A.I.mpact, Part 3: Who’s On Your Marketing Team of Tomorrow?
Whether they’re in marketing or some other discipline, it’s odds-on that when artificial intelligence in the workplace is mentioned to somebody, one of the first thoughts to leap into his or her head is, “Will it take away my job?”
It’s a fair question. And people have been asking it for a really long time.
The tension between machine learning and human intellect has been routinely reflected in popular media: In our past two entries in this series, we did call-outs to an Avengers movie, Star Trek and Mad Men for their takes on AI as a nefarious change agent who’ll suck the humanism out of our lives. Or just zap us to atoms.
But as far back as the 1950s, the notion that computers might supplant us in the workplace was very much on people’s minds. So much so it was a comic premise of a 1957 Tracy-Hepburn comedy, Desk Set.
The movie’s third lead? The computer installed by Spencer Tracy’s inventor in the reference and fact-checking department of a fictional TV network. Its name, EMERAC, was a take on the pioneering ENIAC, but also sounded suspiciously like ipecac, the emetic. Though Kurt Vonnegut had already beaten them to that joke in a 1950 short story, EPICAC.
In Desk Set, Kate Hepburn’s staff worries they’ve become expendable. When EMERAC mistakenly pink slips everyone in the company, including the CEO, it’s a gag with unnerving undertones.
How will AI really affect the structure and staffing of a marketing team as we move forward? Will a “marketing department” even be recognizable a decade or more down the line?
What disciplines will be valued? And what jobs may end up as obsolete as punchcards, vacuum tubes and EMERAC?
US Army photo of ENIAC via Wikipedia
Global demand for data scientists will outstrip supply by 50% by 2018.
The economics of automation and AI
The fact that technological advances can cause job losses, at least in the short term, is taken as economic gospel. Yet whether or not it results in long-term unemployment has nearly always been up for debate.
As far back as Aristotle, thinkers were debating the issue of machines displacing human beings at various types of labor. Would innovation disrupt employment over the short haul, but reward society with more productivity, expansion of opportunity, even the creation of new industries and wealth?
Technological unemployment was the term coined for job displacement. But as the Industrial Revolution progressed, it became obvious that technology was delivering huge benefits to every level of society, from white-collar to working class.
So any worry about technology causing permanent harm to the job market got assigned another term: “Luddite fallacy.”
But the shifts that artificial intelligence and robotics may create might be significant enough to create serious upheavals in business, banking, manufacturing, retailing and every other sector, especially since those changes are happening with remarkable speed. The questions are big enough that the World Economic Forum has tried to address it, Stephen Hawking has chimed in with dire warnings, and leading economists are attempting to model the impacts.
In other words? Nobody really knows anything yet.
But people and pundits are more than willing to hazard guesses:
- 93% of the major investors polled at the 2016 edition of Web Summit in 2016, a major technology conference, felt that AI would eliminate significant numbers of jobs and that governments weren’t prepared for the problem.
- On the other hand, 89% of that group also predicted Hillary Clinton would be the next President of the United States.
- Researchers at Oxford University estimate that up to 47% of U.S. jobs could be automated within the next two decades; whether that number includes politicians and economists was unclear.
- A new White House report sees the potential for AI to create new career opportunities, but may hit hard at lower-level jobs.
- Forrester foresees how 16% of U.S. jobs will be replaced by AI by 2015, while while the equivalent of 9% will be created — a net loss of 7%, with office and admin support staff taking the biggest hits.
- But higher-wage jobs may suffer, too. Mary “Missy” Cummings, director of the Humans and Autonomy Lab at Duke University — and a former Navy pilot — noted at a recent MIT conference how commercial pilots might be in the crosshairs since they “touch the stick for three to seven minutes per flight and that’s on a tough day,” with the rest accomplished via autopilot.
- A survey of business executives by the National Business Research Institute (NBRI) found 38% of enterprises are already using AI technologies of one kind or another, and 62% will employ AI by 2018.
One bright spot if AI becomes prevalent in the workplace may be an explosion of entrepreneurship as people innovate new businesses and livelihoods outside of those made obsolete by machine intelligence systems.
That’s especially true if social safety nets or even a “universal guaranteed income” are used to mitigate the impact of job losses.
Given a social safety net, entrepreneurship is almost to be expected. One economist at the University of Chicago has demonstrated what’s called the Peltzman Effect: when more safety features are introduced into cars, the number of fatalities and injuries doesn’t fall, because people compensate by taking more risks. The same may prove true in business.
