The Double-Edged Sword of Generative AI and Why it’s Part of a Broader Problem for Public Relations

Doug Downs (00:08):

Back in 2001, Dean Kamen promised the world a revolutionary personal transportation device. One it said would eventually render cars useless. The Segway, a motorized personal vehicle on two wheels beneath a platform the rider stands on while holding handlebars and controlled as the rider distributes their weight. Kamen decided to keep his new invention a secret from the world until it was ready to unveil, but he was so convinced his invention would change the world. He met with Jeff Bezos who said this invention was revolutionary. He met with Steve Jobs who said the invention was bigger than the PC and cities would be redesigned around it. The media started to catch wind of this, and the hype went into overdrive. Some people thought it was a teleportation device, a personal hovercraft or a backpack helicopter. There was so much hype that by the time it was unveiled, Diane Sawyer of a BC News said,

Diane Sawyer (01:10):

Because I'm tempted to say that's it, but that can't be it?

Doug Downs (01:18):

Despite the lead up hype, sales were terrible right away it was a safe form of personal transportation, but there were several high profile accidents as most riders just didn't know how to ride it. The Segways were expensive and impractical. The infrastructure and regulations needed for them to become a mass form of transportation just weren't there. Electric scooters came out in 2017 and those were just more practical, and the secrecy of the project didn't help either. They couldn't do any market research to find out what people wanted, so they couldn't tweak the project at all. Today, Kamen's master invention is fodder for Mall Cop jokes.

Paul Blart, Mall Cop (02:01):

In the event that you approach an assailant here's what I want you to do. You're going to pull up left hip forward, placing your right hand on your away hip, thus giving the illusion that you have a gun. Of course, we both know you don't, but you know what we do have? Our voices.

Doug Downs (02:27):

The company has now been sold twice, most recently to Chinese Company NineBot. The original two-wheeled segway is no longer in production. Despite the initial excitement, the segue was ultimately deemed impractical. Generative ai, while powerful also faces challenges such as ethical considerations, biases and outputs, and the fact it relies on pattern recognition and lacks contextual understanding means it can't strategize, but we will use it and maybe a lot today on stories and strategies, why generative AI might be part of a bigger problem in marketing and PR.

(03:20):

My name is Doug Downs. Off the top, we played the theme from Paul Blart, mall Cop, composed by Waddy Wachtel, the movie distributed by Sony Pictures releasing. My guest this week is Joe Zappa joining today from Brooklyn, New York. Hey Joe.

Joe Zappa (03:36):

Hey Doug. Thanks so much for having me.

Doug Downs (03:37):

How are things in Brooklyn? Up in Canada's Rocky Mountains, we've got, I'd say it's a few inches of snow right now. Not a ton, but we do have a little bit. You have any,

Joe Zappa (03:47):

Oh, we're doing better. No snow sun is shining today.

Doug Downs (03:50):

By the time this publishes, we're recording in December. By the time this publishes in early January, you will have the white picturesque and maybe not downtown Brooklyn, but you'll have it in the outskirts at the very least.

Joe Zappa (04:04):

Yeah, here our snow is covered in mud from people walking in the streets, so it's very picturesque.

Doug Downs (04:11):

Joe, you're the founder and CEO of Sharp Pen Media, a content marketing and PR agency serving adtech and MarTech, your former journalist and previously the editor of the MarTech Media site Street Fight. Okay. Chat, GPT or more broadly, really generative AI because lots of competitors out there, but more and more coming out, Google is really ramping up. How do you see generative AI affecting the balance between strategy and tactics in marketing and public relations?

Joe Zappa (04:47):

Yeah, well I'm glad you started with the strategy versus tactics question because I think that independent of generative AI is where a lot of PR and content marketing agencies go wrong. So I think often, even if the client doesn't say this, we're brought on to achieve a strategic outcome, but we actually start with tactics. So anyone who's been on either side of that, hiring agencies, being in agencies, how many times have you started engagement? And the first thing is like, alright, what channels are we doing? What stories, whom are we emailing? What reporters matter to you? That kind of thing. The problem with that approach is that you end up just regurgitating whatever the company is already saying on its website. You never take the time to strategically analyze what should the story be? Where is this company going, and how are PR and content actually going to help us achieve those outcomes?

