The Omnichannel Grocer's Dilemma

The biggest question for the future of online grocery commerce comes down to the last mile.

As in, who's going to drive it—retailers or customers?

Food retailers who want to use online commerce will have to make a hard choice between three major options: delivery, store pickup or both. They'll have to look at capital and labor investments, how much (if anything) to charge, how to meet the challenge of Amazon and other competitors, whether and how to use third-party services, and, above all, what their customers really want—and what they'll want in the future.

Some retailers are at least experimenting with offering both. Walmart offers in-store pickup at about 600 of its 4,600 U.S. stores, covering 100 markets. It also is dabbling in delivery, conducting test markets in Phoenix, Denver and San Jose, Calif. Hy-Vee, a chain of more than 240 supermarkets based in Iowa, is offering both home delivery and in-store pickup. Kroger, which has been a pioneer in store pickup through its ClickList platform, has offered home delivery for years in Denver through its King Soopers banner; it is now experimenting with it in the Cincinnati and Memphis markets.

Publix Super Markets is one of the grocers using third-party services like Instacart to make deliveries.

These retailers' approaches differ in some significant ways. Kroger and Walmart are using third-party delivery services, usually crowdsourced ones like Instacart, Shipt, Uber and Lyft, for their experiments in home delivery. Hy-Vee is using company-owned trucks, as is Randalls, an Albertsons-Safeway banner that started home delivery in Houston and Austin in late November. Hy-Vee doesn't charge a delivery fee for orders over $100, while Randalls charges $12.95 for orders under $150 and $9.95 for orders of $150 or more. Some retailers take in-store pickup literally, making shoppers enter the stores to get their orders; others offer curbside service, or even, have designated pickup lanes in their parking lots.


Information on which option is currently more developed, click-and-collect or home delivery, is hard to come by. Retailers participating in one or both tend to keep those data confidential. But data on the popularity of online service, in whatever form, are more available. And it leads to the first question retailers have to ask themselves: whether it's worth investing in online service at all, when it's still a very small piece of the U.S grocery market.

Most estimates of online grocery business in the U.S. range from only 1.5 percent to 4 percent by dollar volume. A MarketForce survey of more than 10,000 consumers in February 2016 showed that only 5 percent had used online ordering with delivery in the previous three months, and only 2 percent had used click-and-collect. According to a 2016 survey by Morgan Stanley, only 2 percent of groceries in the U.S. are currently purchased online.

On the other hand, the current state of affairs could just be a sign of the online market's untapped potential. Food retailing has traditionally lagged behind most other retail sectors in technological innovation. Packaged Facts predicts that online grocery sales will hit $100 billion in sales, giving it about 10 percent of the total U.S. grocery market, by 2019.

Many industry analysts believe that, of the online shopping options, click-and-collect has the most potential.

"Home delivery is not profitable for anyone, including Amazon, and the infrastructure to get it right is expensive," says Randy Evins, senior principal, industry value engineering (IVE), SAP Food Drug & Convenience at SAP Retail. "Grocers' bottom line is slight to begin with."

Walmart offers online service for pickup at stores, in some cases delivered to the customer's car.


Click-and-collect has the advantage over home delivery in terms of both grocers' current (and likely future) capabilities, and customer preferences.

"I think the growth profile of that service [click-and-collect] is higher than delivery," says Randy Burt, a partner at global management firm A.T. Kearney. "It fits well with existing customer routines."

Those routines often are tightly timed. Click-and-collect allows customers to decide exactly when they will fetch their groceries, while even the most sophisticated delivery service usually involves at least a one-hour arrival window.

"I think we will see the click-and-collect model," says Deborah Weinswig, managing director at Fung Global. "I don't think the customer wants to wait in their home, in the specific delivery window, for their items to be delivered to their home."

There's also the question of consumers wanting to see their food before they buy. In a 2014 BMO Capital Markets survey, those who said they were unwilling to buy groceries online were asked for a reason. The most popular one by far was "prefer to pick out my own foods," cited by 37 percent of respondents in New York, 34 percent in Los Angeles and 31 percent in Seattle. (The second most popular response was "Do not want to pay for delivery service," cited by 16, 19 and 19 percent, respectively.)

"There's a lot of elements of food that are experiential, and that physical interaction is going to be very hard to digitize," Burt says.

That applies more to certain foods than to others, says Jordan Rost, vice president of consumer insights at Nielsen.

"The store still plays a crucial role in consumers' ability to decide what foods to buy, particularly for fresh produce," Rost says. "Compared to the average CPG purchase, consumers are more open to a click-and-collect model for fresh produce and frozen foods. As retailers try to figure out the economics of offering fresh foods online, the click-and-collect model is an easier way to offer consumers online convenience and an in-store experience."


Click-and-collect allows customers to pick out such foods themselves, if that's what they prefer, and leave others to store employees, Weinswig says.

"A lot of those items that are less interesting to pick out, they can actually have someone pick out for them," Weinswig says. "Those items that are more important to them, like meat and produce, it allows them to be involved in that part of the visit."

"The store still plays a crucial role in consumers' ability to decide what foods to buy, particularly for fresh produce."


Beyond specific categories of foods, click-and-collect offers an in-store experience that can be potentially valuable to both customer and retailer, by exposing them to other products and potentially prompting unplanned purchases.