“Companies that do not adapt applied AI will quickly perish or diminish in market size and value.”
AI will be a $37 billion industry by 2025.
93% of automation technologists feel unprepared or only partially prepared to tackle the challenges associated with smart machine technologies.
AI will force change — and help manage it
Programmer/pundit Scott Brinker came up with a revealing model (which he called Martec’s Law) for how technological change and organizational change happen. The challenge for businesses lies in the fact that they don’t follow nearly the same curve:
Technology changes evolve exponentially, while organizational changes happen logarithmically. Companies are constantly having to adjust and re-adjust to tech advances. And with the accelerating pace of tech change, that presents an even bigger issue.
Artificial intelligence and Deep Learning systems are driving many of those changes, but they’ll also provide a means for marketers to close the gap between tech’s potential and their own organizational capabilities.
AI will, first and foremost, deliver efficiency that’s impossible to achieve using traditional methods or human intelligence. Arduous, repetitive and complex tasks can be assigned to AI agents, from CRM engagement through customer service bots to far more sophisticated predictive analytics functions.
“Anything that seems rote or mechanical, there is no reason for humans to do—it’s all going to go to AI.”
As the chart above shows, deploying AI to tackle these tasks isn’t an optional scenario. It’s a competitive mandate, compelled by the huge tides of data we’re generating and the multiplying number of touchpoints we’re creating, where consistency and seamlessness of personalized experience is becoming the customer’s default expectation.
Only AI implementation will allow us to keep up with the escalating workload we’ve created within 21st century marketing departments.
An AI jobs map
Here’s a topline grid of what kinds of jobs AI will take on in tomorrow’s workplace – and why:
Important points to remember?
- Simply because AI may be able to take over any type of job is no guarantee that it will.
- When the Industrial Revolution eliminated many jobs, it also drove innovation and new forms of work and wealth benefitting far more people than were displaced.
- Despite the qualms of people like Bill Gates and Elon Musk about AI, the majority of AI researchers agree that fears of machine superintelligence are groundless.
Marketing functions of the (very near) future
Before we outline the structure of tomorrow’s prototypical marketing department, let’s lay out the functions it’ll have to manage, since those will dictate exactly what roles need to be filled.
Here’s a simplified model of the tasks of a marketing team in the near-future omnichannel universe of customer engagement:
- Through Analytics, marketers will seek to better understand prospects and customers, and search out new opportunities to engage and sell.
- The Experience they deliver will have to be remarkable, customized, engaging and consistent across every single touchpoint.
- Communication will, in a social channel-dominated world, entail ongoing two-way dialogues that are personalized to every customer.
How does AI factor into these?
- In Analytics, collecting and analyzing all the data necessary for finding actionable insights that can power 1:1 engagement — in real time, yet! — is only viable using AI.
- The customized Experience marketers must deliver to each customer will only be reliably consistent across all their various touchpoints if AI is coordinating every step of the sales journey.
- Communication will depend on machine intelligence due to the scale of engagement across social media and mobile, as millions of customers each demand an instantaneous and personalized dialogue with a brand.
A more integrated (and innovative) team
For more integrated and seamless customer engagement to really work, marketing teams will need to be far more integrated and seamless than they are today. Silos and barriers between key players and resources will have to be torn down in order to give an enterprise’s marketing the agility it’ll need to stay competitive.
Today’s divisions – between marketing, advertising, leadgen, sales and other disciplines – will need to disappear so marketers can prioritize the customer experience and journey.
One global study by the Association of National Advertisers, the World Federation of Advertisers, EffectiveBrands and Forbes predicts that tomorrow’s most successful companies will follow exactly that model. They’ll use hub-and-spoke structures with a CMO (or even a “Chief Experience Officer”) in the center, with managers interconnected in a constant web of collaboration.
Graphic source: Forbes
AI will actually expedite this. By intelligently automating some workloads, organizations are free to eliminate many traditional, silo-style managerial hierarchies. Processes and workflows can be more streamlined, so marketers can focus on innovation and strategy, not the time-sucking minutiae of program administration or execution.
Who goes, who stays?
Gartner projects that 30% of our jobs may be taken over by AI by the year 2025. But that doesn’t mean marketers will get kicked to the curb en masse by AI. Many will be liberated, so their energies can be devoted to those areas where creativity, empathy and human engagement matter most. Or they’ll work in jobs that hybridize the respective strengths of human beings and AI.
