Customers have a new way to experience your company that doesn’t involve you. Will you market to their AI surrogates?
But most of all, it will depend on your ability to manage the impact of artificial intelligence (AI).
Major technology companies are investing heavily in AI assistants to act as intermediaries between companies and customers, creating a shopping environment where humans can offload vast swaths of the customer experience to machine-learning algorithms.
As AI matures, organizations that can’t compete with the tech giants to hire data scientists—which is to say, most of them—will face a life-or-death struggle to develop strategies for defending their market share, maintaining customer relationships, and retaining relevance.
AI is already capable of automating routine retail purchases, but it won’t be long before it can take over far more complex transactions. Customers will be able to hand over any purchasing decisions they don’t find particularly interesting or engaging (though not necessarily unimportant) to an AI that has learned to calculate their preferred combination of features, price, and ethical considerations. Only companies that can deliver customer experiences with personal meaning will continue their direct relationships with buyers. The rest will be nudged into the background, where the “customers” will actually be emotionless, thoroughly informed AIs.
Almost every company will have to pay constant attention to how AI assistants think, learn, and evolve and then adjust their marketing strategies accordingly. To avoid being backgrounded to AI-based proxies for human customers, companies whose products and services lack inherent emotional resonance will have to find ways to create it. And where that’s not possible, companies will need marketing strategies that persuade gatekeeper AIs to choose them.
The complexities of modern life have driven our brains into cognitive overload. To cope, customers are pulling AI engines close, relying on their recommendations while pushing traditional brand interactions into the background.
We happily accept the shows Netflix suggests, the music Spotify plays, the cleaning regime Roomba chooses, and the results Google displays. We have come to trust these machine-learning systems so much that 35% of Amazon purchases and 75% of Netflix selections are driven by machine-learning recommendations.
So it’s not a huge leap to imagine that many customers will welcome the opportunity to tell an AI assistant that they need socks, knowing they can trust the algorithm to parse that as “buy me six pairs of black organic cotton crew socks for less than $15 a pair from the vendor offering the best deal.”
But it creates a challenge for the hosiery manufacturer: how do you make sure your socks are the ones the AI picks? Or, alternatively, how do you make buying your socks such an enjoyable experience that your customers would rather do it themselves than offload it to the AI?
To understand how to tackle this challenge, it’s first necessary to understand the difference between buying decisions customers don’t consider worth significant attention and those that stand out. The first group can be categorized as candidates for “background shopping”: rote purchases of products the customer doesn’t want to spend a lot of time on and can offload to an AI that compares products and makes the decisions.
For background shopping, the AI will begin to winnow the available options based on parameters the customer sets, such as preferred brands, product features, or price points, as well as emerging values-based considerations, such as transparency, ethical sourcing, fair pay and benefits, and gender and minority equality. Then it will rank the results by its own criteria, generated by tracking past purchases and gathering comparative information that humans can’t—or don’t bother to—assemble.
Over time, the AI will learn and adjust both individual and group preferences by combining the makeup, habits, and buying patterns of a household (or a company, in the case of B2B purchases) with changing dynamics in the market, such as price changes and new product introductions.
For example, a family with two college-aged kids may double its coffee consumption when the kids come home for visits. The family’s AI will track the calendar and add more coffee to the week’s grocery order in preparation for school breaks. It may look for a better price on that coffee at other stores or add a different brand of coffee that the kids prefer. It may also discover a new brand it thinks the whole family will like better. Or it may even invent a custom blend at a local coffee roaster and order a batch for everyone to try.
Background shopping will, in short, be so convenient, sophisticated, and attuned to our lives that we’ll barely have to think about it. Products and services will arrive almost as soon as we know we need them—and maybe even before.
Companies that want to avoid the background will have to make an even greater effort to communicate directly with end customers and place (or keep) themselves front of mind.
“Basically, brand managers have to determine what makes certain consumers or market segments think about shopping for their products or services as a meaningful experience instead of a chore and then figure out how to make more consumers think the same way,” says Niraj Dawar, professor of marketing at the Ivey Business School at Western University, London, Ontario, Canada, and author of Tilt: Shifting your Strategy from Products to Customers.
There are four broad approaches to doing this:
Brands, retailers, and entertainment are already beginning to blur into each other. Just look at any blockbuster movie with its spinoff merchandise, such as toys, clothing, and video games, and character appearances on screens and at live events.
