Listen to the podcast:
Omnichannel marketing seems like a simple enough concept. Consumers like to shop online, offline, and across different channels, so firms need to meet them wherever they are. But coming up with an omnichannel marketing strategy is a lot more complicated than just collecting cookies and tracking purchases. A new study that appears in a special issue of the Journal of Marketing in collaboration with the Marketing Science Institute explains why omnichannel is not a panacea.
There are three big challenges to making it work. Those challenges are outlined in the study, along with some solutions that include using machine learning and blockchain technology to harness the full benefits of omnichannel marketing. Wharton marketing professor Raghuram Iyengar is a co-author of the paper, titled “Informational Challenges in Omnichannel Marketing: Remedies and Future Research.” The other co-authors are: Tony Haitao Cui, marketing professor at the University of Minnesota’s Carlson School of Management; Anindya Ghose, marketing professor at New York University’s Stern School of Business; Hanna Halaburda, technology, operations and statistics professor also at NYU Stern; Koen Pauwels, marketing professor at Northeastern University’s D’Amore-McKim School of Business; S. Sriram, marketing professor at Michigan University’s Stephen M. Ross School of Business; Catherine Tucker, management and marketing professor at MIT Sloan School of Management; and Sriram Venkataraman, marketing professor at the University of North Carolina’s Kenan-Flagler Business School.
Iyengar joined Knowledge@Wharton to talk about the findings. Listen to the full podcast at the top of this page or keep reading for an edited transcript of the conversation.
Knowledge@Wharton: Not only are firms trying to execute omnichannel marketing better, but researchers like you are trying to understand it better, even while the rapid evolution of technology makes that a moving target. What does this study add to the literature?
Raghuram Iyengar: Omnichannel certainly is a very hot topic. When companies are thinking about omnichannel, they sometimes want to think about distinguishing from multichannel. The big distinguishing aspect of it is multichannel has different ways in which you’re reaching the customer. Omnichannel is that as well, but it should be in synergy.
If you are, for example, a customer of REI, you might have a mobile application, you might have emails coming in. And if they are pursuing an omnichannel strategy, they are hoping that the customer is seeing different pieces of information in conjunction with each other and, in some sense, are complementary to each other.
Carrying that out is not that easy because you need to have a good sense of what the data is like — all the different touchpoints that the customer has had with REI or any other company — and then be able to execute it on the back end. Putting it all together is not as simple as it seems.
“Especially in the last year or so, consumer behavior has changed. What was working the year before perhaps is not going to work today.”
Knowledge@Wharton: The paper is organized into three distinct challenges and remedies that are easy to follow. The first challenge is about data. What is the issue?
Iyengar: Let’s say you go to Nordstrom on their website and shop for something. And then you decide to go to the store and shop for something else. The hope would be that Nordstrom would have all your data in one place: going online, going to the store, perhaps using a mobile application. But the reality for many companies is that much of their customers’ data is very siloed. Why? Because different departments are in charge of different parts of the journey. There might be an online department. There might be an in-store department, and so on and so forth.
These people are looking at different snippets of data, so sometimes in large companies, the data becomes siloed. This could be for various reasons. Some of it could be political, because some people want to take charge of data that is perhaps more important for revenues. And some of it could be that analysts just don’t know where the data is.
Knowledge@Wharton: How do we apply technology solutions to that?
Iyengar: Some of it, of course, is forcing the silos to be taken away within the company. Again, this is easier said than done, but it has to be top-down. Companies have to realize what is the value being added by some of those silos being taken away.
Another set of solutions comes from more machine learning-type examples. It may not be for every retailer, but you can imagine in certain regulated industries, even if they do want silos to go away, it may not be easy. In financial industries, for example, one part of the company might be interested in certain applications, and another part might be in another. But because of regulation, they can’t talk to each other.
There is something called predictive learning, which is a type of machine learning where you can imagine data sitting in different places, and a central kind of — you can call it an algorithm process — [where] each of the data by themselves is anonymized. In that sense, you can mix in the secret sauce, so to speak, without any of the ingredients coming together. That might be a good analogy. That’s one way to do it where silos [exist] because of regulation. There are these kinds of solutions which, increasingly, many companies are thinking about.
Knowledge@Wharton: Let’s go to the second challenge that you present in the paper, which is about marketing attribution. What is that, and why is it a problem?
Iyengar: Let me give you an example and go with Nordstrom again as the representative company. Imagine you get an email. You [think,] “Well, that’s interesting. Nordstrom is sending me an email. Let me look at what the offer is.” I’m assuming Nordstrom also sends some catalogs, perhaps. You might go into the store. Let’s imagine that Nordstrom’s data is not siloed, and at some point, they see that you have bought something. Marketing attribution is about which part of this touchpoint was responsible. Was it the email? Was it the catalog? Was it something that the salesperson did within the store? Perhaps all of them were responsible for that conversion that happened. But they’re also thinking about how much of that conversion can be credited to each of the different touchpoints. That’s what attribution is all about. How do you attribute the conversion at the end, or lack thereof, to what happened along the way?