A lot of tomorrow’s marketing jobs are beyond imagining right now: In 1986, what would you have thought if someone said they earned their living as a “webmaster”?
That said, a lot of the present-day jobs that deal with the more mechanical mundanities of marketing will be lost:
- Customer service and CRM will migrate to AI; customer dialogues conducted via the keyboard are increasingly being handled by bots, and conversational UIs will advance to the point where voice communication will be a customer-machine activity.
- Persona development, lead generation, acquisition and conversion will be managed by machines. Hands-on manual processes for creating buyer personas will be obsolete when AI can build real-time prospecting profiles using detailed social data. On the sales side, leads will get fed to specific salespeople (or “personalized” salesbots) based on data-based “matchmaking” to ensure the right fit between prospect and sales rep.
- Market research and business intelligence is a natural fit for AI, whether it’s charting social sentiment, analyzing online communities, identifying audience trends or microtargeting block-by-block, house-by-house sampling programs. One U.K. agency is using IBM’s Watson platform to generate psychological profiles to help in ad targeting. The focus group may never go away, but will it someday be facilitated by an “employee” like Jia Jia?
- Content management and development for CRM, email, websites and digital advertising will increasingly fall to AI, allowing publishers to do more with fewer human staffers. “Cognitive content platforms” like Persado already use semantic algorithms to create optimize email subject lines, and crunch massive amounts of data to script personalized online ads; tomorrow, each person’s customer journey will be de facto unique and utterly customized.
- Media buying is a natural fit for AI. When IBM tested Watson for programmatic ad buying, it was able to make instantaneous decisions about whether or not to place a bid while learning which buys were most effective, resulting in a 35% lift in campaign results.
Human assets still offer human advantages
There are aspects of marketing, advertising and the “art of persuasion” that AI may find tough to take on, at least in the near term, according to a 2013 study by Oxford researchers about the future of employment and job automation.
The paper took an in-depth look at the history of labor and analyzed 702 occupations, and found roadblocks to automating certain occupations, with four main skills posing the thorniest challenges:
- Originality: The ability to come up with absolutely new creative ideas or approaches to solve a problem, rather than creative synthesis or recontextualization of existing ideas into new forms, where AI is already making advances.
- Fine arts: Knowledge of theory and techniques required to compose, produce, and perform works of art, like dance, visual arts, drama, and sculpture.
- Social perceptiveness: Being aware of others’ reactions, and understanding why they react as they do.
- Persuasion: Persuading others to change their minds and behavior.
Any job which relies heavily on these talents is unlikely to be entirely deputized to an AI…at least, for now. When McCann Erickson Japan pitted an AI “creative director” against a human one to develop separate campaigns for Clorets, consumers voted for the latter. This demonstrates how AI can’t yet reproduce the emotional connections inherent in a good creative campaign — but the AI-versus-human voting ran close.
A garden of growth, not scorched earth
A lot of AI experts shake their heads at the dire forecasts of doom-and-gloomers. To them, the story will be about job creation, not attrition or eradication. Particularly when it comes to marketing.
Ari Sheinkin, VP of marketing analytics at IBM, says their work with Watson has proven the power of AI to actually create new marketing jobs, versus discarding existing workers. “When we roll out a new capability, our experience is the exact opposite of what you see in the press. We go out and hire new people because our capability is so much more. I think you will see more of that.”
New roles for data scientists, automation specialists, robot monitoring experts, content curators and more will probably populate the rosters of in-house marketing departments and external agencies in the years ahead. Those may only be the beginning.
“There are many ways of going forward, but only one way of standing still.”
Franklin D. Roosevelt
Marketing AI’s happy ending?
Is there a secret to thriving as a marketing professional in our AI-powered future?
Sure. It’s a simple constant that’s held true during any period of change and disruption: Flexibility and agility always win the day.
So marketing pros need to stay educated about the changes going on around them, and assertive about exploring ways to partner with machine learning systems to deliver better experiences and value to their clients and customers.
Fact is, maybe the last word on this already got said back in 1957.
By the end of Desk Set, the dreaded EMERAC turns out to be a helpmate for its human colleagues, not a replacement. It’s a tool they’ll be able to use to improve the work they already take pride in doing.
As some people prophesy and panic about how AI will affect tomorrow’s marketing jobs, that’s a plot point worth remembering.