To understand this changing landscape, Kellogg Insight interviewed Jim Leczynski, clinical associate professor of marketing at the Kellogg School and former Google vice president of consumer solutions. He discusses the current state of e-commerce and how companies are thinking about reaching customers in a multi-channel world.
Because in e-commerce nothing stays still for long. “Artificial Intelligence is playing a growing role consumer journey to buy decisions,” says Lecinski. “AI is predictive and genital Capabilities are now able to offer shoppers a superior online experience.”
This interview has been edited for length and clarity.
Kellogg INSIGHT: I think a lot of people assumed that the pandemic would lead to a permanent rise in e-commerce. Has this happened? How important a role do you see e-commerce playing in the future?
Jim LECINSKI: Before the pandemic, e-commerce adoption was steady. During the pandemic it soared, then fell and is now back to its pre-pandemic adoption rate. It’s never going to be 100 percent. We have to think about it from the consumer’s perspective. Where they choose to shop – online or in person – depends on what they need to make their purchase decision. And this varies from customer to customer. Some consumers want to taste or touch a product—lie on a mattress, try on shoes, or check the ripeness of fruit. Other customers will never go to a retail store for anything. They are happy to have a mattress or any other item that appears in the box. It depends.
INSIGHT: Can you talk about how companies like Amazon, SHEIN or Temu approach e-commerce? What is the value proposition for brands that tap into these markets?
LECINSKI: There are over a hundred of these online marketplaces, such as Amazon, Tmall in China, Lazada in Southeast Asia, or Mercado Libre in Latin America. Some, like Amazon, offer products in multiple categories. Others are narrower — for example, they only offer hotel rooms or handmade crafts. But what they all have in common is that they bring together many buyers and sellers in one place, where a transaction can be completed. It is estimated that Amazon, eBay and Walmart Marketplace together receive more than three billion visitors each month in the US
These marketplaces also provide a range of services to brands, including platform advertising and back-end fulfillment. For example, Fulfillment by Amazon can handle a brand’s picking, packing, shipping, inventory management and customer service needs, including returns. Of course, these additional services come at a cost. Overall, it’s not uncommon for e-marketplace fees to amount to almost half of a brand’s overall profit margin on each sale.
INSIGHT: Over the past decade, we’ve seen the rise of direct-to-consumer (DTC) brands competing alongside traditional brands that also sell online. Can you talk about the advantages and limitations of DTC strategies?
LECINSKI: DTC brands use what we call a single-channel path to market—there’s only one way to buy. Warby Parker started this way. If you wanted to buy their glasses, you either went to their website or opened their app. This channel strategy serves some buyers, but not all. Some customers want advice, expertise and service when they try a product. This is why, for example, Macy’s still has makeup counters. Other customers want an in-store option for immediacy. Even same-day shipping or Instacart in three hours isn’t good enough. They want it now.
So the DTC route only appeals to a certain set of buyers and, by default, means those brands can only scale to a certain size. Even successful single-channel DTC brands struggle to grow much beyond about $300 million. This is because they have absorbed all the buyers whose service requirements are met through this uni-channel.
INSIGHT: Interesting. So if you’re a single-channel DTC company looking to grow, what are your options?
LECINSKI: One option is to expand into omnichannel, where you sell online and in stores, whether it’s your own stores or places like Walmart or Macy’s. Warby Parker did just that. It is sold through more than 200 retail stores as well as online. It now has $600 million in sales—half of which is DTC, and half of which is brick-and-mortar. For each DTC brand, each store expansion option erodes the profit margin. But they have no choice if they want to keep growing.
The other option is to go omni-channel, which connects everything across channels for a seamless experience. But an omnichannel strategy is very difficult to execute and not many brands do it well.
INSIGHT: How come? Can you give us an example of a company finding success with an omni-channel strategy?
LECINSKI: H&R Block was originally a one-channel brand. If you wanted your taxes done, you call on the phone, make an appointment, put your receipts in a shoebox, go to the branch and the accountant takes your taxes and calls you when they’re done.
