The rise of the artificial intelligence in marketing

The rise of the machines

Anna started working for IKEA in March 2005. As is common with young and inexperienced recruits, her early days there were tricky – some customers called her ‘retarded’. But Anna stayed cheerful, helping visitors find their way around the huge list of IKEA products. She had memorised more than 120,000 of them, from beds to kitchen utensils. Because she was good at languages, Anna worked in a number of countries – the Czech Republic, Sweden, France, the US and the UK. Even when half of the questions she was asked were ‘sex-related’, she would respond, helpful as ever. But eventually it got too much for Anna. She retired in 2015, having clocked just over ten years of service. IKEA’s decision to take Anna offline (or Ask Anna, as she was also known, the virtual assistant on the company’s website) is surprising. It goes against a huge surge in chatbot deployment and consumer adoption. Facebook Messenger announced it had 34,000 bots in November 2016, barely six months after launch. As Forbes revealed last autumn, while many businesses are enthusiastic about chatbots, customers cannot always recognise when they are talking to a robot instead of a human. With part of the customer experience entrusted to machines, this can lead to confusion and, potentially, brand damage.

Technological turning point
Interacting online with robots is only the start. With rapid technology advances, consumers are increasingly engaging with machines in the real world. This is one of the symptoms of what economist Carlota Perez calls a ‘technological turning point’. Unsurprisingly, Japan, with its admiration for technology, has been a pioneer. Japanese telecoms group SoftBank launched The rise of the Artificial intelligence in marketing Pepper, the first robot capable of reading and responding to human emotions, in June 2015 in Japan. The company took the machine – dubbed the ‘robot with a heart’ –international in 2016. By the end of 2016, the company had sold 10,000 units worldwide (costing $1,600 per robot, plus a $360/month data price plan). Importantly, it claimed Pepper was paying its way through improved customer experience and revenue growth in the retail chains where the android is a sales assistant.

Still in Japan, robot receptionists – one of them strikingly shaped as a dinosaur – have been welcoming guests at the Henn-na hotel in Nagasaki since the summer of 2015. This builds on a long tradition of highly automated hotels – available for workers who miss the last train home or young couples looking for privacy. In Nagasaki, the hotel aims to have robots as 90% of its staff. While able to operate 24/7, the receptionists only start checking guests in from 3pm, and requiring human intervention such as help in checking the passports of non-Japanese visitors.

Many observers, such as technology writer Clint Boulton, are bullish about humans and machines complementing each other, pointing out that the automation of repetitive tasks will create new roles. These will improve customer experience and make work more interesting for employees. He notes that, at AT&T, “bots pull sales leads from multiple systems, enabling staff to spend more time with customers.”

Robots, it seems, can positively affect revenue growth, cost reduction and customer experience, arguably the three pillars of a healthy corporate strategy.

The rise of robots
These examples, however, are only the tip of the automation iceberg. According to the International Federation of Robotics, the global average for robotic density in manufacturing (measured in number of robots per 10,000 jobs) was 66 in 2014. With investment growing at more than 40% annually, the Federation forecasts that 1.3 million robots will come into service by 2018.

Twenty-one countries score above this average of 66 (the UK is 21st). The top three countries were South Korea, Japan and Germany, with density ranging from 300 to 500, up to eight times the average.

One of the obvious ramifications of a rise in the use of robots is the impact on jobs. A September 2016 report by Forrester predicted that robots will eliminate 6% of jobs in the US by 2021. In transport, the driverless car trials run by Uber in Pittsburgh in 2016 make cab and truck drivers one of the likely targets of this rapid transformation.

The fast-food industry is ripe for a shake-up. Robots, in the form of kiosks or touchscreens, were always able to take orders. With, the Pepper generation, they can read emotions and are starting to appear behind the counter – Pepper already operates in Pizza Hut in Singapore. In tandem with backroom robots that are incomparably better at flipping burgers than humans, they may soon provide a fully robotic fast-food experience at, say, McDonalds (one robot can cook 360 patties an hour, according to Gizmodo).