Anyone who enjoyed the immersive Star Wars: Secrets of the Empire VR experience offered in shopping malls around the world had a meaningful interaction with the Star Wars brand—and with the technology brands that enabled it.
Big-box retailers nearly eliminated the experience of connecting with neighbors and catching up on local news in local shops. But now, by creating engaging peer-to-peer experiences, companies are reviving that experience, generating emotional value for customers in ways that AI can’t replicate.
For example, boxed meal services like HelloFresh and Blue Apron let people go online to order complete meals with cooking instructions and all the ingredients, premeasured and precut. These companies push the most intimidating aspects of the home cooking experience to the background, while allowing families and friends to connect by cooking together, enjoying the resulting meal, and creating an experience worth talking about. Services that send regular curated deliveries of clothing, wine, or toiletries follow the same model: backgrounding the boring, foregrounding the fulfilling.
It’s hard to make a purchasing decision when you feel under- or uninformed. Companies can turn that uncertainty into meaningful shopping by helping customers understand how their products work and teaching them how to get the most from their purchases, whether they’re housewares, power tools, or luxury automobiles.
They could do this with curation, online and in-store classes, and gamification. A mobile phone manufacturer competing on the quality of its cameras can go beyond offering in-store and online photography classes to allowing students to submit their photos to a digital art gallery where the public votes on the best images. The manufacturer might even send the photos with the most votes to experts, who then curate the best of the best for inclusion in the company’s advertising.
Another possibility is to embed technology into the product so that using it becomes an additional opportunity for the company to deliver information. For example, a golf club could include Internet of Things sensors that monitor the user’s swing and provide feedback to a smartwatch in real time, while a winemaker could offer a wine club experience that pairs a monthly wine selection with recorded or live webinars with wine experts.
The craving to improve ourselves and our surroundings is a core human trait. Consider the person who chooses a firm, locally grown tomato at the farmer’s market over something mealy grown far away from the supermarket.
Companies can also tap into that desire by offering, sponsoring, or partnering in tools and communities where customers can share goals and monitor progress for both personal aims, such as weight loss, and societal needs, such as charitable fundraising.
Influence marketing is a variation of this approach. Dozens of companies are already hiring celebrities to post pictures of themselves enjoying a product on social media, resulting in increased sales to people who aspire to a similar lifestyle.
The more of these avenues companies can use to deliver meaningful experiences to customers, the better their odds of remaining relevant in the age of background shopping.
The ideal solution to avoiding backgrounding is to create a shopping experience so phenomenal that customers actively seek it out. Audi, BMW, and Mercedes stand apart from the Japanese brands that offer lower prices and better repair records by giving American customers a unique experience. They can pick up their new cars at architecturally breathtaking delivery centers in Germany. There, they first enjoy a hands-on demo and a meal with automotive experts and visit the neighboring museum of classic cars. Once they pick up their new wheels, they can embark on a European road trip that includes recommendations for hotels (Mercedes even throws in a free night’s stay) and preplanned routes that show off how enjoyable the luxury vehicles are to drive. At the end of the trip, the cars are shipped home at a lower sticker price than the buyers would have paid in the United States.
Realistically speaking, however, some products simply don’t have much potential for emotional resonance. These products will inevitably be backgrounded, if they aren’t already. In these cases, companies need strategies for marketing to the AI assistants, largely by emphasizing features the human customer might want.
To catch an AI assistant’s attention, brand managers will have to expand their definition of “value” to include any possible relevant characteristic. Even the humble roll of toilet paper can stand apart from the pack in the right context: made from recycled materials or with eco-friendly processes for resolutely green shoppers, unscented for shoppers with sensitive skin, or in industrial-sized packs for those buying for a big crowd.
Smart speakers and other AI-assisted technologies are being adopted at such a blazing pace that Dawar predicts that every household will have some kind of AI handling many tasks as soon as 2025. Indeed, he believes AI will become the dominant shopping interface, with customers automating all but the most meaningful tasks.
That means companies with good reason to expect they’ll be backgrounded will need to start strategizing right away about how to stay in front of the algorithms.
In the brick-and-mortar world, retailers practice trade promotion management to make sure their products catch shoppers’ eyes. Manufacturers pay retailers for the best spots on the best shelves in the store. And when a manufacturer wants to increase sales volume with a promotion, it drops prices, which gives the retailer a financial incentive to put more product on the shelves for the length of the sale.
