Marketing is overflowing with data. It informs and guides nearly every decision we make as marketers. AI, and, more specifically, machine learning, can help. But, ultimately, we are just people trying to create meaningful connections with other people who are making purchasing decisions based on very human factors.
Just like a copilot, AI assists but is not in command. You are. In episode 1 of season 2 of Emodo’s FIVE podcast, Everybody Gets a Copilot, Jake Moskowitz and his guests explain how AI assists and elevates the thinking and roles of marketers.
Hear from Shelly Palmer, CEO of The Palmer Group; Rishad Tobaccowala, author and senior advisor to the Publicis Groupe; and Michael Stich, Chief Business Officer at agency VMLY&R.
Jake and his distinguished guests explore how AI fits into the marketing org chart and address timely questions. When is the machine your partner, as opposed to the tool you’re using to accomplish your goal? What happens when machine learning replaces the chain of human roles — and training — in a marketing creative department? What are the biggest gaps to implement AI in marketing?
Explore the human-AI connection with them.
Host Jake Moskowitz and his guests explore the human side of AI in this first episode of the new AI series. Shelly Palmer shares some common, compelling examples of machine learning, muses about how to be an “AI co-worker” and offers some sharp advice to marketers on how to vet the AI claims of vendors. Rishad Tobaccowala cautions about AI as a buzzword and describes the three turds on the table (a reference to one of the chapters in his book): Ignorance, fear and science fiction. Michael Stich returns in the new season to talk about AI as the marketer’s co-pilot and how viewing the AI-powered future through that lens can help marketers see AI in a different light. Jake and Jeremy Lockhorn talk about the movies that have shaped our perceptions of AI and weigh the accuracy of Hollywood’s depictions.
And, of course, Jake hits the FIVE List.
The FIVE List: The human side of the AI discussion
- Marketing is about ideas, the domains of people.
- Marketing’s job is to connect with people.
- Scale and complexity are so great, marketers need help (from AI).
- People make purchase decisions based on very human factors.
- Algorithms are programmed by people.
Rishad Tobaccowala, Author, speaker and Senior Advisor to the Publicis Groupe.
Shelly Palmer, CEO of The Palmer Group, columnist at AdWeek and regular commentator on CNN and CNBC.
Michael Stich, Chief Business Officer, VMLY&R, frequent speaker and writer on a variety of digital marketing topics.
Jeremy Lockhorn, Global Head of Partner Solutions at Ericsson Emodo, speaker, mobile marketing expert.
In this episode, Rishad Tabaccowala references his book: Restoring the Soul of Business: Staying Human in the Age of Data and Shelly Palmer references free weekly courses and gatherings offered by The Palmer Group.
The Five podcast is presented by Ericsson Emodo and the Emodo Institute, and features original music by Dyaphonic and The Small Town Symphonette. This episode was edited by Justin Newton and produced by Robert Haskitt, Liz Wynnemer and Jake Moskowitz.
Transcript of S2 E1: Everybody Gets a Copilot
Everyone gets a copilot in their life, whoever you are and whatever task you have. As marketers we get copilots.
Let’s talk AI.
Welcome to FIVE, the podcast that breaks down the big transformations for marketers. This is the start of season two. This season is all about AI in marketing. Here’s where we start, episode one: Everybody Gets a Copilot.
I’m Jake Moskowitz.
There’s a lot of talk about AI, and not just among techies, data scientists, and sci-fi fans. AI is already transforming the ways businesses operate, how today’s products and services work, and how careers are defined and built. And the pace of AI driven change is only accelerating. The impacts of AI among consumers, brands, and agencies will be significant. As marketing is one of the most likely industries to be impacted by AI. But how? What do marketers need to know now, and why? That’s what we’re setting out to explore.
My name is Jake Moskowitz, I’m the head of the Emodo Institue, an orginazation that focuses on research and education of data challenges in marketing. I’ve been involved in marketing and data analytics for over 20 years. The Emodo Institute is part of Emodo, a wholly owned subsidiary of Ericsson, the global telecommunications company.
