It has been predicted that machines will eventually overtake human intelligence thanks to the advancement of computer power and increase in data collection. But while fears that robots will one day wipe out the human race might be extreme and distant, the rise of artificial intelligence (AI) does have the potential to disrupt brands and those working within them in a far more immediate way.
So does that mean marketers jobs are at risk, and that a new way of thinking is required in order to adapt to this transformed employment market?
Programmatic buying and other automated techniques are already having a huge effect on marketing efficiency and the way the industry operates, and marketers at all levels now need to consider how AI and machine learning might affect their roles.
As Bank of England Governor Mark Carney stated in December last year, we are “in the midst of a technological revolution” that will “destroy jobs and livelihoods well before new ones emerge”.
‘Intelligent agents’ or AI will destroy 6% of all jobs in the US by 2021, according to research by Forrester, which shows the biggest effect will be felt in transportation, logistics, customer service and consumer services. A study by Oxford University suggests the top five jobs at risk of automation are loan officers, reception and information clerks, paralegals, retail salespeople and taxi drivers and chauffeurs.
These things don’t destroy jobs, they just change what jobs are needed.
Larry Kotch, Brainbroker
A separate study by Oxford University and Deloitte at the end of 2015 suggests that for marketers specifically the risk is less pronounced. For ‘marketing associate professionals’, it is fairly unlikely (33%) that their jobs will be automated over the next 20 years and for ‘marketing and sales directors’ it is very unlikely (1%).
Meanwhile, a poll of 800 senior marketing and sales professionals across the EMEA region, conducted by Oracle, shows the use of emerging technologies is set to surge by 2020.
It reveals that 78% of brands expect to provide customer experiences through virtual reality in the next four years, while 80% of brands will be using chatbots for customer interactions by 2020. It also finds that 48% of brands have implemented automation technologies in sales, marketing and customer service, with another 40% planning to do so by 2020.
Need for experts
But rather than destroy the need for a human workforce, there is an argument that the move towards automation will simply create new and different jobs.
And as technology use grows so does the need for experts on that technology. For a marketing professional, understanding how and when to use automation tools and managing these will be a big part of the role in the future.
“These things don’t destroy jobs, they just change what jobs are needed,” says Larry Kotch, co-founder of Brainbroker – a startup aimed at helping brands blend technology and talent. Instead, he says it “pushes people into the managerial, consultative and search function rather than the implementation function”.
Brainbroker acts as a consultancy, matching the latest software-as-a-service (SaaS) and marketing tools with clients looking to use the capabilities those tools offer.
Kotch says that although “people say marketing automation is taking jobs, the sheer number of these tools is warranting consultants for the tools themselves”. Since hundreds of tools are launched each year, he says there will be a “premium on having a human tell you what you should be using based on a specific use case”.
Kotch also believes that the onus is on the people in those jobs to “get a monopoly on how [tools] work” so when the transition happens, “they are the best placed people to take that forward and pioneer new roles”.
He adds: “There’s going to be an element on shifting rather than deletion of roles. The quicker you can get on top of this world, the more likely you will profit personally.”
Tech entrepreneur Jeremy Hindle, who co-founded entry-level recruitment app Headstart together with Oxford University student Nicholas Shekerdemian, agrees and says it is “not something people should be afraid of”. He advises anyone who is nervous about the effects of AI to read up on the subject and look into improving their technical skills.
He adds: “I’m not talking about the need to be amazing at programming, [but] reading around topics so they can understand concepts and feel more comfortable and position themselves in their roles as things develop – be engaged with what is going on.”
For younger generations it is also about education, according to Avril Murphy, vice-president of sales EMEA at Neato, which designs robots for the home. She says: “We are seeing four- and five-year-olds learning how to program – that is the way forward because you are already developing the skill set.”
AI in customer-facing roles
Yet despite the scope for professionals and knowledge workers such as marketers to adapt their roles to technology, many examples of automation and AI-based services replacing humans already exist and are making headlines, particularly in customer service functions.