Later, he bought a software company and renamed it H&R Block online, so they became omnichannel. But problems arose when people doing their own taxes had questions. Customers would call their local H&R office, but the agent wouldn’t be able to help because their online system was completely separate.
H&R Block has now stitched all these pieces together, using a persistent customer ID, making it an omni-channel strategy. It took them many years and an investment in technology to get to this point.
INSIGHT: Let’s talk about this technology investment. Why is it so hard to develop an omnichannel strategy?
LECINSKI: Well, for this to work, companies need a unified view of the customer regardless of the channel they use. The problem is that customers enter their ecosystem in many different ways. They might sign up for a newsletter, attend an event, fill out a warranty card. Companies need to know you are you, however you enter.
So the technology required to deliver this omni-channel experience is one that absorbs all of those past interactions, anonymous or not, and uses AI to make predictions about what the consumer might buy next.
The goal is a seamless experience for the user on the front end, stitched together by the technology on the back end. It’s easier if you’re H&R Block, with your own stores and software. But what if you’re Nike and you sell at Dick’s or Footlocker in addition to your own channels?
INSIGHT: Then I can imagine it becomes a question of how to integrate various data sources.
LECINSKI: Right. There are two ways. One is to have a formal data sharing agreement between the companies they pass data back and forth to. Alternatively, I can have what is called a “data clean room” in which each company shares anonymized customer lists, which are then merged. From there, Nike can ask Facebook, Google, or Instagram to build a model and show Nike ads to people who look like those anonymous customers.
INSIGHT: I can imagine that the changing landscape of data collection—including Google’s phasing out of third-party cookies during 2024—will change the way data is collected and developed for online marketing.
LECINSKI: Yes. Either marketers lose all personalization and revert to broad, untargeted, untailored strategies—which is not a good choice because customers want brands to talk to them as if they know them—or they use what we call “data zero-party” or “first-party”. Zero-party data is information that customers volunteer, such as when they are incentivized to tell a company their birthday for a 10% discount. First-party data is obtained through business transactions, such as when check out online and enter your shipping address.
To compensate for the loss of cookie targeting, companies can create ways to collect this data—for example, by asking people to sign up for a loyalty program. There is a long list of ways I could capture first-party data. is a big push right now for marketers.
INSIGHT: Can you say more about it? What does zero and first party data collection allow companies to do?
LECINSKI: If I have this data, I can use technology like a customer data platform (or CDP) to deliver a personalized experience. A CDP ingests customer data from various sources and then creates a unified view of the customer. It then uses artificial intelligence to predict the next best experience for that customer—the product, offer, timing, and messaging the customer or customer group receives is informed by this data. Thus, customers gain timely, relevant engagement with the brand. This will improve a brand’s business results and improve customer satisfaction. An example of this approach is when you receive a personalized message and offer from Starbucks in the app based on your past purchases.
INSIGHT: Are there certain industries where this happens faster than others?
LECINSKI: It doesn’t happen quickly in heavily regulated industries like pharmaceuticals, finance or healthcare.
Industries that will move quickly first are unregulated and already have a wealth of first-party data and a higher degree of buyer frequency. Certain types of retail are at the forefront of this because they have recognizable, often repeat shoppers. For example, I might only buy one sweater from your brand every few years, but I shop at Jewel, Safeway, or Instacart every week, or go to Starbucks every day. It’s growing fast, especially if you have a technology like Instacart at the forefront.
INSIGHT: This is interesting because many of the companies you mentioned are more traditional brick and mortar companies and not DTC companies.
LECINSKI: Correctly. Now the other thing that is needed is some vision. It’s not just your situation, but do you have the leadership, vision and will to move in this direction?
INSIGHT: So it’s a question of leadership.
LECINSKI: One hundred percent. It always is. Often companies see this as a technology problem. Well, technology can help, but it will never completely solve it. It is a matter of people, leadership and strategy of where to play and how to win.