To some people, AI sounds scary. You know, cyborgs, robots, the end of humanity. To others, AI’s just a distant, irrelevant topic, the boring chatter of nerds and futurists. To most, AI just so unds complicated. But if you’re a business leader, or a career minded marketer, AI really shouldn’t be any of those things. It’s simply a set of capabilities that if well understood, can make your business better and your marketing more effective. And maybe most important, understanding AI will help you become more successful, because AI will impact a wide range of marketing jobs, capabilities, and functions. So understanding AI means opportunity, or put another way, not understanding AI, is likely to mean lost opportunity. Somehow, someway, AI is powering just about every industry you can name. Whether it’s automotive, telecommunications, or education, finance, gaming, healthcare, logistics, or countless other sectors, there’s some AI algorithm doing something in the background to make it work, probably lot’s of them. For example, AI is playing a huge role in world wide efforts to fight the covid-19 pandemic, it’s being used to determine early diagnosis of people who might be covid-19 positive. It’s being used to project infection rates, ICU capacities, and death rates in likely hot spots. AI’s also being used to shape and identify potential vaccines to push forward through the trial process, so lots of different uses and applications in just that one particular area. AI can be applied to all kinds of tasks to make them more scalable, smarter, faster, more accurate, make it better in numerous ways and even make some tasks possible for the first time. That sounds pretty good, right? Well not so fast. Throughout the first few minutes of this episode, I did what a lot of people do. I talked about AI as this big sweeping, singular thing without any degree of specificity or any real details. But when you hear sweeping, glowing statements like that, it almost sounds like AI is this infallible ingredient that produces solely positive outcomes and incredible benefits like something you might see promoted on a cereal box like: Cheerios, now made with real AI.
Just so we’re all starting on the same page, let’s take a minute to talk about what we’re talking about. AI, what is it really?
The term AI gets thrown around a lot, without definition or specificity. So it really is kind of a vague meaning, the truth is that artificial intelligence takes many forms, can be fraught with a variety of issues, as different models and depending on the application, can mean different things. And every company applies it and approaches it differently in just about any application, and in all those variations there are all kinds of potential issues. In your business you may have noticed a growing trend of vendors touting their proprietary artificial intelligence of their super smart algorithm. After this season of FIVE, you’ll know whether or not something’s of value beneath the surface, whether or not there’s something meaningful behind the vendor claims. You’ll know how to spot red flags, what those flags may be telling you. You’ll know where to probe, what questions to ask, and what you’re looking for behind the buzzwords. In fact let’s take a few steps beneath the surface, behind the buzzwords now.
In most marketing scenarios when people talk about AI, they’re talking about machine learning, which is essentially an algorithm that gets better as you feed it more data. Generally, uses and applications of machine learning are only growing more and more prominent in marketing and beyond. Some applications of AI have become pretty high profile. For example, autonomous vehicles. It is fair to say that AI makes automated driving possible, but you have to step down a little bit to understand how that actually works. Automated driving is made possible by three categories of machine learning algorithms: one that identifies the objects around your vehicle, one that determines what those objects are doing or are about to do, and one that decides on behalf of your vehicle what to do about it. Note that each of those categories involves multiple algorithms that have very specific roles, there’s no all in one algorithm that’s just programmed to drive a car.
To some degree AI, particularly machine learning, already enhances a wide range of marketing solutions including things like hypersegmentation, dynamic creative, inventory quality filtering, dynamic sites and landing pages. But there are lots of things that can get in the way of an algorithm’s success. If you don’t know much about AI, you might not even know why things went south or even that they went south at all. So it’s these issues where we’ll put a lot of our attention in upcoming episodes.
One more thing about machine learning, throughout our series, we’re primarily focused on supervised machine learning, that just means that the algorithms learn to predict outcomes by learning from specific, intentional data inputs. Essentially, you show a computer a bunch of examples, you tell it the right answer for each example, and then based on these examples it’s seen, the computer starts to predict the right answer on its own. Feed it more examples, and the predictions get more and more accurate, well that’s the goal anyway.
Okay before we go any further, we need to talk about something else. Up to this point I have done something a lot of marketers do. It’s not intentional and it’s not good. You know we as marketers are constantly chasing metrics and conversions, optimizing creative and media, tracking ratios and percentages. How many impressions, clicks, views, visits. You know what we don’t talk about? People.
Seriously, we’re an industry of euphemisms like audience, consumers, visitors. We define personas, targeting segments, talk about personalized this and that, and we buy and sell impressions. So often we talk about programmatic innovation, demand sides, sell side, viewability, fraud, safety, that the human element, the real human element gets a little lost. In the day-to-day grind it’s easy to forget that we’re actually just an industry of people, trying to make a meaningful connection with people. Most AI in marketing is there for the purpose of either helping or reaching people. So in my view people-based themes are essential for framing our perspectives and discussion around AI, and there are actually quite a few of those. In fact to be specific there are five.
Marketers are people, and in an industry so complex and so dependent on scale, the people who make it work need help.