Voice assistants like Amazon’s Alexa and Google Assistant have been making headlines in recent months, thanks to their rapid proliferation of the market. At last count, NPR and Edison Research estimate that one in six Americans own a voice assistant device, or about 39 million people. In response to this growth, retailers have been quick to explore voice search and how they can surface their products on these devices.
Niraj Dawar, professor of Marketing at Ivey Business School, and author of “TILT: Shifting Your Strategy From Products to Customers,” says that voice is just the first step in a much bigger transformation — the rise of artificially intelligent assistants. Encompassing voice, as well as mobile devices, and other technology that can track consumer behaviors, AI assistants will deliver the products that consumers need most without them even having to ask.
Niraj believes that AI assistants will monitor a consumer’s consumption of laundry detergent, for example, and understand when the consumer will need more, what brand they prefer, and where to get the best deal. Then the assistant will automatically purchase the detergent that meets the consumer’s most important criteria. The need to make a trip to the local supermarket or big box store will essentially be eliminated.
“That will change just about everything for retailers,” says Niraj, “from their business models to the kinds of products that they sell.”
In the following interview with Sidecar, Niraj explains how AI assistants will transform the retail industry by usurping the coveted “last customer interaction” from retailers. He also details how retailers can remain relevant in a world where consumers trust AI to do their shopping for them.
Ellen Harvey: You’ve written quite a bit about AI assistant technology and how you expect this to impact the future of retail. Why do you think it’s an important topic for retailers to consider in the future?
Niraj Dawar: I think it’s worthwhile looking at the past in order to be able to think about the future. Retailers over the last 20 or 30 years have become increasingly concentrated, and they have taken possession of the last customer interaction in the supply chain. As a result, they’ve become very, very powerful. The concentration of retailers is such that in most developed markets today, the top three retailers account for as much as 60% to 80% of all groceries sold, for example. They are powerful retailers that are global, such as Walmart, Carrefour, and Albert Heijn.
That ownership of the last point of interaction has allowed retailers to do things like develop their own private label brands and develop their loyalty programs. Together, these two things are extremely powerful. You talk to a brand manager in any consumer goods companies over the last 25 to 30 years, and they will tell you that retailers have staked out a very critical position in the value chain.
Now, what is happening is that with technology emerging, retailers such as Amazon that were born on the web are changing that last customer interaction. They are changing it by examining how to reduce the customer’s costs and risks of interaction.
An early step in that direction was one-click buying and easy returns. The Amazon Dash Button is another example. It is a very simplified way of reordering products that you consume on a regular basis. It’s also an intermediary step towards the next stage where artificial intelligence platforms will monitor consumers’ consumption of products and be able to determine what the consumer needs next.
What will soon happen is that, that last point of interaction is going to be owned by whatever AI platform the consumer uses to monitor their consumption and to reorder products that need to be reordered.
Retailers have traditionally relied on a model where consumers will make a visit once or twice a week to the grocery store within three to five miles of where they live or work. They will walk around the store looking for brands and products. They will be subject to influence of coupons and buying the promotions within the store. Then they’ll go through the checkout process, load those groceries into their car, drive home, and then sort those groceries in their pantry and fridge.
That whole model can be called into question because retailers are imposing huge amounts of cost of interaction on the consumer by expecting them to do all of these things.
The AI assistant automates a lot of this. It turns it into a subscription model. The AI monitors what the consumer is buying and consuming and is able to reorder regularly consumed products automatically. That will change just about everything for retailers: from their business model to the kinds of products that they sell, to a greater emphasis on experiential products and so on.
EH: How will retailers’ business models change, especially from the marketing perspective, with the rise of AI assistants?
ND: The business model is ripe for change with the advent of AI assistants acting on behalf of consumers. The first area where we are already noticing the impact is on products that are regularly consumed. Everything from laundry detergent to shaving blades and so on. There’s about 300 to 500 products that every household consumes on a regular basis and if these can be automated, then what happens to the physical retailers that need to rely on these products as the foundation of what they sell?
