Grant Gilkerson, data science director at 84.51º, Kroger’s analytics and e-commerce arm, told Food Dive. “There will be a role that brick and mortar will play, no doubt, but people are going to the e-commerce channel.”
But this doesn’t mean diving into sophisticated web applications or having to learn complex data crunching programs. While third-party consultants might be best to help with building the actual e-commerce site, Instacart Chief Business Officer Nilam Ganenthiran told Food Dive the data part is much easier for retailers.
“Retailers have been dealing with big data since before 'big data' was a term, if you think about loyalty programs, prices that change." Ganenthiran said. "Retail is a data-centric business, and people don't appreciate how much data really is in the system. Probably more than any industry. Retailers are highly comfortable and numeric and data-oriented. We've just been helping in bridging that for the e-commerce world.”
“Retailers have been dealing with big data since before 'big data' was a term, if you think about loyalty programs, prices that change. Retail is a data-centric business, and people don't appreciate how much data really is in the system."
Vice president business development and strategy, Instacart
Dave Clements, head of global retail at dunnhumby, agreed. While e-commerce produces extremely rich and detailed data, he told Food Dive that a retailer does not have to dig very deep to make a demonstrable difference in sales.
“You don’t have to interpret to recommend action,” Clements said.
But data shouldn’t be just a buzzword, Ganenthiran said — the type of thing that grocers collect simply because they can and because others say it is powerful.
“Data is only as powerful as what the party can do to use it to help grow their sales,” he said. “Our philosophy is we work in concert with our retailers, sharing back with them because they will use it to help grow their sales. And that's a win for us, a win for them, and it's a win for the customer.”
The data crystal ball
When a grocery store enters the e-commerce space, it aims to translate its core offering online and, oftentimes, onto a mobile app. But Chris Bryson, founder and CEO at Unata, told Food Dive that is easier said than done. Unata creates and manages e-commerce sites for grocery stores, and was acquired by Instacart in January. Bryson said It’s very difficult to create an online replica of a grocery store because of the complexity of the store.
But the website or app, for all intents and purposes, is the online store. Megan Moglia, vice president of customer experience and loyalty at 84.51°, said that the online experience is all-encompassing. Shoppers need to be able to find the product, its price, its accurate size, and if there are any promotions associated with it.
“How do we create an experience that truly solves the customer's problem?”
Vice president of customer experience and loyalty, 84.51º
“How do we create an experience that truly solves the customer's problem?” Moglia said to Food Dive.
However, all the variables at play, including available items — which are not standard, even for a firm like 84.51º that only works with one retail company — basket size and the challenge of finding those items make grocery e-commerce daunting, Gilkerson said.
Personalization based on user behavior is one of the places grocers can try to differentiate themselves online. Moglia said the quest for better personalization is one of the things that drives 84.51º. If a user went to Kroger’s website and did a search for “chicken,” she said, it makes the most sense if the search results are relevant to the shopper. That search could pull up a wide variety of products, from chicken breasts to chicken-flavored instant ramen to Chicken in a Biskit crackers. If the system could learn that a particular shopper is searching for the meat, the search function would pull up more targeted selections in the future.
“We have other kinds of personalization services that we hope are like a quick start to the basket building process,” said Gilkerson. “One of our personalization recommendations is called 'Start My Cart.' ... It's showing you as soon as you come to Kroger's website ... before you even start typing anything, as soon as you land on that page, it's showing you at the customer level the types of products that you're typically buying when you do your grocery shop.”
Kroger’s investment in data is paying dividends. The retailer has seen massive growth in e-commerce sales, both through delivery platforms and ClickList, its click-and-collect service. As overall sales grew last year, digital sales posted a 90% increase. And sales are still climbing; according to a transcript of the retailer’s most recent earnings call last month, Kroger CEO Rodney McMullen said digital sales were up 66% in the first quarter of this year.
“It's one thing if you can get a customer to try ClickList or try grocery e-commerce once. The bigger challenge is to engineer an experience where they want to continue using it,” Gilkerson said.
At dunnhumby, Clements said, extremely rich data about what shoppers want, the path they take to finding it, other products they have considered, and the product information they look at before adding items to their carts helps build better shopper profiles.
“If you know what a customer has bought before, you can serve up ... immediately, they're only seeing the promotions they're really interested in."
Head of global retail, dunnhumby
“If you know what a customer has bought before, you can serve up ... immediately, they're only seeing the promotions they're really interested in," he said.
Using this data to personalize products recommendations to shoppers is one of Clements’ favorite parts of e-commerce. This can sometimes take the form of reminding shoppers of items that they usually get, but don’t have in their basket, like a weekly shopping trip that is missing the usual quart of 2% milk. But it can also take the form of comparing a shopper’s basket to others, recommending items that others generally pick up. Or looking at items in a shopper’s cart and recommending popular items to go with it — like potato chips and canned baked beans with a package of hot dogs.
