Learn more about Hidden in Plain Sight. PR 20/20 Shows Businesses That Benefits of AI are Already Here.
Carl Lewis: Welcome to the Connected Enterprise podcast, where our guests share how they stay connected in their business lives. I'm your host, Carl Lewis, from Vision33, and my guest today is Paul Roetzer of PR 20/20. Paul, welcome to the podcast. Tell us about your background, PR 20/20, and your role there.
Paul Roetzer: Thanks, Carl. I'm the founder and CEO of PR 20/20, which is a marketing services and consulting firm. I started it in 2005. I also founded the Marketing Artificial Intelligence Institute, which is an event and media company focused on artificial intelligence for marketers.
Carl Lewis: That’s one reason we wanted to talk with you – you know people across all the industries we associate with. What industry trends are you working with? What are people excited and making plans about regarding automation?
Paul Roetzer: For us, the big thing is helping marketers understand what artificial intelligence (AI) is and how to adopt it. Most people consider it abstract and overwhelming, so they avoid learning about it. They think it's like what they see in sci-fi movies.
Paul Roetzer: Which is the furthest thing from the truth. What you see in movies doesn't exist. AI drives efficiency. It reduces costs by intelligently automating repetitive data and data-driven tasks. It increases performance and revenue by improving your ability to make predictions. How it does that, and the terminology, is where it gets confusing. But we try to help people understand it's not that complicated.
Paul Roetzer: If you look at everything you do and try to find data-driven repetitive tasks or areas where you could make predictions, you realize how AI could help you intelligently automate. To clarify, automation is when you write rules to tell a machine what to do; intelligent automation means the AI figures out those rules without you or enhances your ability to create rules. Instead of telling the machine what to do, the machine figures out what to do, what to predict, and what actions to recommend.
Carl Lewis: The problem I find when I speak with customers looking at these technologies is that whether it’s AI, the internet of things, machine learning, etc., the biggest challenge is they can't go somewhere to see it work.
Paul Roetzer: Yes.
Carl Lewis: What are the biggest challenges people face trying to get their heads around this?
Paul Roetzer: The understanding piece must come first. Reading about it isn’t enough. I give 20-30 talks a year around the world about this stuff, sometimes to very technical, advanced audiences. Mostly, though, the audience is small and midsized business marketers who've never done anything like this. The best way to help people understand is to let them see it at work, like you said. The second is to give them a bunch of use cases. I’ve learned to say, "You'll use AI dozens of times today and not know it." Any time a machine (hardware and software) sees, hears, generates language, and understands – that's AI.
Paul Roetzer: When you unlock your phone with facial recognition, you’re teaching a machine to see and recognize a face. When you talk to Siri or Alexa, its ability to hear you, understand what you're asking, and generate a reply is AI. When Netflix recommends shows, Spotify recommends songs, or Google maps predicts how to get to point B faster because of an accident – that’s all possible because of artificial intelligence. In marketing and sales, we guide people to think about AI in everything we do as consumers. That's how marketing, sales, and business operations will be soon. It's going to get smarter, and the challenge is understanding that it's not hard and knowing where to use it to be smarter.
Carl Lewis: I'm a simple guy at the core, but over the last year, I noticed AI happening. I shop for musical equipment – guitar amplifiers, etc. And when I'm online, researching something, I see reminders that I looked at this guitar and am I ready to buy it?
Paul Roetzer: Yes.
Carl Lewis: Things like that happen all around us, and we become oblivious to some of it. How can listeners make these technologies work well consistently? How do you get your toe in the water?
Paul Roetzer: There’s uncertainty around definitions – sometimes AI and machine learning are considered the same thing. But artificial intelligence is the umbrella term. The best definition is from Demis Hassabis, the co-founder of Google: AI is the science of making machines smart. It’s an umbrella science, and machine learning is the primary subset. Machine learning is all about making predictions based on historical data. That base level understanding of what we're talking about is the key to figuring out how to apply it. The other key is that in every case that currently exists, it's built to do what's called a narrow task.
Paul Roetzer: There’s no general artificial intelligence where you buy a marketing platform and it does your email, social, advertising, etc. – you have to teach it. When you get started, you're looking for repetitive tasks that take a lot of time and energy or that would be more valuable with intelligent automation. Then you find a solution that's built to do that. Sometimes it’s in your existing technology – you may be working with a major tech platform company with built-in AI features. Start by turning on new features that do things smarter, then find vendors building things like email send-time optimization, media buying optimization, email subject line writing, or social media sharing. Those are individual tasks.
Carl Lewis: Let’s say I'm the owner of a business that employs 500 people. As an executive, how do I start understanding this?
Paul Roetzer: There are many great classes. We built the Marketing Artificial Intelligence Institute for that purpose: to make AI approachable for all business people, not just marketing and sales. Google and Amazon have free training. Microsoft has an AI Academy. Big tech companies built around AI need more educated buyers. They need people in every facet of business to demand smarter solutions. But to create that demand, they must educate the marketplace.
Paul Roetzer: They have free videos, downloadable reports, and entire sections of their sites dedicated to AI education. Another good starting point is a class on Coursera. It’s an online learning platform called “AI for Everyone.” It’s designed for business leaders, not technical people, and it’s an in-depth understanding of AI fundamentals. There are also three books I recommend: Human + Machine: Reimagining Work in the Age of AI, Prediction Machines: The Simple Economics of Artificial Intelligence, and The Algorithmic Leader: How to Be Smart When Machines Are Smarter Than You. They’re introductory books with case studies, use cases, and examples across many industries.