To attract the consumer into the store, they carry Tide laundry detergents and Coca-Cola but if those things can be delivered automatically, then the AI based retailers take a big chunk of the business that traditional retailers have been relying on. This is coming. What that means for the business model of retailers is that they have to rethink where they get their traffic and where they get their margins.
EH: Will retailers shift to a more experiential shopping model?
ND: Absolutely. Retailers will have to rethink what kind of products drive meaningful experiences for consumers and those will be the kinds of products that physical retailers will continue to emphasize. Right now, they do bring in customers who are looking to touch and feel the tomatoes or they want to smell the peaches before they buy them.
The experiential aspect of retail is important for certain products. Not all products. And if we start to separate those kinds of products that require an in-store experience in order to purchase versus those that don’t, those that can be easily sold online, that separation will become much more stark.
Retailers will have to realize that the traffic that they used to generate on the basis of Tide laundry detergent and Coca-Cola may be gone within a few years. The reason for consumers to come into the store may evaporate with the advent of AI platforms.
EH: And how does the model for online retailers change given these AI assistants?
ND: First, there will be a much greater emphasis on acquiring and holding onto customers. We’re seeing this in things like the Amazon Prime program, which recently announced that they have over 100 million subscribers. The reason that’s important is because if you can acquire the customer once and sell them a wide variety of products, you’ve only incurred the cost of acquisition once and you’ve built the trust with them, you’ve developed a relationship with them, and you can now sell them a wide range of products.
That’s called economies of scope, where once you’ve built a relationship with the consumer, you’re amortizing the cost of building that relationship over a range of products as opposed to selling one product more often. That’s the critical implication for online retailers is acquire your customers early and hold on to them. Sell them a wide range of products.
EH: You wrote in your article for Harvard Business Review that push marketing, getting platforms to carry a product, will become more important while pull marketing, persuading consumers to seek products, will become less. Can you explain that a little more and how you believe retailer strategies will change in light of this?
ND: That’s a critical question, and I think it’s important for retailers, in particular, to realize that once the AI becomes the consumers’ preferred interface with their world, that’s the interface they’re going to use to order products. The AI platform will operate automatically to order a wide range of products. So marketers now need to sell not just to the end consumer, but they need to sell to the platform. They need to sell to the AI.
For a brand such as Tide or Pantene shampoo, this would be very similar to the challenge they faced in the last 20 years, which is that retailers are concentrating. Brands have to work very closely with the retailers in order to get on to the shelf and manage their space on the shelf so that it is profitable both for the retailer and the brand.
That challenge will be transposed into the online space, into the AI space, where the AI becomes the retailer. The AI platform operates as a retailer and can list or de-list brands and products depending on how they’re performing. It becomes critical to understand the algorithms by which the AI operates and makes those decisions.
EH: In this future where AI assistants own the last interaction with the consumer, how will retailers be able to build brand loyalty?
ND: I think AI platforms do, at some level, pose an existential threat to the brands as we know them. If the AI platform becomes the trusted source of information and advice about products, the question becomes why would a consumer pay a premium to purchase a branded good? The reason they were paying that premium in the past and are paying it now is because the brand offers a promise of superior quality or a promise of a better life, an experiential promise.
What if the AI platform can offer a promise that consumers will always get the best product and the best prices if they rely on it? And then the AI platform switches brands based on how good a deal it can get the consumer and the fact that many other brands meet the minimum standards required. When that happens, the AI platform starts to become the trusted source of advice, brand switching may increase as a result and brand loyalty, therefore, will decline. That’s a real possibility.
EH: Today, shopping is a relatively small part of what how consumers use AI voice assistants. Mostly they use them to listen to music, set timers, or check the weather. What do you think will change in the near future that will actually increase shopping on AI assistants?
ND: What I think will change is the, first, the availability of products and services through the AI assistants, so travel services, banking, stock market, investments. We are already seeing Uber is on there. The range of products and services available will move consumers to use the AI assistants as their primary interface to the world.
When that starts to happen, that reinforces the built-in advantage that AI assistants have and it increases their hold on consumer behavior. The range of product availability plays right into the hands of AI assistant, and as that increases, consumers will rely more on these AI assistants to manage their lives.