(05:38):

So I like to think of strategy as involving three things. How does marketing support the business trajectory or the business objectives? What's the message and then where and how are we going to distribute the message? So then if you think about how Chat GPT or generative AI might play a role in PR and content, where do they come in? I think mainly in that last bit, which is where and how are we going to distribute the message specifically with how, right? It's about content production, but that comes third after you've already analyzed how is marketing supporting where the business wants to go and what is the message That is the critical thinking driven strategic work that I think really separates good agencies from poor agencies and good marketing from bad marketing, and I think the outsize attention or what I would call the outsize attention generated by Chat GPT and marketing over the last year, what it illuminates is that we are, as marketers generally too, focused on tactics and not sufficiently focused on the strategy that makes those tactics successful.

Doug Downs (06:51):

Why do you think that is? I agree with you often all have had companies come to me and say, we need a newsletter or we need a social media presence. We need an open house. If you work in public engagement, we need an open house. That used to be the thing in public engagement. Is it because they're such easy proof points? Look, I published the newsletter. Look, I held the event. Why do you think we end up so focused on tactics and events?

Joe Zappa (07:19):

Yeah, one reason for sure is measurement. So we're in the age of digital marketing. Everyone wants to be data-driven or pretend, and it's much easier to link leads generated by events or certain forms of owned content to deals so that the marketing team can go back to the COO or CFO whomever and say, yeah, well, we did this thing and then it produced a results. So aren't you happy and are we making good use of your money?

Doug Downs (07:52):

Yes. So proof points easy to show. We're a real busy shop, so how should companies integrate generative AI into their communication strategies without losing sight of the core narratives that hopefully they've built, and can we just create a new GPT that actually instructs us what those core narratives should be in advance? I've recently gotten into that creating my own GPTs for different things that I want to produce. The strategies are kind of just laid in there, including the keywords.

Joe Zappa (08:27):

I mean, I think it's mainly at present Chat GPT and similar tools are for research and inspiration. So if you want to find out about something, you've never written a blog post about location data. You say, write me a blog post about location data and X, Y, or Z, and it can do a pretty good job of educating you quickly on what that would look like and how someone in the industry might talk about that. But that doesn't tell you anything about how your company specifically should be talking about what your company's assets are as opposed to the assets of competitors. It doesn't tell you what your customers specifically care about, which not only can Chat GPT not tell you that, but often the client can't even tell you that, right? Often marketers are themselves divorced from what their customers actually think about them. So to me, all of, again, Chat GPT is useful for that tactical purpose. It can help you with efficiency, it can provide some ideas, jog your brain. It can be a little partner in dialogue for you, but it doesn't yet solve the strategic problems of coming up with what the message should be, how the message is going to help you meet broader business objectives.

(09:40):

But I don't know, maybe you've seen something different. It sounds like you are using it for strategic work.

Doug Downs (09:46):

Well, when you stop and think about it, I'm plugging in direction to the GPT so that it's regurgitating along the format that I've already as a human decided this is where I want this to go. It really, it truly is tactical. It feels strategic, but it is tactical and it's producing things at a faster pace. If it comes down to this is how I show non Comms leaders, my CEO or the VP of Human Resources, which tend to swallow up communications often within their portfolio, and it's just a completely different discipline, how do we demonstrate the value of our work? Then it comes down to the key performance indicators that we established to show no for that four hours. You're right, I didn't produce a thousand newsletters or whatever the case may be. I sat and thought strategically and often coupled with that, let's face it, the human habit as comms pros in the office or even working from home is we go to meetings all day. How are we supposed to find time to ideate and be creative and really provide that value if all we're ever doing is going to meetings and doing tactical things?

Joe Zappa (11:01):

Yeah, it's a good question because I think, let me just cut the BS and say that I think most teams do not know how to calculate this value. They don't have time to calculate it. There's no one, unless you're in a big company, even in large ad tech companies, I'll ask them, okay, well, we're talking about measurement. I'm like, do you have anyone on your team? They might have 20 people on their comms and marketing team, and I'll say, do you have anyone who is a data person whose job who has skills in that area and whose job is to figure out what's working? What does the data say? How can we optimize? There might be one person, and then usually what they'll tell me about that one person is they're so busy, they're getting so many requests from everyone across the marketing and sales organization that they don't even really have the time to assess the efficacy of content from a data-driven perspective.

(11:56):

In addition, when people say they are collecting data on the efficacy of their content, it can sound very intimidating and like, oh my God, they must be doing such sophisticated things. I need to do those things. And then you talk to 'em about it and it's like they're measuring white paper downloads and maybe if they're sophisticated, they're able to tie those downloads to specific sales opportunities so that they can then say, okay, well last quarter, these seven sales opportunities, they had downloaded a white paper at some point so we can say that the white papers are generating X in pipeline and then maybe X in deals, but it's a much less advanced science than I think people make it out to be.