“The issue with privacy is that customers may not want to give you that data.”
Knowledge@Wharton: What are some of the solutions in shoring up marketing attribution?
Iyengar: There are many. Let’s start with the simpler ones. In fact, what I’ve seen some companies do quite actively is this idea of testing and learning. Going back to that email example I was using, sometimes a company might say, “Well, let’s try to see what happens if we don’t send that email.” Then in a systematic manner — like test versus control — people are randomly assigned. Some people get an email; some people don’t. And then they track throughout the entire customer journey to see what happened to people who got the email versus not. What are they trying to do? Holding everything else constant, they’re trying to change one part of that journey to then be able to see what are the impacts of changing that one part.
We can be a little bit more demanding in terms of coming up with an experiment. We change multiple parts of that journey in a systematic manner. That’s basically this idea of testing and learning. For example, recently I was talking to Hershey’s CMO, and she mentioned that especially during the last year, they’ve been experimenting with different types of media mixes to see what works and what doesn’t.
It’s all about trial and error. If you knew the answer at the onset, you’d go ahead and go with the answer. But many times, the context is changing. Especially in the last year or so, consumer behavior has changed. What was working the year before perhaps is not going to work today.
Knowledge@Wharton: The third and final challenge in the paper is about data privacy. We hear about these issues every single day, especially with General Data Protection Regulation (GDPR) in the European Union and other measures being proposed in this country. What do you and your co-authors say about this issue of data privacy?
Iyengar: When you think about all the wonderful things that omnichannel marketing can get you, which is the synergistic view of the customer, the 360-degree view where you can see the customer at all the different touchpoints, the issue with privacy is that customers may not want to give you that data. Especially with GDPR coming in and the California Privacy Rights Act coming in the U.S., a lot more control is being given back to the consumer. For example, the latest update from Apple is basically demanding that consumers give their approval for certain apps to track their information. All of this is giving a lot more information and control back to the customers, and now it’s up to the customers whether they would like to see some benefit from the data that they’re sharing.
What you’ve seen is that privacy is not one-dimensional. It’s not a yes vs. no answer. It’s something that customers have to think about. How comfortable are they with sharing? I’ll give you an example. I like drinking coffee. If I go to a website and it says, “Well, based on your preferences, here’s a coffee blend that I would recommend.” Great. Thank you. On the other hand, I don’t want my healthcare records being shared. I think it’s a question of what context you are in. It’s a question of what exact benefit we are getting as a customer, such that you might be more inclined to share that data.
“Do we actually need the most granular data for making appropriately good targeting decisions?”
Knowledge@Wharton: Your paper also says that customers feel more comfortable when there is data transparency, when companies are telling them what they are sharing and letting them opt-in or opt-out. You also talk about using blockchain technology to help address privacy issues. Can you explain how that can help?
Iyengar: For those of us who might be uncomfortable or unfamiliar with blockchain, think of it as a distributed ledger. You might have your own personal tracking of the money that you’re spending on different things. Think of blockchain as a giant audit book where things are being tracked, and it’s publicly available. But once the record is in there, it’s immutable. It doesn’t change as fast. The idea would be that you can imagine customers giving up certain information within the blockchain and firms being able to access that information to appropriately target customers. This is a great way of keeping track of what information firms are using, and then consumers or customers can demand appropriate compensation for using that information.
Knowledge@Wharton: Omnichannel is an emerging area of research for marketers. What do you hope to study next?
Iyengar: I think this idea of privacy and using machine learning and new technologies is very interesting to understand. This is some of the work that I’m doing with my colleague, Wharton marketing professor Eric Bradlow, and grad student Mingyung Kim. We are looking at, for example, the following question, which is something very much of interest to Apple and other companies: Do we always need the most granular data to make good decisions?
Imagine we have individual data from customers. Then let’s imagine we have slightly more aggregated data, which might be groups of people, and so on. Do we actually need the most granular data for making appropriately good targeting decisions? At what point does some of the more granular data have a lot of noise, whereas slightly more aggregated data smoothes out that noise? We are trying to see what kinds of models can be built on slightly more aggregated data that might do quite well. What does this mean for privacy? You can imagine individuals may not want to share their particular data per se, but they might be comfortable if they are part of a data set: “We are aggregated with other customers.”
I think that’s an interesting area of research where it intersects between privacy and machine learning and other kinds of models. That’s something I’m pretty excited about, to see how we can use data of different kinds to still make good decisions and, at the same time, respect people’s privacy.