So even if marketers can adopt tactics to survive the encroachment of automation onto their roles, they will still have the headache of working out where human employees are needed within the areas they are responsible for, and where these can be dispensed with in favour of AI.
Last year, Uber acquired an AI startup, as the race to launch driverless cars accelerates; Amazon tested the ‘Amazon Go’ store concept, which allows people to grab items and leave, with the products charged to their Amazon account on exit; and Starbucks let customers order a coffee via a chatbot. All are potential future replacements for tasks performed by humans.
Google Home and Amazon Echo virtual assistants also made an appearance in 2016 as part of a growing trend towards internet-connected homes. Intelligent virtual assistants provide information, add events to personal calendars, control other internet-connected services in the home such as heating, and can place orders on other websites.
The workplace upheaval that AI brings is being lauded as the fourth industrial revolution but experts warn that the upheaval and disruption could be like nothing we have seen before.
“AI and machine learning is a different beast,” says Kotch at Brainbroker. “It will be foolish to say it will follow that same trend that it did with the industrial revolution, computers and the internet.”
He says: ”It is something different and far more worrying in terms of just how much it can replace a human’s job.”
The fact AI is not yet replacing professional jobs such as marketing roles at scale means businesses and their workforces have some time to think about how their company should operate and what the jobs of the future look like. Retaining a human control element to automated processes may be one key step.
Controlling the effect on jobs
Harry Armstrong, senior researcher of technology futures at innovation foundation Nesta, focuses on researching the effects and implications of potentially disruptive technologies and believes that people can have control over how technological automation impacts jobs.
“One thing that is forgotten every time this automation conversation comes up is how much control we have on the design of jobs and the tasks within the jobs,” says Armstrong. Although repetitive jobs can be automated, the decisions affecting those processes may still require human input. He therefore advises the designing of jobs that require more task discretion rather than repetitive tasks.
When industries create totally automated systems, employees are quickly forced to conform and therefore the “narrative that we have control of, what those jobs will look like and what they do” is lost, argues Armstrong.
In some sectors a mix of automation and human interaction will be vital to how those products or services are received. It is the effect on the end result that matters the most rather than the technology, according to Rurik Bradbury, global head of research at messaging service LivePerson.
One thing that is forgotten every time this automation conversation comes up is how much control we have on the design of jobs and the tasks within the jobs.
Harry Armstrong, Nesta
The company believes that bots are not ready to be the primary customer service agent, meaning customer service jobs are not redundant yet but will change.
“Brands that have tried bots have forgotten that consumers don’t care [whether it’s a] bot or not, they just want to get customer service,” says Bradbury. “They treat bots the same way as an agent; they are either happy or unhappy.”
He advises a “hybrid of bots” where “you have humans behind the screen but you can scale up their effectiveness and how much they can accomplish selectively”. For example, simple tasks such as updating an address or a credit card, occasions where there is not much chance of a misunderstanding, can be done by a bot.
This gives customer service representatives the time to tackle the more open-ended or difficult problems and means customers do not have a bad experience with the bot or the human.
Headstart also requires a hybrid of machine learning and human input. The job app uses machine learning algorithms for a matching system that considers personality, interests, skills and demographic background, as well as traditional criteria such as qualifications and experience – asking students applicable questions using an AI-powered chatbot.
Hindle at Headstart says the human input on recruiting is a requirement because the algorithm learns based on how recruiters hire. He says: “That will always be the case because human factors and culture change and therefore no matter how good your machine learning algorithm gets it has to learn more – and for it to learn it has to have a positive response that it got it right.”
The majority of the disruption that AI will cause depends on how the people using AI-powered products react. Pictures of Terminator-style takeovers often accompany the headlines surrounding the use of robotics and machine-based services so there is a certain amount of fear attached.
US-based Neato sees some resistance to its robot vacuum cleaners, particularly in the UK “where [Britons] haven’t embraced the technology as much”, according to Murphy.
“The UK has the lowest penetration of dishwashers in Europe, so it could be a pride thing about not being a big deal to do your own washing up or vacuuming,” she says.
