toggle mobile navigation

How Artificial Intelligence Will Redefine Marketing Jobs

Posted by Mike Kaput on June 7, 2018

agreement-business-businessman-872957This post originally appeared on the Marketing Artificial Intelligence Institute's blog. The Marketing AI Institute was created and is powered by PR 20/20. 

Will AI replace your job?

To be honest, no one knows for sure. Not yet. Which makes it more important than ever to ask the question.

We know AI is advancing rapidly. Image and voice recognition have become serious technologies. Natural language processing and generation are progressing. Neural networks push the envelope of how well machines learn without human intervention.

This has led to some insane advancements. We all now have AI voice assistants in the form of Siri and Alexa. Self-driving cars are viable technologies—and self-driving trucks may be next. Google built AI that taught itself how to be superhuman at a board game—without human teachers.

It's still very early in AI. But these advancements mean we need to move beyond hype and start asking tough questions.

How might AI redefine marketing jobs? How might it redefine our entire industry? And how do we train ourselves and others to survive potentially massive disruption?

AI has massive potential to make our lives and businesses more successful. It also has the potential to create big winners and big losers in the economy. (See this McKinsey study that seriously looks at which human jobs will be replaced by machines.)

Google CEO Sundar Pichai called AI "more profound than electricity or fire." But marketers are still largely in the dark about AI’s implications.

We wanted to start a conversation in the industry about this. We're not seeing a lot of people talking about the very real potential economic disruption here. Yet we're seeing a lot of AI advancements progressing with scary speed towards this disruption.

We believe the marketing industry is approaching an inflection point when it comes to AI. And we see three broadly possible futures that could result.

Not all of them are likely. And they won’t happen overnight. Some or all of them might not happen at all. Or, we may be missing pieces of the puzzle that aren’t yet clear to us.

These aren’t really predictions. They’re jumping off points. We need to start asking questions about our businesses, industries, and careers in the coming age of AI. Here are a few talking points to start with.

1. The most likely scenario

At this stage, it feels safe to say that AI is here to stay in a meaningful way. AI systems can recognize images, process language, and analyze huge datasets.

AI powers Amazon and Netflix recommendations. It self-drives cars and trucks. It routes your UPS deliveries. AI runs your smart home. It powers Siri and Alexa. It trades stocks, sometimes better than humans. You probably use AI daily whether you know it or not.

Now, the extent to which AI will transform and disrupt industries is an open question.

One scenario is what we'll call the artificially intelligent future.

In this scenario, at a bare minimum, AI significantly streamlines our lives and work.

Expect AI to massively reduce friction between your intent and your outcomes.

Need to order something on Amazon? Say the word—and expect Alexa to recommend a complementary purchase.

Anticipating a grocery run? Your online shopping cart might already be full thanks to smart predictions from AI.

We may even get to a future where you get stuff you didn't even order—because a retailer anticipated your needs.

In marketing, AI will work behind the scenes to produce, perform, and predict at scale. Many tasks that eat up much of a marketer's day will be automated.

This includes things like drafting social shares or finding hashtags or uploading posts.

It includes things like writing the best headlines or A/B testing emails.

It may even suggest new campaigns or activities you didn't think to explore.

Marketers will have much more time for the strategic actions that drive results. They'll need that time. Because this future is already approaching. AI is already changing how marketing works.

  • It enables more personalized experiences at scale.
  • It enables voice to overtake typed queries as the dominant form of search.
  • It recommends content based on your preferences.

Marketers need to start thinking about how their content gets found in the age of AI. It's not just keywords and blog publishing now. And that fact alone torpedoes a lot of marketers' gameplans.

In this artificially intelligent future, marketers will probably love their jobs more. After all, you'll have what you've always craved:

Marketing that actually works. Clarity into the marketing that doesn't work. And data to prove both to the people who matter.

But to get there, you'll need to pay very close attention to how the game is changing.

You'll need to risk slaying some sacred cows, both in your company and in the market.

Because AI is likely to change how we interact with content, how we find products, and how we buy.

If this sounds transformative, you're right. It is. And it's what we consider the most likely outcome. But it’s not the only possible outcome. The future could get a whole lot weirder.

2. The pretty likely scenario

In our first scenario, AI permeates most things you do. And marketers need to adapt accordingly.