Carl Lewis: Those are great suggestions; thank you. But I’m going to change the subject. I ask all my guests these questions and get intriguing answers. Communication is the heart of business. With all the technology and how fast it changes, how do you do most of your business communication, and is it changing?
Paul Roetzer: Absolutely. Email is still the predominant way to communicate, but businesses are moving to online collaborative tools. We use Yammer in addition to email. The biggest change is conversational. People can go to websites and chat, either with a bot that's routing them based on what they're asking, or a human. The other element of conversational becomes voice – the ability to ask questions through intelligent assistance. That's a fast-emerging area of communication.
Carl Lewis: We’re very similar – email and Yammer. The benefit is that the consulting group has their own section, so I don't get bombarded with their stuff. You mentioned the new part of adding voice to electronic conversations. I wonder how long it will take us to add visual. It's out there, but it seems like we're a little slow on the uptake.
Paul Roetzer: It’s another important part of ours. We do so much through Zoom. There’s also GoToMeeting, etc. Five, ten years ago, whether it was because bandwidth wasn't good enough or the tech wasn't there, you didn't do much video conferencing. Whether it's remote employees, clients, podcasts, or whatever, there's almost always a video element to the communications.
Carl Lewis: More companies have a lot of distributed workers, so that’s a real plus.
Paul Roetzer: It's great.
Carl Lewis: Paul, you work and partner with a lot of critical third parties, customers, and vendors outside your company. What’s the most challenging part of working with those organizations?
Paul Roetzer: When we're working with vendors, we mainly do content creation management with HubSpot agencies. We manage the platform for them. But we don’t have some elements of marketing in-house, like video and web design work, so we used industry partners. The key with any partnership, whether you're hiring an agency or finding vendors, is trust. How do you assess someone before you've engaged with them, and how do you know they can deliver?
Paul Roetzer: And once you've established a relationship, how do you maintain that trust to where you know they're a valuable partner and can create the value you need? We look at many of our partners, vendors, and clients similarly – there's a mutual value exchange. That's the biggest thing for us: How do we build trust and then maintain and grow it?
Carl Lewis: You mentioned that AI is good at attacking repetitive tasks. Have you automated things in your business?
Paul Roetzer: Analytics reports. We would write analytics reports every month for clients about what happened on their website, with data from Google Analytics, HubSpot, and other marketing sources. Now we’ve devised a way through a natural language generation tool. It's not technically AI, just a smarter way to automate stuff. We trained the machine to write the reports. So, something that took six to eight hours per report takes about six minutes now. We still have to provide consulting and strategy, but the machine writes the actual narrative. We took years to train it, but it's now a template; you pick it and plug in the data, and it runs it and writes the report.
Paul Roetzer: We're about to launch a smart newsletter that’s written and distributed by AI. We're going to integrate a knowledge management solution that's a layer over online chat. It will enable people to discover content and answers more intelligently than a standard search or asking a chat bot. We also use transcription, which people rarely think of as AI, but anytime there's spoken words – like this podcast – machines can transcribe them. They’re getting so good that soon you might not need a human in the loop. There will always be typos and such things, but society will become more comfortable.
Paul Roetzer: For example, in consumer life. A few years ago, iPhone owners suddenly had transcribed voicemail. You’d go to your voicemail and not have to listen to it. That’s AI. That's a machine taking voice in and generating what it believes was said. We're constantly looking at things like text to speech, speech to text, image recognition, and content extraction. There are so many possibilities. In three to five years, 80% of what we do as an agency will be intelligently automated. I'm not saying it will be autonomous and we won't need marketers anymore – it will be machine-assisted. And like the analytics reports, we may save 90% of the time we used to spend on tasks.
Carl Lewis: I’ve been in the industry for twenty years, and I spoke to a customer who said – and this is the first time anyone's ever said this to me – that introducing a new software reduced his employee count. Normally, we just re-task those people to do other things; it sounds like that's what you expect to happen, too. It's fascinating. I admire your ability to break this down into what people are experiencing daily that’s AI at work, even if they don’t realize it.
Carl Lewis: It’s so common in daily life that we don't think of it as AI – just conveniences someone's delivering to us. And thanks for the suggestions about how to get started. I know people struggle to find a place to understand AI; they want to do it just themselves in a room where they don’t have to ask dumb questions.
Paul Roetzer: There are no dumb questions. No one understands this stuff. There may be people who talk like they do, but most people who can explain AI are PhDs and the engineers building it. And they struggle to make it approachable to the average person. So I came at it. I was a journalism major with a liberal arts degree. I’m not a scientist or an engineer. We worked for years to make this stuff understandable to ourselves. Now, when we explain it to others, we go back to the beginning of how it was for us and assume it doesn't mean anything to them either. We've spent years trying to make this make sense the way we finally understood it as nontechnical people in the AI world.
Carl Lewis: That reminds me of one of the greatest tasks I've ever thought about doing – surveying business leaders. Asking, “What did you study in college, if you went to college? Did you end up doing what you thought you’d do? If not, what do you do?” That would be fascinating because I studied to be a theologian and ended up in ERP and accounting.
Paul Roetzer: Oh my gosh.
Carl Lewis: My brother was a CPA, and we laugh because he sells real estate. It seems our early dreams are rarely where life takes us.
Paul Roetzer: I was premed. I went to college to be a doctor and ended up in journalism.
Carl Lewis: Thank you for sharing that because that’s fascinating to me. Well, Paul, thank you so much. This has been great. This discussion was meaningful to our listeners as they try to grapple with these topics, and I appreciate your ability to make it easily understood. That's not a common gift. It’s easy to make these podcasts too long, especially when we can converse as easily as you and I have. So, we'll say farewell today – and until we meet next time, everyone stay connected.