Clements said this is an extremely successful tactic. One in five shoppers actually purchases items that are suggested to them.
“It’s really kind of being a personal shopper,” Clements said.
This personalization can also work with coupons. One of the areas Unata works on is integrating paper coupons with digital ones. With a weekly print ad, Bryson said, no more than a quarter of the weekly specials at a grocery store — which can be about 1,000 — can be featured. Paper has its limits.
“With digital, you don’t have a limit,” he said. “We can show the shopper, ‘Here are items you’ve bought before that are on sale and you should know about.’ We can make the specials curated by what the shopper has seen.”
And it works. Bryson said at Canadian grocery Longo’s, Unata was able to increase shopper spending by 16% with targeted coupons.
Clements said that even though this level of personalization leverages a lot of shopper data, customers generally don’t seem to mind. Bombarding the customer with ads or blatant suggestions, on the other hand, may cross the line.
“Customers see things as helpful when they are helpful,” he said.
Monitoring in-store behavior
Imagine being able to watch a shopper as he enters the cookie aisle, looking for his favorite snack. You know what the shopper is looking for. You know where he is looking to find it, and you know how long he searches. And you know if he’s ultimately successful in getting the cookies into the shopping cart.
This isn’t a mundane scene in a movie or television show. It’s exactly the kind of data that Instacart gathers. The grocery shopping and delivery service, which is currently partnered with retailers in more than 200 markets nationwide, gets data from the in-store shoppers who fulfill orders placed through its online platform.
Instacart shares all data it collects with its retail partners — which Ganenthiran said sometimes blows their minds.
“I think that when people think about Instacart first, they think, 'OK great, it's this website I can go on and get some sales,' " he said. “Absolutely, that's all true. I don’t think they think of the partnership on the operational side, on the core business, on the brick-and-mortar business can get better using the data signals online.”
Because Instacart’s shoppers are looking for a very specific list, and because they are constantly checking in — when they get to the store, when they are looking for an item, when they find an item (or when they do not) and when they enter and exit the checkout line -- it’s easy to use this data to discover operational issues in the store.
"I don’t think they think of the partnership on the operational side, on the core business, on the brick-and-mortar business can get better using the data signals online.”
Vice president business development and strategy, Instacart
For example, Ganenthiran said, there was a retailer who was getting an abnormal number of Instacart shoppers finding there were no strawberries at a certain time of day. Instacart sent people into the store to take a look at what was going on. They found that there were berries on the floor, but the ones that were there were in poor condition — the type of berries that Instacart trains its shoppers not to purchase. Instacart worked with the retailer to change its restocking flow.
“That was a gap that got filled,” Ganenthiran said.
Instacart also uses its data to help retailers improve their merchandising, Ganenthiran said. They can suggest what stores should carry, based on what the end consumer is searching for and not finding. They can tell a retailer when it has a problem with products being out of stock — including the day and time more is needed. Ganenthiran said that these are things stores already track, but Instacart’s shopper-level data tends to be more detailed and helpful.
The shopping service also collects checkout-level data, so it can measure how long it takes the average consumer to check out — and how to further streamline the process.
Ganenthiran said he is used to retailers’ surprise at the level of granularity that Instacart’s data can give them, but they accept it quickly — even if they don't use it to take immediate action.
“It’s usually not a problem of believability because the data is the truth. It's just telling you the reality,” he said. “I think what it comes down to is it's always two-fold. One is actionability — can I actually change the outcome by doing something based on this data — ... and the second is prioritization.”
Bryson said it's time to streamline e-commerce, going from something grocers have to something that works better. Many grocer e-commerce websites suffer from “island syndrome,” where different providers run different parts of the ordering and fulfillment process. For example, he said, if a grocer has a place where shoppers can custom design cakes at the bakery, the back end for that area might be run by a different provider than the rest of their site. Bryson said he sees this ending in the next two years, as grocers invest more resources into their online channels.
Moglia said getting everything in place to present an accurate representation of grocery shopping is a significant challenge. After all, she said, shoppers cannot touch, smell and taste products online.
"How are you using this content online to make a more pleasant and functional experience, rather than adding to basket from list?”
Head of global retail, dunnhumby
Clements agreed, saying that a majority of shoppers still feel like going into a brick-and-mortar store is better than ordering online. E-commerce is a less impulsive shopping experience, he said, so brands need to work harder to win the same shoppers.
“How do you close that gap?” he asked. “...How are you really getting across the fresh experience? Where the products are sourced from? Using more and more content, whether its straight webpages or increasingly videos, how are you using this content online to make a more pleasant and functional experience, rather than adding to basket from list?”
But on top of that, a solid and working integration with store loyalty programs is essential. Kroger is actually working on some of this sophisticated technology, hoping to roll “Kroger Edge” shelf-side information that interacts with shoppers’ smartphones and shopping lists out to about 200 stores this year.