How AI Is Making Its Way Into Marketing

1

Artificial intelligence is firmly in the spotlight this year as global tech firms continue to invest and set out their AI ambitions. As the technology takes hold, how is it making its presence felt in marketing?

No longer the preserve of science fiction, artificial intelligence (AI) is becoming technology fact. So far in 2016, AI technology has grabbed headlines as the focus of Apple’s first acquisition—in the form of Emotient Inc—and Facebook CEO Mark Zuckerberg has resolved to build an AI assistant to run his home and help him at work.

Google has also gone down in the history books after its DeepMind team developed an AI programme capable of defeating human world champions of complex Chinese board game Go. It’s an achievement reminiscent of IBM’s milestone moment when its cognitive system IBM Watson thrashed human contestants in the U.S. game show Jeopardy in 2011.

So as AI finds backing from the world’s large tech firms, what are they buying into? Data scientists use a range of terms to describe AI and its related technologies, such as machine learning, deep learning, and cognitive computing.

What they have in common is the move away from rules-based programmable systems towards those requiring human-like intelligence, whether that’s visual perception, speech recognition, or decision-making. Specifically, they are systems that learn from the knowledge they acquire and are able to infer meaning and act upon it.

While there is a long journey ahead for AI and all it could achieve, the technology is gaining an early foothold in marketing. From market research through to personalisation, AI is becoming pervasive as marketers look for ways to improve data insights and scale responses without unnecessary costs.

Market Research Accuracy

By combining the sciences of deep learning, computer vision and cognitive neural science, U.S.-based Emotient has developed software that measures micro facial expressions such as joy, anger, and surprise to infer real emotional responses, attention, and engagement.
Advertisers are using the new technology to gauge reactions to new creative and replace traditional market research techniques.

“Emotions drive spending, but we’ve always struggled to understand how people really feel,” said Emotient CEO Ken Denham. “Billions of dollars are spent every year acquiring customers and trying to understand them and what they feel about products and the customer experience. The reality is we’ve been guessing.”

When deciding which of three new fragrances of laundry detergent to launch, one FMCG company used Emotient’s AI software to measure participants’ facial expressions when told they could take home one of the fragrances. The technology accurately predicted the preferred fragrance, which contradicted the results of focus groups and surveys the participants had done. “It’s very difficult for people to tell you what they really feel,” said Denham.

The BBC has used AI as a new form of market research to drive B2B sales conversations. New Zealand-based AI firm Parrot Analytics worked with the British broadcaster’s BBC Worldwide division to determine demand for some of its programming internationally. Viewer sentiment for certain shows was measured based on the nature of activity across social media sites and blogs, leading the BBC Worldwide team into new dialogue with potential customers.

Natural language analysis also underpinned carmaker Kia’s use of IBM Watson to guide its decisions on which social media influencers to work with for the recent Super Bowl.

Real-Time Automation

The rise of programmatic advertising is creating a greater need among advertisers and marketers for AI-led automation tools.

By 2020, more than half of all digital video advertising revenue in Europe will be programmatic, according to a study by analytics firm IHS and video inventory management platform SpotX. In the U.K., programmatic’s share of digital video revenues is set to increase from just over 32% in 2016 to more than 60% over the next four years.

As such, AI and its technologies are increasingly becoming part of the day-to-day tools for marketers, with mobile video platforms such as LoopMe using AI to optimise ad placements across devices and by viewers.

When integrated with programmatic ad exchanges and direct publishers, this is giving marketers new opportunities for real-time campaign optimisation.

Personalisation At Scale

AI is increasingly being deployed by marketers to improve and scale personalised customer interactions, making use of the improving technologies around visual perception and natural language processing.

Online retailer Shoes.com is using Sentient Technologies’ AI to enable product discovery for its footwear customers on Shoeme.ca. The merchandising application was launched in November, whereby customers browsing a range of boots can click on various product features they like to see more personalised recommendations, powered by Sentient Aware, the company’s visual filter tool.
The technology is providing detailed insights for use across the business. “By engaging around product, it allows them to understand the customer’s intent and individual product preferences,” said Nigel Duffy, Sentient CTO.

This helps the business move beyond what Duffy calls “groupalisation”, when a customer receives a personalised offer based on the average preferences of a group rather than one truly based on his or her own preference.

Conversational AI processes the interactions between customers and company to create a highly engaged customer experience and forms the basis for Digital Genius’s work for Unilever. The company created an SMS-based AI personal recipe assistant called Chef Wendy for its Knorr brand in developing markets where Internet access is limited.

The text-based conversation is enabling Unilever to build up profiles and a detailed database for future promotions, while scaling its brand-to-customer interactions without a necessarily big increase in cost.

Another of two-way AI interaction is outdoor brand and retailer The North Face’s use of the sophisticated IBM Watson technology through its partnership with Fluid. At the tail end of last year, the company beta-tested an interactive online shopping tool that asked shoppers context-seeking questions. The natural language tool helped guide the shopper journey and replicated an in-store experience online.

During the 60-day trial, 50,000 consumers used the service, spending two minutes longer than before on the site. Satisfaction levels were high, with three-quarters of customers saying they would use the tool again.

“I think this tells you that consumers are ready for some type of AI,” said Cal Bouchard, senior e-commerce director for The North Face.

About Author

Angela Rumsey

Angela is a B2B journalist specialising in marketing and retail.

1 Comment

  1. Personalisation at scale is a great benefit from artificial intelligence. There is another area that you don’t explore in great detail in this article – the sheer scale of decision making automation that AI gives us.

    We are creating billions of pieces of new data all the time. Each time we build new ways of analysing that data we are creating new layers of data. The more we go down the road of AI and machine learning, the more we will need AI to help us manage all of the learnings that derive from that technology.

    There’s no way we will be able to keep up with all the insights and decisions that the AI technology will develop.