Will Artificial Intelligence Take My Marketing Job?

While we can talk about the goal of making this enrich people's lives and careers, there's always going to be a reality that if there's a way to intelligently automate things that previously required humans, they may not talk about it in public, but behind closed doors. They're trying to find a way to become more efficient because if they don't, their competitors will.

Sometimes when we talk about artificial intelligence, there can be almost a paranoia about computers taking over everything. Think there's any risk of marketers being replaced or really losing jobs at all by machine learning and artificial intelligence?

Yeah, so this is something that's almost always the first question I get. So I've done probably 50 or 60 talks on AI over the last two years and the audience always wants to know, like, oh my gosh, am I going to lose my job? So what I tell people is no one can look out five years from now and start truly accurately predicting the impact on jobs. Like anybody who claims that they know one way or the other is, is just guessing, maybe an educated guess, but they're just guessing because the technology moves so fast. It's primarily intended to augment our knowledge and capabilities. I think generally speaking, we can all agree that that's mostly what it's designed to do. It's meant to take very time intensive, data driven manual tasks and automate those where we can in the hopes then that marketers will spend more time being creative, being strategic, being empathetic, considering the emotions and the implications of the actions we take, in theory, that's where it's going to go. I do worry. I think there's a reality that many of the organizations racing ahead are publicly traded companies that have responsibilities to their shareholders. And while we can talk about the goal of making this enrich people's lives and careers, there's always going to be a reality that if there's a way to intelligently automate things that previously required humans, they may not talk about it in the public, but behind closed doors they're trying to find a way to become more efficient because if they don't, their competitors will. So I just look at it as it's a critical conversation that we have and that we illuminate because otherwise I think we might get into a place where three to five to seven years down the road, we do see more automation than maybe we wanted to in the process and we lose some of the human side of the brand. So do you guys think about that? I mean, do you consider the implications as you're building things a little bit.

I mean, I think you're right though. This is like a very long scale process. I've heard an analogy in the past made to like electrification. People being concerned about, you know, if we don't have or if we have electric drills or something on the factory floor that that might actually like replace certain workers. But they're going to be so much more efficient. You earlier referenced Google Maps. It's not preventing people from figuring out where they're going, but man, it makes it a lot easier. And every now and then you do get some pretty strange recommendations about making a turn or something that doesn't make sense if you're very familiar with the route and people are still there to correct and guide. So I think it's going to make people tremendously more efficient. But also like you mentioned for marketers in particular, there's a tremendous amount of creative work that marketers do. Content creation in and of itself, it's very difficult to automate or even to augment. Things like tone and style, which can be so critical to marketers when they're creating that content are very hard to fake.

With machines, I think there's going to be roles that will get replaced. People far smarter than us have done studies on it. Like the University of Oxford in 2014 did a study on the computerization of 702 jobs. And they were basically trying to predict out how likely a job was to be completely intelligently automated. And telemarketers, I believe, was number one on the list. And so I actually said that one time at a talk in New York and a guy like had an existential crisis in the crowd because he was a telemarketer apparently. And so he, the hand went up right after the talk. He's like, I'm a telemarketer. Like, am I gonna be at a job? And I was like, well, you may have, you may spend 10 years training the algorithms that replace what you do. I don't know. But that's where we're at. It's like, there's gonna be things where you and I could sit there and be like that totally going to be intelligent, automated. Like, why wouldn't it be? Like a machine can do it better than a human every time. So like an example would be again, a really tangible one. The Associated Press writes all their earnings reports using natural language generation, a very rudimentary form of AI and in some cases not really even considered AI anymore. But if you give it structured data, the columns and rows and then you train it, the story to write, give it synonyms, get it to branching logic. The machine can literally write thousands of them in a second versus a human that would have to sit there and pull the data and double check the data and all that so that the writing of earnings reports was intelligently automated. Now is it AI? You could probably argue both ways, but the reality is it doesn't have to be from a, you know, if you think about it on a one to five scale of full autonomy, it doesn't have to be a 5 to disrupt entire career paths and jobs. So if your job was to write earnings reports, you don't have a job writing earnings reports anymore. You now you may be doing more human interest stories and you may have built a more fulfilled job. Like you may do more enjoyable things and be more creative. But if that was all that was there for you to do, you're out of a job now. And I think there's a lot of roles in marketing that are like that. If your job is to ab test landing pages, you're not going to have a job. Hopefully you find something else to. But like that's not going to be there. You would think media buying won't really exist for humans in the future. Like that to me is a really practical thing. I've got a million dollars to spend. I've got a thousand places I could spend that money. How in the world is a human going to figure out the optimal way to do that? Like I graduated from college in 2000, there was like five places you could put your money in 2000. Google was a year old. The iPhone didn't exist. We didn't have social media. Blogging wasn't really a thing. Influencer marketing wasn't a thing. Like it was trade magazines and I don't know, like buy ads, send direct mail, telemarketing. Like there just wasn't that many places for it. So that's the kind of thing where you look at the complexity of doing that job now and you think about what machines excel at and there are definitely jobs that I could sit here and make a list of 20 of them for you that I don't think in five years will exist doesn't mean that person won't have a career anymore. They just may be doing something other than the thing that a machine can do better.

Yeah, I tend to agree. I mean within the broad fields of marketing, sales and services, you know, none of those are going to be replaced overnight, but particular tasks within that field. So that could be like lead qualification or some within sales or Parts of keyword research within marketing, or ticket routing within services, things like that, that are fairly redundant and there tends to be a correct answer.

I think we'll see. Intelligently automated analytics is one that I can't wait for it to get better, because trying to find insights, just take Google Analytics, for example, or even HubSpot's performance data. There's so much to analyze, and most marketers aren't analysts by trade. Like, we don't come out of school learning how to dive into spreadsheets and run pivot tables and, like, find insights. Out of the hundred things you could extract from last week's performance data, which five matter? Like, that's a really hard thing to scale. And so for me, like running an agency, I look at it every day. It's like, well, what are the things we are core to what we do, the success we can create for clients and that we're not really good at? And it's almost impossible to scale that thing. And analyzing analytics is right at the top of that list. So we've actually built a system where it takes insights, we use natural language generation, it takes the data, and it spits out a report. We basically just answer the same 10 questions about what happened on the site. Top pages, entry points, top conversion points, and we have the machine write the narrative. So we can literally write 25 client reports by pushing a button where it used to be 5 to 7 to 10 hours per client per month. So we took something. We didn't replace anybody's job. Like, nobody lost their jobs because of it. But we took an inefficiency, created greater value for the client and reduced the time we had to spend doing it.

Right. That also probably provides an opportunity to really dig in deeper on any of those insights that it may surface and provide greater value.

Yeah. The human looks at it and says, that makes a lot of sense. Like, I wonder why you can dig a little deeper. But the human doesn't have to sit there and stare at spreadsheets all day or look at the analytics data and the graphs and figure out what it's saying and then try and train someone straight out of school to do that. Like, it's just. It's not an efficient way to scale.

And I don't think anyone's gonna miss.

No, nobody has complained yet in the office that they don't have to do that thing anymore.