Digital marketers, accustomed to using software that helps them think about marketing, are now transitioning to a time when software will do much of the thinking.
It’s called cognitive marketing, and research firm IDC expects that half of all companies will use this emerging generation of computer intelligence for their marketing and sales efforts by 2020.
IDC Research Manager Gerry Murray (who is speaking at our MarTech conference next week) predicts that just a year from now, the current group of several dozen applications employing cognitive computing will become “hundreds.”
Such applications, he told me, utilize “processes akin to the human brain, [by] taking signals and drawing conclusions.”
The converging drivers — massive computing power, massive amounts of data and a booming population of connected devices and sensors — are of course instrumental in delivering this new kind of marketing. Equally important is the ability to deliver cloud-based services, as most marketing services do these days — except that the cloud occupant here may well be IBM’s Watson supercomputer, or an equivalent, communicating through an API to the operational platform, also in the cloud.
Murray noted that we’re already seeing the advance guard of self-learning intelligence, in the form of chat bots, automated segmentation of audiences, image analysis, natural language recognition and automatically delivered personalization of content.
It’s more than the seemingly omnipresent predictive analytics, he said, which delivers scores for whether leads are likely to turn into customers or makes other predictions, such as the best product to offer a prospect in order to arrive at a sale.
In many kinds of predictive analytics, the prediction is accomplished by seeing what worked in the past — which customer profiles resulted in the most sales, for instance — and then matching those patterns against new prospects. For those applications, Murray said, “the swim lane is very narrow [and] they become obsolete over time as things change.”
By contrast, he noted, cognitive marketing most commonly utilizes a neural net, employing “a platform with a very different model [from] predictive analytics.” While cognitive-based machine learning often involves prediction, just as a human makes dozens of predictions as part of daily existence, Murray pointed out that an overall cognitive marketing system has a much higher level of independent decision-making than do many current predictive applications.
Moving up the food chain
With cognitive-based marketing tools, Murray said, “marketers will spend less and less of their time clicking through user interfaces,” a present-day situation where rules-based process-flows often involve dozens of manual selections, just to get, say, basic audience segmentation set up.
“Marketers need to move up the food chain of decision-making,” he said.
A cognitive marketing platform can do operational decision-making, such as setting segmentation rules that are customized for each visitor, and not just for groups of visitors, while the marketer can focus on strategy, platform oversight, making recommendations and the use of what Murray describes as “ethics and governance.”
And, he suggested, cognitive marketing may even redefine how brands relate to customers.
“Philosophically, what do we do as marketers?” he asked. “Do we just sell stuff, or can we improve people’s lives?”
Cognitive processes might lead to a mass delivery of hyper-personalized content and services that are light years beyond what is possible today.
Or brands might differentiate themselves by the quality of conversations conducted by their intelligent agents on, say, messaging apps or cognitive TV. Today, brands talk about their personality, which is mostly shorthand for their graphics and the public’s impression of them. But imagine their personality when you’re able to have an actual conversation with a cognitive agent designed for a specific personality.
Companies will be able to go beyond being “just a utility company, just a cosmetics company,” Murray said, so that they can have personalized relationships with their customers. Essentially, they become familiar strangers, like the friendly neighborhood barber who is more than a barber.
It would certainly constitute an appropriate full circle if these enormous shifts in the availability of massive computer intelligence and data lead us to a marketing environment where distant brands become more like your corner store.
Comments