Who to Follow: Paul Roetzer (@paulroetzer )
Paul Roetzer really likes marketing. He also really likes AI. Basically, he was a guy we at Phrasee knew we needed to talk to.
A 13-year marketing veteran, and the founder of The Marketing AI Institute, Paul found his way into the marketing field shortly after graduating from Ohio University’s E.W. Scripps School of Journalism in 2000. By 2005, Paul had founded his own inbound marketing agency, PR 20/20 which went on to become HubSpot’s first partner agency in 2007.
Author of 2011’s The Marketing Agency Blueprint and 2014’s The Marketing Performance Blueprint, Paul’s insatiable thirst for knowledge about what makes marketing effective eventually led him to begin looking into emerging artificial intelligence technologies and how they could be leveraged in automated marketing strategies.
Then, in 2016, Roetzer founded The Marketing AI Institute, which endeavoured to make AI technologies more approachable and actionable for marketers across the globe.
Combine an extensive understanding of AI’s emerging role in marketing with a lively social media presence, and do you know what you get? One of Phrasee’s very phavourite follows!
— Paul Roetzer (@paulroetzer) April 22, 2018
“The biggest issue I see with so-called AI experts is that they think they know more than they do, and they think they are smarter than they actually are” https://t.co/2yKm2mHrek
— Paul Roetzer (@paulroetzer) April 10, 2018
Now that we’ve etablished Paul Roetzer’s AI and social media bona fides, let’s see what else he has to say…
Paul Roetzer tale of the tape
Favourite food: Pistachio delight (my mom’s dessert speciality)
Pets: Two cats
Dream job as a child: Doctor
Last big purchase: LEGO NASA Apollo Saturn V Building Kit (I’m a total space geek).
Guilty pleasure: Million Dollar Listing. It’s my one reality show.
An interview with Paul Roetzer
For those who might not know, what is the Marketing Artificial Intelligence Institute, and how does it make life easier for brands and marketers?
We launched the Marketing Artificial Intelligence Institute in 2016. AI is abstract and overwhelming to most marketers. Our mission is to make it approachable and actionable. We want to give marketers who take the initiative to learn and apply AI a competitive advantage in their careers.
What inspired you to create the Marketing Artificial Intelligence Institute?
I developed an interest in AI in 2011 when IBM Watson won on Jeopardy! I didn’t understand how the technology worked at the time, but I had a vision to someday use it to intelligently automate marketing strategy. My interest grew into an obsession as I read about and researched the history and application of AI in other industries.
After a couple years of flirting with the idea of building a technology company to bring the marketing intelligence engine to life, I decided to tell the story of AI instead. Thus, the Institute was born.
How does AI drive marketing performance in the current landscape, and how will it drive marketing performance in the future?
It depends on the use case. AI is built to perform very specific, or narrow, tasks. For example, Phrasee uses it to write email subject lines that outperform human written copy, thus increasing open rates, click-throughs and sales. BrightEdge uses it to surface opportunities and insights for SEO that bolster organic traffic. Crayon uses it to track the digital footprint of competitors, which can be used to inform marketing strategy.
The vast majority of marketing AI technology today is focused on using machine learning—a subset of AI—to make predictions based on historical data. The algorithms look at what has happened before and predict what will happen in the future. Sample uses cases include predicting email clicks, content performance, lead conversions and customer churn.
The future of AI will be about prescribing strategies and tactics based on goals. The marketer’s primary role will be to curate and enhance algorithm-based recommendations rather than to devise them. Input a goal, say, for example, 500 qualified leads, and the machine will run models and recommend actions based on probabilities of success. But, this is a very hard use case to solve, and I’m not aware of any companies that are even close.
Humans are limited by their biases, beliefs, education, experiences, knowledge, and brainpower. All of these things contribute to our finite ability to process information, build strategies, and achieve performance potential.
Algorithms, in contrast, have an almost infinite ability to process information. They possess the power to understand natural language queries, identify patterns and anomalies, and parse massive data sets to deliver recommendations better, faster, and cheaper than people can.
What questions should a brand ask itself before adopting AI into its marketing strategy?
- What repetitive, manual marketing tasks do we do every day/week/month/quarter that could be intelligently automated?
- What opportunities are there are to get more out of our data—discover insights, predict outcomes, devise strategies, personalize content and tell stories at scale?
- What are the AI capabilities within our existing marketing technology stack? Can we drive greater efficiency, personalization and performance without having to add more vendors?
How long will it be until all of the world’s marketers are all replaced by AI?
I don’t see that happening. Some jobs will be replaced. Others we can’t imagine will be created. But, for the foreseeable future, this is a story of human knowledge and capabilities being augmented by machines. That doesn’t mean there won’t be painful transitions at times, but the transformation will be mostly positive for brands and marketers. And, most importantly, for consumers.
What areas of marketing do you think are ripe for AI disruption at the moment, and what is standing in the way?
Anything that requires data to inform decisions and actions. A couple that immediately come to mind are media buying and marketing automation. The irony of marketing automation is that it’s largely manual. Marketers set all the rules that tell the automation software what to do. That will change.
It’s still early. While venture capital money is pouring into AI-powered marketing companies, many of them are unproven, with limited case studies and an unknown ability to scale.
Also, unfortunately, AI is becoming an overused term, which causes confusion in the marketplace. The reality is many of the companies claiming to be AI-powered aren’t really that advanced, yet.
What are the biggest problems that you envision arising in the AI marketing space in the next 2-3 years that people might not have thought of yet?
Because AI is built to solve narrow use cases, marketers may need to add a collection of new solutions to their marketing tech stack. One for email send-time optimization. One for email subject-line writing. One for content recommendations. One for predicting content performance. One for etc., etc., etc.
This brings its own complexity and integration challenges. But, assuming you can solve that, there’s a bigger challenge on the horizon.
There’s a lack of AI talent available to the big marketing platform companies who are all racing to make their solutions more intelligently automated. Each of the AI use cases we’ve discussed is a potential feature within a larger platform. So, the logical play is to acquire promising young AI companies.
Thus, the AI companies marketers come to rely on over the next 2-3 years will likely get acquired and absorbed into larger marketing software companies. Then, it’s back to square one.
What is your least favourite thing about working in marketing?
The need, or, at least, the perceived need, to always be connected.