Targeting Marketing and Cognitive Computing

Stephen DeAngelis

August 2, 2016

When it comes to marketing, Delshad Irani and Ravi Balakrishnan decry the fact that overblown claims are being made about artificial intelligence (AI). “In the marketing world,” they write, “it’s caused the customary outbreak of confusion. Thus resulting in a series of obtuse declarations like ‘AI will change everything. EVERYTHING!’ that are typical of the industry.”[1] It’s easy to fall into that trap. I published an article with that very headline; but, I wasn’t addressing the marketing industry; rather, I was discussing how artificial intelligence is becoming so ubiquitous that it touches our lives in ways we don’t even realize.[2] Nevertheless, their point is well made. In spite of their concern about hyperbole, Irani and Balakrishnan admit that artificial intelligence (or cognitive computing) is already helping advance targeted marketing. They rhetorically ask, “What can intelligent, learning, evolving machines do for marketers and their brands right now?” Their answer:

“Provide actionable insights from a massive amount of unstructured data that’s available today. Data that will grow to 44 zettabytes by 2020, ‘an unprecedented online milestone that will occur in our lifetime’ according to Cisco. 44 doesn’t sound like much until one learns: ‘A zettabyte is roughly 1000 exabytes … an exabyte alone has the capacity to hold over 36,000 years’ worth of HD quality video … or stream the entire Netflix catalog more than 3,000 times. A zettabyte is equivalent to about 250 billion DVDs.’ Artificial intelligence, machine-learning, cognitive computing, call it what you will for now, these systems consume deluges of data, learn with every fraction of an interaction, and answer questions in natural language and with real-time relevancy.”

Taylor Stockwell (@stockwet), an enterprise digital marketing leader at IBM, argues that marketers are always looking for the next big thing. “The next new, cool thing comes along,” he writes, “and we pounce all over it, right? Why do we do this? It’s simple — we have big problems to solve and little time or resources to do it. If there’s a technology that can do it for us, then we’re sucked in. Cognitive is on its way to being that next shiny object.”[3] Stockwell goes on to provide his definition of cognitive computing:

“Cognitive systems, as we’re using the term here, are technologies that use natural language processing and machine learning to learn, reason and understand natural language. While some systems can understand natural language, cognitive systems are able to learn and reason, moving them away from the deterministic computing systems we’re used to when we fire up our iPads or laptops. Cognitive computing promises much to the world.”

I define cognitive computing as a combination of semantic intelligence (i.e., artificial intelligence (including machine learning) and natural language processing) and computational intelligence (i.e., advanced mathematics); and, I agree with Stockwell that cognitive computing promises much to the world — including the marketing world. Like Irani and Balakrishnan, Stockwell believes that cognitive computing is going to have a disruptive influence in the marketing arena. He states, “There are a number of disruptive technologies either here or coming soon (drones, IoT platforms, blockchain) and cognitive is right up there with them.” He quotes Harvard Business School professor Clayton Christensen who notes, “Disruption displaces an existing market, industry, or technology and produces something new and more efficient and worthwhile. It is at once destructive and creative.” Stockwell adds:

“Disruptive innovation destroys inefficiencies in order to create new, more efficient opportunities. Applying this to cognitive — where you transfer the embodiment of human touch points to a system — you can see numerous opportunities to both destroy and create — whether that’s happening at the person, process or product level. We’re going to see new business models created through cognitive systems, new processes that make transactions more efficient and products that alter how we think and interact with technology.”

He believes cognitive computing will be a boon for targeted marketing because it can deal with more data than past systems. To underscore his point, Stockwell cites a Gartner study that concluded, “75% of marketers are struggling with the 20% of the data they can see, never mind the dark data they can’t see.” He goes on to explain the difference between light and dark data:

“We often refer to data as structured vs unstructured. That’s an accurate view of how data can be categorized. It’s a worthy framework. But, analysts are talking about a new way to look at data — light data vs dark data. This is a more meaningful view of data because you can immediately assess your company’s data gaps. Light data refers to that data you know about, have access to, and can use for insights. It can include any form of data, whether structured or unstructured. Dark data refers to data that, well, you don’t know about. Or, maybe you know it’s there, you just can’t access it or do anything with it. It is a latent asset — something with tremendous value if you can tap into it.”

Stockwell asserts, “Cognitive shines in the dark.” It’s an assertion with which Irani and Balakrishnan would probably agree; nevertheless, they caution, “The biggest challenge to the marketing industry embracing cognitive computing systems is to resist the temptation to use AI for AI’s sake.” In other words, marketers need to know what cognitive computing capabilities exist and understand how those capabilities can be applied to the marketing challenges they face. The world is going digital and that means more and more data is going to be generated. Chuck Martin puts it this way: “The Internet of Things is going to expand the concept of target marketing. By a lot.”[4] He explains, “The promise of one-to-one arrived in various forms over the years. Mobile certainly pushed that along. But much of the targeting has been based on normal and rational things, such as where a person is, where they have recently been and maybe a little of what they purchased along the way. IoT technology here and on the horizon will extend that well beyond location data.” The right kind of analysis can turn that data into knowledge and insights. Sandeep Sehrawat, a Digital Marketer at Survtapp, asserts, “A marketing campaign without analyzing and targeting audience is a big waste. Digital marketing will help you to reach targeted audience. It helps both global and local businesses in lead generation. Search engine marketing, search engine optimization, banner promotion, video marketing, etc. are the effective digital marketing campaigns that help business to reach their targeted audience. You can reach to the most specific audience based on age, group, activity, interest, location, etc.”[5] As Martin indicated, Sehrawat only scratches the surface when it comes to the types of data that can be used to pinpoint the right audience for a particular product. The point is, cognitive computing systems can handle the oceans of data being generated and analyze it for actionable insights that can benefit any marketing campaign.

 

With so much media vying for people’s attention, getting the right message to the right person at the right time is becoming more important than ever — even as it is becoming more difficult. Cognitive computing can help.

 

Footnotes
[1] Delshad Irani and Ravi Balakrishnan, “How Artificial Intelligence can help marketers sell better and more,” ET Brand Equity, 6 July 2016.
[2] Stephen DeAngelis, “Artificial Intelligence will Change Everything,” Enterra Insights, 11 July 2016.
[3] Taylor Stockwell, “Cognitive Computing Is Marketing’s Shiny New Object,” Business to Community, 6 July 2016.
[4] Chuck Martin, “IoT Target Marketing: Detecting Sitting, Standing, Walking, Driving And Current Mood,” Media Post, 28 June 2016.
[5] Sandeep Sehrawat, “Why Businesses Should Focus on Digital Marketing?Customer Think, 29 June 2016.