In the era of AI-enabled background shopping, we will need a digital version of trade promotion management that creates new customer channels driven by product placement within the AI platform.
Dawar suggests that to become solid background brands, companies will have to master the equivalent of premium shelf placement for AI. Indeed, he predicted in the May 2018 issue of the Harvard Business Review that AI-assisted shopping will induce companies to make significant cuts in their current spending on advertising, listing and shelf placement fees, and commissions for brick-and-mortar retail. He foresees them redirecting that budget toward offers and innovation strategies designed to get AI shopping assistants to suggest or even automatically offer their products.
How will that work?
For one thing, Dawar says, companies will have to figure out how the AI decides what to show customers and how to appear in as many search results for the target market as possible, at the lowest possible cost. This will be an ongoing process of observing an AI assistant’s choices and reverse-engineering them to determine what criteria it uses to select the products it buys.
While this collected data will help companies understand their markets and identify opportunities to address a particular individual, group, or segment more precisely, it will also require them to optimize continually for each of those individuals, groups, or segments.
When purchasing decisions are being made by AIs rather than customers, brand managers will have to include AI providers in their strategies for getting products offered and featured.
A manufacturer will likely make deals directly with AI providers that are themselves retailers, such as Amazon or Google. However, the growth of the Internet of Things and smart appliances suggests that we can also expect an explosive increase in third-party AI providers.
By interfacing with multiple retailers, these third-party AIs could generate reverse auctions to ensure an optimal combination of products, services, and price—as, for example, a smart refrigerator’s AI requesting bids from several supermarkets’ own AIs for a total shopping list or even individual items.
The combination of multiple vendor and third-party AIs potentially creates millions of sales channels. This could theoretically destroy manufacturers’ and retailers’ control over pricing, forcing them to track hundreds of AIs in real time to see how prices are changing and how that affects their ability to fulfill orders.
It’s also likely that some AI providers will create special offers to increase adoption of their own AI assistants and make themselves customers’ preferred data collection, communication, and shopping channel.
For example, providers could offer discounted or free smart devices for accessing a specific AI platform. Or they could offer discounts on specific items, house brands, or entire orders to people who shop using that platform. As this occurs, companies will have to determine how to win the provider’s favor and be included in those special offers—or to break through the noise and compete with those offers.
When AI is making more of the decisions about what to buy and when, companies will also have to draft their own AIs into the fight for relevance.
Some are already doing so.
Leading retailer Costco has always relied on providing its members with a treasure-hunt atmosphere where shoppers discover additional items they didn’t know were available but must have. It is now testing AI in its bakery to predict demand across multiple variables and change production runs accordingly. During a sports event, for example, the AI will suggest making more of items that sold well during similar events to ensure that people looking for game-day snacks don’t walk out empty-handed.
Similarly, Chinese fast-fashion shoe brand Aimiqi has developed an AI application for intelligent product design that keeps track of hot colors, top-selling fashions, and design trends in real time. The AI combines the variables into new shoe concepts faster than human designers can think them up; Aimiqi then manufactures the shoes on a tight turnaround so it can hit shoppers’ “gotta have it” emotional button before anyone else can.
However, one critical fact remains: brand managers lack the skills to pry open the black box of AI shopping algorithms to see how they make decisions, and while the number of data scientists worldwide may be growing, it’s still small. Most companies simply can’t compete with the tech giants for data science skills. That means developing both in-house AIs and strategies for interacting with third-party AIs will remain a challenge for some time to come.
“Brand managers are struggling with this, and not just for consumer goods but for banks, insurance companies, and any other products and services that could be automated or the decision-making around buying them augmented using AI,” Dawar says.
In other words, the race is on.
The post-algorithmic customer experience is already developing. Companies have only a brief window of opportunity to develop strategies to optimize for or avoid AI assistants on an ongoing basis. They must find a way to remain meaningful, or at least to compete on features and price. Otherwise, their choice about maintaining relevance or being shoved into the background will be made for them. D!
SAP worked with more than a dozen industry experts to uncover five trends that will determine the customer experience over the next decade. The Future Customer Experience: 5 Essential Trends report, examines each of these trends and offers recommendations for how brands should respond now to prepare.
Read more thought-provoking articles in the latest issue of Digitalist Magazine, Executive Quarterly.