We need to do a better job of marketing and communicating because [people] have this perception and fear factor of having a robot in their home and [whether] they’re opening up their door to something that is going to be detrimental in the long term.”
AI-driven investment advisor Munnypot’s co-founder Simon Redgrove believes people will embrace its offer as they did online banking. The service was created because its founders believed existing auto-advice services were too robotic and wanted to create a more natural experience.
Consumers can use the service to answer questions about their life goals and can interact with different scenarios to understand how their finances could be affected by different decisions. “I’ve heard bad feedback and fear [about] using services like that but we are not experiencing that with ours,” says Redgrove.
He adds: “Robo-advisors have demonstrated that you can get the guidance you need; it’s subjective and not biased. People embrace it in the same way they do online banking and [it will be] interesting to look at [customer behaviour towards AI] in 12 or 18 months time.”
The onus of reskilling
For those already in the jobs that are being automated, it is down to individuals to adapt to change, but also to be aware that “it’s not going to happen overnight”, Neato’s Murphy says. “Not every taxi driver in London has to fear for their job. Other opportunities are going to emerge. There will be more opportunity in manufacturing or development that will attach around automation and AI.”
Armstrong at Nesta does not believe the onus on reskilling is as straightforward. He says: “There isn’t really a clear understanding of whose responsibility it is – the industry, the state or a personal responsibility.”
He adds: “If we are about to see a big disruption or at least continued change from technology, those are the questions we need to start answering now, coming up with plans for processes and structures to deal with them.”
It is not just self-driving cars and stores without cashiers that will have an effect on job losses. The biggest problem is digital technology, according to Alex Warren, senior account manager at technology PR agency Wildfire and author of Technoutopia, which looks at the negative effects of digital disruption on society.
He says that although Google is seven times the size of General Motors, it hires less than a quarter of the number of workers; WhatsApp is worth $19bn (£15.4bn) but has 55 employees; and Instagram was worth $1bn (£811m) when Facebook bought it and had around 12 employees at the time.
The danger is that digital applications are becoming increasingly “influential in our lives” but “don’t have to play by the same rules as traditional companies,” says Warren.
He says technology companies are “not job creators but job destroyers” and are “creating a massive wealth divide of a handful of billionaires and an army of people on zero-hours contracts, working in a gig economy or not working at all”.
Warren believes that the world has got “trapped in the idea that technology is always progress – it’s technological determinism”. He adds: “[It’s] the idea that if something is new and is technology-based, it is automatically progress. We can’t stop it, there is nothing we can do about it, we just have to accept that fact.”
A realistic timeline
Technological advances will continue and while the effects of that progress are becoming clearer, the speed at which it will happen is not. Getting a clear timeline depends on breakthroughs that could happen in a couple of years or a few decades.
Armstrong at Nesta predicts a “subtle shift rather than a sudden disruption” in terms of a timeline for change.
“It may be more of a problem in the short term than the long term,” explains author Warren. “In the short term there is going to be a massive wealth divide that will happen as a result, there will be job cuts. There will be lots of cheap goods and services, which is great, but it will have a knock-on effect.”
Warren also predicts a generational divide where those growing up in a digital world will not want the jobs that are being disrupted. He says: “[Young people won’t] grow up wanting to be a driver. Obviously those jobs are going and we’re not going to miss them.”
What does AI mean for marketing jobs
- Roles that involve a high level of social intelligence and original ideas are less likely to be replaced by machines.
- Adding in an element of human control to decision making will ensure marketers do not relinquish all responsibility to artificial intelligence.
- The move towards automation will free up time to spend on other, more strategic areas of business.
- The introduction of new tools and automated processes will change the job role; adapting now will allow marketers to lead the way.
- Machine learning requires human input, so there will be a demand for experts in specific areas to work with technology rather than against it.
- The creativity and business understanding required in marketing roles means they are relatively safe from being replaced, although purely executional roles, for example in media buying may be more vulnerable.