In this second scenario, it goes one step further. AI begins to directly augment your work.

We call this possibility the artificially augmented future.

You won't just use services and tools with AI baked in. AI will be your very real coworker. In some cases, AI may even be your boss.

This sounds sci-fi, but we think it's actually fairly likely.

Many tasks handled by marketing manager level pros and below will be automated. In fact, mid-level marketing job functions today will effectively not exist.

AI systems will do these jobs cheaper and faster than people can.

Now, we're not trying to say marketing manager job functions aren't valuable. They are. Very much so. And we're not saying companies want to eliminate this job function.

We're just predicting they'll have no choice.

Firm A and Firm B compete in, say, the manufacturing industry. Firm A's marketing department consists of the following:

  • One marketing executive who owns strategy and budget.
  • Three marketing managers who manage day-to-day campaigns and staff.
  • One or more teams of creators: content, visual, video, etc. who create marketing assets.

Firm B's marketing department looks like this:

  • One marketing executive who owns strategy and budget.
  • One marketing manager who manages and integrates marketing AI systems. These systems plan, produce, personalize, promote, and perform.
  • A network of freelancers and independent creators to fill gaps as needed.

Oh, and we forgot to mention that Firm B's marketing destroys Firm A's.

Firm B uses AI to dramatically improve performance. Firm B makes every single marketing dollar work overtime. And Firm B uses AI systems to identify rich opportunities completely unknown to Firm A.

Firm A, however, spends another year burning money. They fail to produce meaningful marketing results. And they watch Firm B eat their lunch daily.

AI's advantages compound. There is a very real first mover advantage to this technology. When Firm B uses AI, their moat widens more the longer they use it, since their systems will learn and improve.

They will capture even more efficiencies.

They will reinvest their time and resources into even more AI-powered marketing activities.

And they will rack up more and more wins while Firm A tries to catch up.

Firm A's executives become aware that Firm B is doing its marketing very differently. They will realize that their human teams at some levels are liabilities, not assets. They can’t perform as well as intelligent machines. They can’t move as fast. And they can’t scale. 

We can worry all we want, but it's just math.

As late as 1984, there were an estimated 40,000 telephone switchboard operators in the U.S.

Consider: we have to link to the phrase "telephone switchboard operators." Because there's a very real chance younger members of our audience have no idea what this is.

The market is the market. Platforms change. Consumer attention goes elsewhere. And if there is a vastly superior way to do a job, companies will eventually do it that way.

Even if the vastly superior way is to have an intelligent machine do it instead of a person.

We don't see this future as unlikely. It probably won't happen next year. But 10 years from now? Lots of smart money is betting that some version of this plays out.

The good news? Good marketing will be much cheaper to produce. And it'll work better. Which will create significant wealth.

This levels the playing field for smaller businesses. They won't need the massive budgets required today to compete with the big boys.

You're seeing an early version of this with social media. Smart practitioners at small businesses are leveraging the technology to outcompete big brands. They're winning because they take advantage of free distribution of their messages. They would have needed a huge TV or radio budget 20 years ago to do the same thing.

Similarly, AI may empower you to scale your marketing—without scaling your budget.

That's powerful. But it also means you might not need the human teams you needed in the past to scale.

It's difficult, if not impossible, to predict exactly how this plays out right now. But we'd bet money that the top, most in-demand marketers end up being one or all of the following:

  • Highly strategic executors. You are an idea machine that constantly brings revenue-generating activities to the table. Then you mobilize tech and teams to turn ideas into reality.
  • Trusted advisors. You are a client and customer service rockstar. You excel at creating massive value for people through your consultation and connections.
  • AI shepherds. You get AI. You know the tech, the tools, the possibilities. You can assess, adopt, and apply AI tools to achieve meaningful marketing outcomes. You're an "AI shepherd," who guides a flock of intelligent machines.

If none of these describe you or the track you’re on, you may find it challenging in this future to find meaningful and well compensated marketing work.

These roles are already evolving in the market. Glassdoor, a job site, listed out 15 of the top AI roles advertised on its site. The list doesn't just include engineers. It also includes AI copywriters, AI trainers, and AI-enabled marketing and product managers.