EH: How quickly do you anticipate consumers will adopt AI assistant technology?
ND: Technologies today disseminate very, very rapidly. So smartphones reached a billion consumers over 10 years. I think AI assistants will beat that record.
EH: Are there certain types of retailers or certain types of products that will have a larger play in these types of technologies?
ND: Yeah. I think we’re seeing the first application is to products that are low involvement, reasonably low priced, regularly consumed, and just a chore to purchase. We’re least likely to see the AI assistants operate where shopping is an experiential necessity. In other words, you have to go and touch and feel the product. You have to try it on or whatever before you purchase it, and where shopping is fun.
If the consumer considers shopping fun, they are not going to delegate it to an AI assistant. That’s an opportunity for retailers to counteract the force of the AI assistants. Brick and mortar retailers need to make shopping increasingly fun and to increase the experiential aspect of the shopping in order to attract consumers to the stores.
There are some products that you cannot make more experiential and more fun. It’s difficult to make shopping for laundry detergent fun. There’s going to be a very large number of products, 300 or 500 products, where retailers may need to concede to automation. Consumers are not going to come into the store to purchase those items, and that brings us back to the idea that they need to rethink their business model.
EH: You wrote in your article that discounts and price will be a big way to differentiate products on AI assistants and that certainly will be a challenge for smaller retailers who don’t necessarily have the resources to discount as steeply as an Amazon or a Wal-Mart. So in this future how can a smaller retailer succeed?
ND: That’s a good question. I think as price becomes a primary driver on the AI platforms, all retailers will react by making themselves as different as possible and appealing to a more focused market segment. And I think that’s going to happen as a result of an emphasis on price. Price competition will drive differentiation.
Small and medium retailers understand that already. They understand that because they don’t have the economies of scale of the larger retailers so they understand the importance of differentiation. They understand it more than large retailers that have operated on the basis of price, and I think that will carry over into the world of AI assistants.
EH: So, for example, a smaller retailer who’s focused on just hiking equipment will focus on how to market that and differentiate that in a much better ways than a Wal-Mart or Amazon.
ND: That’s exactly right. They will be able to offer a level of knowledge, a level of service to the end consumer that the larger retailers can’t because the larger retailers’ attention is spread across too many product categories.
EH: So for those smaller mid-sized retailers, would AI assistants be the best channel for them? Would they be better served on a channel where they can market their brand more?
ND: Yes, exactly. It might be a proprietary channel. For example, consumers who are highly involved in hiking may be willing to take the time to exit the general AI platform and go to a more specialized platforms in order to find the right product for them.
EH: What are the biggest opportunities that lie ahead for retailers with AI assistants?
ND: I think the message for the retailers is that AI assistants will become the primary interface by which consumers interact with their world. That is a huge challenge for traditional retailers but there are also tremendous opportunities, and I think that’s where they need to focus. They need to understand the opportunities. The AI world is still in its infancy, and as it matures, there will be tremendous opportunities for all types of retailers.
Retailers should focus on service. They should focus on things that make the physical world different.
Let me give you an example from Amazon’s history. In the early 2000s, many traditional, small bookstores disappeared because of Amazon. Once consumers had the convenience of buying from Amazon, it was no longer viable for the Mom and Pop bookstores to continue to serve those customers who valued convenience. Amazon had the economies of scale. It got the customers. It got loyalty. It got all sorts of things because of their size.
Traditional, smaller retailers did go out of business but those that reacted, reacted in a very interesting way. They reacted by making the real world much more important in the world of books for consumers.
That meant having leather couches in the stores where patrons could come in and sit and read the first chapter before buying the book. They added coffee shops within bookstores, where the experience of a coffee and a book together was more than the experience of either the coffee or the book alone. They had reading clubs, where readers could come interact and spend an evening talking about the books that they had read. They had author signings and meet-the-author evenings.
Those were not things that could be done online at that time, and so they survived by reacting to the online world, by making the physical world much more real. And I think that’s about to happen across a wide range of product categories, not just books. This emphasis on experience and service will translate to online retail as well, as AI assistants gain a greater foothold among consumers.