But these (arguably low-skilled) roles are not the only jobs that robots can eradicate. US author Marshall Brain points out that airline pilots could be easily replaced by robots, given that much of their work environment is already automated. And there are other examples of machines taking on (at least part of) jobs seen until now as the preserve of knowledge workers: Amazon’s Alexa is helping our children with their homework and surgeons are operating remotely using robotic arms.

But this evolution could plant the seeds of deep social unrest, for if robots help us in stores and hotels, drive our cars and pilot our planes, do our kids’ homework, treat us when we are ill, trade stocks and fight wars on our behalf, who needs salespeople, receptionists, taxi drivers, pilots, teachers, doctors, traders or soldiers? And how will all these deposed people earn a living?

And what about marketers?
Intelligent machines already fulfill a good chunk of the 21st century marketer’s mandate. As we have seen with Anna and Pepper, robots have extended their role into the way loyalty and CRM strategies are executed. It follows that performance indicators such as ‘customer lifetime value’ and ‘net promoter scores’ are now dependent on machines and artificial intelligence (AI).

Undeniably, the marketing skill set has already changed to include disciplines, such as SEO and data analytics, that are closely connected to AI. In its 2017 survey of services marketing trends, ITSMA highlights a range of marketing priorities that are climbing up the to-do list of B2B CMOs across the world. Marketing technology and automation systems, for example, which ranked 31st in the 2011 survey, is now the eighth priority for marketers, the largest jump of any of the priorities tracked by the association in that six-year period. Digital and social marketing, perhaps the best example of a digital skill enabled by AI, was eleventh six years ago and is today fifth. Lead generation, management and nurturing, a marketing competency, which has fast evolved to rely on statistical and CRM engines, has shot up from ninth in 2011 to second in 2017.

The question is: what will this new phase of marketing digital transformation look like? The ITSMA report brings a helpful perspective. The topranked areas for marketing development and training blend what could be termed machinecentric competencies with softer, customer-oriented skills:

  • On the data and machine side: marketing automation (3rd), social media (6th) and data analytics (9th).
  • On the content and customer side: content creation (1st), accountbased marketing (4th), thought leadership development (5th) and storytelling (8th).

Digital user experience, ranked seventh, is the ideal fusion of these two categories of skills, and arguably marks the attributes of a successful digital marketer. As they map the customer journey, marketers need to blend on- and offline worlds, understanding, testing and measuring crossover points between channels. Data analytics can help to decide journey steps and shortcuts as well as next best offers. In the design of attractive, easy-to-consume content such as thought leadership, creativity comes to the fore. Serving up the right content in the right context will be tested too, needing the help of intelligent machines to deliver the best business outcome possible.

Like many marketers, I believe our profession is a carefully tuned balance of art and science. With more of the science of marketing being taken care of by robots, it may feel natural for human marketers to retrench and specialise only in the art of marketing.

But, how long would this be sustainable? Until a robotic Michelangelo, Ridley Scott, or will.i.am turns up, that’s when. And this may not be very far away.

In August 2016, 20th Century Fox released Morgan, a thriller about an AI humanoid that goes wrong. In what must be the most automated storytelling example ever, Watson, the IBM AI supercomputer, created the movie trailer. Using 100 horror film trailers, Watson learned the dynamics of a trailer. It then processed the 90 minutes of Morgan to generate the first draft of its promotional clip. In doing so, it reduced the trailer production time from between ten and 30 days to just 24 hours. Jukedeck is another surprising example of AI applied to a creative discipline, namely songwriting. The London-based start-up has taught a computer to write songs in seconds, and the result is royalty-free. The idea is to provide affordable soundtracks to advertisers unable to pay for artists’ intellectual property – and has the potential to disrupt the music industry profoundly.

So for marketers, the implication is clear. A successful career means the mastery of a balanced set of skills – some highly numerical and scientific, others more intuitive and emotional. Failing to cover both may lead to experiencing Anna’s forced fate of untimely retirement.


This article was first published in Market Leader Quarter 2, 2017