We implore marketers to work through this scenario. It may not happen. But if it does, you must be prepared. Start to understand exactly where you fit in currently. Look at how you might move into an area where AI augments rather than eliminates your role.

3. The unlikely (but “holy sh*t!”) scenario

This is the least likely scenario. But even if it's very unlikely, it's completely disruptive if it happens. So we do want to address it.

Before we dive in, we want to be very clear:

We work hard to avoid the hype that comes with a lot of commentary on AI. You can't search for AI-related topics without finding articles like "Will AI take all our jobs?" or "Will AI kill everyone?"

These get clicks. But they don't create much value for readers. We started the Marketing AI Institute because of all the hype. We wanted actual analysis on how AI would affect marketing.

But we do need to engage in some speculation here. Because there's a real possibility that AI gets insanely good at things, insanely fast.

Let's call this the artificially dominant future.

Some AI systems can improve themselves without human intervention. They can learn how to achieve their goals better, without being told how to do so.

There are limitations to this.

For example, a chess-playing AI can't decide to give up chess and learn brain surgery instead. It's as likely as your car waking up one day and deciding to be a horse.

But AI systems can improve themselves within their specific domains at scary speeds.

One very specific example is Google's AlphaGo Zero AI system.

AlphaGo Zero is an upgrade of a previous AI system. The old system destroyed human champions at the game of Go, which is like chess on steroids. The old system was trained by being given expert games to study, then it played itself to learn how to get better. The result was that in 2015 and 2016, the old system beat several Go champions, including one who'd won 18 world championships.

And that's nothing.

The new system, AlphaGo Zero, played the old system, its predecessor, and beat it 100 games to zero within 72 hours of learing the game. AlphaGo Zero learned just by playing itself. And within a few days, the system achieved "superhuman performance" at the game, destroying the previous AI that, in turn, destroyed the human champion.

Think about that. A machine started from scratch—and in just a few days it got superhuman good at a highly complex game. A game that most people would probably agree is too complex and strategic for them to master.

That system not only got good enough to beat an average human. It got good enough to destroy the least average humans on Earth at the game of Go. It literally beat the most distant outliers who have this skill.

That's not the craziest part.

The craziest part is that Go, thanks to its complexity and range of possible moves, was consdered a "holy grail" or "pinnacle" of AI research. In short, plenty of experts in the field of AI were stunned that AI had advanced sufficiently enough to beat humans at Go.

Even the best and brightest can't fully predict where this technology is going and how fast it's moving.

This begs the question:

So, what if this happens with peoples' livelihoods?

There aren't a ton of tasks that, individually, are more complicated that mastering the game of Go.

What if an AI system learns how to write better than humans? Or teaches itself how to drive perfectly within city limits? Or figures out how to market or sell or code better than pro humans?

That future looks challenging for almost all of us, to say the least.

Massive job losses in the millions could occur relatively quickly. These losses could potentially happen so fast that workers wouldn't be prepared. And if the workers aren’t prepared, our institutions and infrastructure probably wouldn’t be, either.

It might not be a scenario where you see the writing on the wall and prep accordingly. It might come out of nowhere and happen faster than you think.

One day, it looks impossible a machine could learn your job. The next day, the machine has not only learned it, it's become better at it than you.

This is heavy stuff with massive economic implications. We're talking Industrial Revolution levels of economic disruption. Google CEO, Sundar Pichai, recently told MSNBC, “AI is one of the most important things that humanity is working on. It’s more profound than, I don’t know, electricity or fire.” So, yeah, these questions and possibilities matter.

In this scenario, few people win outright. Beyond that, it's impossible to predict how this changes our lives, our businesses, and our social fabric.

What is certain is that we're not talking about it enough. In schools. In businesses. In the marketing industry.

We are not prepared. That's OK right now. After all, this technology developed fairly fast and we're still finding our footing.

But the fact we aren't even having the conversations is troubling.

We openly understand that we don't know how crazy this future is going to get.

We know we don't know how to prepare the next crop of college students for this.

We can't confidently say we're equipping the next generation of marketers and business professionals for greatness.

So why aren't we talking about it?

We'd like to change that and start a conversation in our industry. Subscribe to the Marketing AI Institute blog to join the conversation. 

MAII Blog

Image credit: Pexels

comments powered by Disqus