After Jeopardy: IBM’s Watson and the Challenge of Advertising


watson_smallADOTAS – Great advertising campaigns are often portrayed in popular culture as the work of lone geniuses, scratching out immortal copy on cocktail napkins in a flash of inspiration.

Those brilliant admen and women of course exist. Big ideas are important. But the most effective advertising has always been built on a foundation of customer knowledge and insight — which is as much science as art. An advertiser with the most creative ads still needs to understand how their campaigns drive sales — especially whether the ads are stimulating purchases that would not otherwise have occurred.

They need to know what brought customers through the door. Was it the television commercials? The radio spots? The circular in the Sunday paper? The direct-mail pieces? The online ads? The mobile ads?

This knowledge is especially important in today’s world of constrained marketing budgets – when advertisers just can’t afford inefficiencies.

IBM has spent many years working on problems of this sort, creating analytics systems that help marketers make better decisions about where to spend their advertising dollars as well as inform creative teams about which customers are most likely to be receptive to their messages and what makes those customers tick.

That’s why I was so captivated recently by the demonstration of the Watson computing system on the Jeopardy! quiz show. In a two-game match, Watson defeated the two all-time Jeopardy! champions.

This is about a lot more than fun and games. Watson represents a tremendous breakthrough in the ability of computers to combine an understanding of natural language with a unique ability to reason. It can evaluate the equivalent of hundreds of millions of pages of material — books, reports, articles and so on — in three seconds or less. It is not stymied by intricate wordplay.

Watson technology could represent a big leap forward in customer insight. It should be interconnected to a retailer’s core systems, including databases, CRM, inventory and order management plus all points of interaction with customers.

The technology’s analytical capability could enable businesses to ask a virtually unlimited series of questions — instantaneously — about individual customers, such as: “What is the next contact we should make with John Smith? When should we reach out to him? What should we say? And through what medium?”

Watson could easily and immediately analyze all the information the retailer has on that customer — purchases (and returns), brands that John usually buys, what he has been browsing online (if he identifies himself), the complaints he has made and the type of customer he is (for example, a loyal customer … or only shops sales … or prefers shopping the Web rather than the store).

It could integrate that data with massive amounts of product data, the retailer’s strategy for that customer, and relevant solutions to come up with the right answer: Tell John about the availability of a new product with a specific warranty plan and free installation. Give him this message in the evening as a notification on his iPad app, and offer pre-purchase with free delivery.

Watson is particularly adept at consuming new information, in real-time, and interpreting its meaning. Historical data is very helpful in understanding what’s important to customers — their motivations, triggers and needs.

But it isn’t very useful in predicting the mission the customer is on at this moment. There are just too many variables at play. However, with even a small clue, such as a bit of browsing data from the Web site or mobile app, or the customer’s current location in the store, a retailer can make much better guesses.

This is Watson’s forte: understanding the meaning of a clue based on a vast repository of knowledge. For the retailer, Watson technology could act as an aggregator, using historical and real-time data in helping the merchant understand what’s important to customers.

Why is this crucial? Because the relevance of a message, offer, or action a retailer takes with a customer is directly proportional to how well the retailer understands what the customer wants or needs right now. This capability is what it will take to transform the billions of emails, text messages, social media interactions and digital coupons retailers are sending customers from spam to something useful and relevant — something worth reading, responding to and acting on.


Please enter your comment!
Please enter your name here