"I think so much of what's missing right now from retail in general is this kind of unplanned purchase," Weinswig says. "I just think entertainment in general is missing from the shopping experience."

Click-and-collect taps into an inherent advantage of conventional grocers: their existing bricks and mortar. By using an infrastructure already in place, click-and-collect gives grocers the opportunity to ramp up online operations without the big capital outlay that would be needed for a delivery fleet and, perhaps, a warehouse.

"If we didn't have the physical brick-and-mortar infrastructure in place, if we didn't have the world's most efficient supply chain, it would be much more capital intensive than what it is," says Walmart spokesperson Ravi Jariwala. "As it is right now, we are using our stores and that efficient fresh supply chain to fulfill online orders. So we're not going to build new distribution centers or warehouses just for fulfilling online grocery orders."

Walmart is also trying a new wrinkle in online service: centers dedicated to pickup of online orders. These stores of about 4,000 square feet sell gasoline and convenience-store items as well as serving as pickup centers, with orders being fulfilled from nearby Supercenters. The first such store was in Huntsville, Ala., and a second opened in November in suburban Denver.

But that's not to say that click-and-collect is investment-free. Depending on what's offered, it will take at least additional labor and software—and sometimes capital investment for things like dedicated pickup areas in a parking lot.

The picking process will in most cases represent the bulk of the expense, in labor and software, both customer-facing and in-store.

"Fulfillment in store is the most expensive part of the process, and grocers need to think about the store as if it were a warehouse, not a display," Evins says.

Burt echoes that thought: "All the work that gets done on warehouse-picking efficiency, that needs to occur in the store now." Training workers to use software, to pick efficiently and not get in the way of in-store shoppers, making sure enough product remains on the shelves for those shoppers, are all concerns, he says: "You want to be very careful about disrupting the shopper experience."

That's why click-and-collect is no panacea. Publix started testing curbside service in August 2010 at two stores, one in Atlanta and one in Tampa, at $7.99 per order. It ended the experiment in January 2012, saying on its website that there simply wasn't enough interest: "The majority of our customers love their Publix and still prefer to have the interaction with our associates."

"All the work that gets done on warehouse-picking efficiency, that needs to occur in the store now."

A.T. Kearney

Now Publix is trying out home delivery—not for the first time. It tested delivery from 2001 to 2003 in company-owned trucks, but called it quits because not enough people signed up. This time around, Publix has forged a partnership with Instacart, a delivery service that uses on-call drivers paid per trip, ordering through Instacart's own app, which is synced up with Publix's. Instacart has similar agreements with Whole Foods Market, H-E-B, Target, Food Lion, Costco and others. Shipt, Instacart's major competitor, has similar arrangements with Kroger, Meijer and other grocers. Walmart is using three other third-party, crowdsourced delivery services, Uber, Lyft and Deliv, in test markets in three cities (one of them for Sam's Club).

The advantage of using a third-party delivery service is that it's the cheapest way to get groceries to someone's door. Drivers usually are paid relatively low wages with no benefits, and are paid by the trip, meaning they don't burn money sitting idle. Plus, in most cases, the delivery service collects fees for each delivery from both the retailer and the customer, meaning the retailer in effect splits the delivery cost with the customer.

On the other hand, the advantage of company-owned delivery trucks, loaded and driven by company employees, is control—of quality, of data and of the customer experience. Quality control is better when a delivery is in the hands of an employee who is answerable when something goes wrong, rather than a stranger. If something goes really wrong, murky issues of liability can arise.

"I think so much of what's missing right now from retail in general is this kind of unplanned purchase."

Fung Global

Control of data and control of the customer experience are closely linked. Customers who get their groceries delivered through Instacart or another third-party service see themselves as customers of that service as well as of the grocer. This view is reinforced when the customer places the order through the delivery service's app, even when it's synchronized with the grocer's digital technology. Such an arrangement places sales data in the delivery service's hands.


Perhaps the biggest potential challenge to conventional grocers who want to do home delivery is the advent of AmazonFresh. Amazon is now delivering a full range of groceries in 10 American markets, plus London, and is likely to keep expanding.

AmazonFresh has less penetration than longer-established pure-play grocery delivery services like Peapod. It also has a higher cost, at least upfront: a $299 annual fee, compared with Peapod's fees of $6.95 to $9.95 per order, depending on size. This may contribute to the disparity between the two services' basket sizes: Peapod has an average basket size of $147, compared with $84 for AmazonFresh, according to a study by 1010data. Once an AmazonFresh customer pays the annual fee, she feels free to place smaller, more frequent orders, which costs Amazon more to execute.

But Amazon has certain things going for it over the long term: name recognition, the world's most popular consumer interface, and a seemingly limitless ability to absorb losses. This could give it an edge in the future, because as the capacity of AmazonFresh expands, so will its efficiency, in both product pricing and logistics.

"Pure online grocers such as...AmazonFresh may well be in a position to undercut brick-and-mortar supermarkets on the price of both delivery and products as they scale up their operations," writes Marc de Speville in his blog on Strategic Food Retail, the consultancy he founded.

In the end, as with most stark business decisions, it's up to the customer to choose one or the other—or both.

"Everyone's trying to figure out the right model," Weinswig says. "And maybe the right model is that you offer both. At the end of the day, the customer has to drive the decision."