California Management Review
California Management Review is a premier academic management journal published at UC Berkeley
by Nikolas Jintri
Camilla Olson did not always want to be a fashion designer. In fact, when her daughter wanted to attend fashion design school, Olson was skeptical: “I took her to [a few design school] open houses so she could see why it was not a good thing to do, but [then] I [was] seduced into it.”
Though she learned to love fashion, her experience running her own design label led to what she describes as an “‘Ah ha!’ moment.” As Olson sees it, the fashion industry caters mostly to “an idealized body shape — the hourglass — thus leaving 80 percent of women without clothing that looks good on their body.” To Olson, this is more than an inconvenience; it is an outright injustice.
That is why she started Savitude — a company that uses artificial intelligence (AI) in order to help women find perfect outfits. Savitude’s AI Curator makes shopping easier and more fun by personalizing customer options. This maximizes loyalty while increasing engagement and conversion.
Savitude’s AI Curator helps fashion retailers present each customer with offerings that are most likely to fit her individual shape and size. It does this by guiding the customer through a series of questions, accompanied by visual aids, along the lines of, “Which of these two images most resembles your hips?” and “Which image is closest to your body weight?”
The narrowing process for body shape culminates in a personalized fashion boutique which then allows for further customization according to design preference. For example, a customer can choose to see garments with necklines that are “less like,” “some like,” or “more like” that of the garment they are currently viewing. The same goes for the garment’s other qualities, such as its silhouette, length, sleeve type, and color. Through this process of elimination, Savitude’s AI Curator minimizes effort, maximizes satisfaction, and increases sales.
It is hard to imagine how an unaided human clothier or store clerk could duplicate these results. For that matter, it is hard to imagine how the typical online shopping experience — with its infinitely cascading barrage of information and options — could ever duplicate these results without AI assistance.
That is why Savitude is at the cutting edge of what V. Kumar and his colleagues describe in their research paper entitled “Understanding the Role of Artificial Intelligence in Personalized Engagement Marketing.”
Kumar et al.’s paper describes how digital marketing has moved consumers from a dearth of information to an overabundance of it — and how AI is now reducing information overload by presenting consumers with tailored options.
This hyper-customized approach to marketing depends upon collecting data about individual customers, then using that data to automatically narrow each customer’s experience to a digestible series of options. This will result in highly targeted marketing that delivers the best possible product or service to each consumer, even while minimizing the time, energy, or fortuity required to find that product or service.
As Kumar and his colleagues put it in their paper’s abstract, “[C]onsumers are ready for a new wave of the [customer] journey in which endless options and information are narrowed and curated in a personalized manner….[T]he AI tool can be used towards this end.”
This is a hopeful development for those who love shopping, as well as for those who hate it. It is also a hopeful development for those who have a product or service to sell, and are looking for ways to individuate their marketing cycle in order to convert more customers.
In the context of AI-driven customer engagement, “curation” refers to the process (mentioned above) that moves customers from information overload to a digestible series of choices. Without curation, customers may find themselves either unable to find a good fit for their needs and desires, or — just as bad — struggling in a state of “analysis paralysis” because they are overwhelmed with too many options.
It goes without saying that neither of these situations is good for business.
While AI curation may sound like a magic pill for both customers and providers, making it work is not as simple as pushing a button. AI is a tool that augments human intelligence — not an actual intelligence in its own right — and like any tool, it is only as effective as the human beings who use it.
Kumar and his colleagues outline four requirements that businesses must have in place in order to make the best use of AI-curation:
Data maturity: Kumar et al.’s article says it best: “A well-developed, well-endowed, and well-connected data ecosystem is fundamental to deriving benefits from learning and AI capabilities.”
Specific goals: As with any tool, if you do not aim for a specific outcome, it is impossible to use AI effectively. Once again, AI is only a metaphor for advanced data analysis; it is not true intelligence. As such, it can only do what it is programmed to do in accordance with company goals — and those goals must be specified in advance by human managers.
Customer care parameters: Companies must understand how and when to use AI so that customers can always receive the human help they need if and when the AI tool is not adequate.
Workforce preparation: AI is likely to eliminate some jobs, add others, and transform all. Companies must prepare themselves by creating a workforce that can use and oversee the AI curation process effectively.
Information ethics: AI curation involves collecting enormous amounts of personal data about customers, so it will become increasingly important to use customer data not only in strict accordance with evolving privacy laws — but also in strict accordance with company ethics. This means that companies must clarify how, why, and to what extent they will use customer data.
AI may not be a magical push-button solution that solves every problem in commerce, but it will certainly transform our world. Kumar et al. predict a near-future in which AI-driven branding, pricing, service delivery, and advertising will become as commonplace as billboards and television ads are today.
AI-driven customer engagement will open many new opportunities, but it will also present many new challenges — the most urgent of which will probably revolve around issues of privacy. When so many people are willingly giving away so much personal data, we must wonder how companies (and governments) will use that data.
There is definitely room for concern, but there are even more reasons to be optimistic. After all, artificial intelligence is really nothing other than human intelligence aided by powerful digital tools. We may augment our intelligence with computerized data analysis, but that does not mean the machines are taking over.
The future remains in our hands.
This post is based in part on the forthcoming academic article “Understanding the Role of Artificial Intelligence in Personalized Engagement Marketing.” by V. Kumar, Bharath Rajan, Rajkumar Venkatesan, and Jim Lecinski.