Doug Downs (12:39):

Yeah, absolutely. I was just reading Reagan's daily newsletter before you and I joined the recording, and Alison Carter article, Alison's been on the podcast, wonderful, wonderful journalist talking about prompt engineers and how this will be one of the new waves in 2024, probably right through 2025, is to hire the prompt engineer, the person who knows how to load Chat GPT. Let's face it, that's one of the primary tools they're going to be using and that these will be easily six figure jobs amongst successful companies. I guess what I am asking is, is that the right approach? Should that be one of the higher paying jobs in the industry when really the output is mainly tactics?

Joe Zappa (13:27):

Well, I think the first thing is that the strategy is indispensable, right? So you can't have those prompt engineers if they're not also creating the strategy or if you don't have someone creating the strategy beforehand, let's be generous and say that the companies hiring the prompt engineers have the strategy taken care of, and then it is a question of tactics. I mean, my reply is essentially, that sounds fine. You do it the way you want to do it. If you want to have a content marketing manager like old school, a real human being who knows the industry and has writing chops and it is going to write all that content, then that's great. That'll probably be especially effective if it's like high consideration B2B and it's not a massive company and volume isn't the play. If volume is the play and you need to be creating dozens of articles a month, then maybe that is the best way forward.

(14:15):

Maybe you make sure you have the strategy and then you have someone who is a writer and editor who will work with generative AI tools to create the amount of content you need, and maybe you have one person doing that, whereas in the past you would've had three to five people. I don't know. I don't know if we're quite there yet. I think probably my guess would be the content is worse, is it? But what companies are really going to be wondering is what do I really need this content for? And even if it's worse, let's say it's 20% worse, if you can save two people from a headcount perspective, maybe it's worth it to you to have content that's 20% worse. I think that really comes down to is it like an old school SEO thing where you're just trying to get pages on the internet and people through the door, or is it thought leadership where it's like you're really winning by building credibility and differentiating your brand

Doug Downs (15:14):

Really good, and there is no clear answer to that. The more futuristic approach, the SEO example, the more futuristic approach is to look at, is it the behavioral qualifiers for what's driving SEO rather than, Hey, I got these keywords. I produce transcripts on my website for every episode, and part of that is I'm hoping the words that you and I say are recognized by Google. So I guess what I'm saying here is it's not black and white. That one approach for one company could be correct and a different approach for another company could also be correct. Yeah.

Joe Zappa (15:51):

Well, and that's the other thing that always gets lost when shiny objects like generative AI pop up and everyone has vociferous debates over the future of marketing and these objects roll in it, people start to make grand pronouncements, right? It's like SEO is dead or fire your marketing team or anyone who's using generative AI doesn't understand marketing at all, and it's like, of course the reality is more complicated. It depends on how does marketing work for your business, but in a way that just brings us back to question one, right? That the way you're going to use generative AI and similar tools is going to depend on why you're doing content in the first place. What's the business objective?

Doug Downs (16:37):

Okay. Tough question here. I know your company, Sharp Pen Media, it offers its clients narrative building, editorial planning, content development, all the things that you're speaking highly of. I think you're onto something here and you and I share a similar background in that we're both journalists, so all questions are good because there should be answers to all questions. I think you're onto something here about generative ai, but others listening might conclude, well, he feels a potential threat from ai, so that's why he's saying this. I guess I'm putting you on the spot here, Joe.

Joe Zappa (17:13):

I think there are two different ways to think about this. So the first thing is we have strategic clients and we have tactical clients, and some are both. I don't think AI is at the point where it's going to be replacing strategy. I think even when AI has become very sophisticated, there will still be a more limited number of human professionals who are setting the strategy and directing the machines, and we do play that role for a lot of our clients. So there's that bit, right? The strategy thing I think is strategic marketing services are more insulated from disruption by ai. Then there's the tactical stuff, and that goes back to what I was saying before. Are we doing really highly nuanced forms of content creation that are meant to differentiate B2B brands who have commoditized technology and really need a sophisticated narrative and sophisticated spokespeople to captivate their customers and help them build credibility?

(18:15):

Well, then we are more insulated as marketing service providers from ai. Are you just an SEO content agency and all you're doing is writing what is content marketing type blog posts and that event, which isn't really what we do and that event, I think you are much more vulnerable to disruption by generative ai, though even there, I will add a caveat, which is I've discussed this with SEO people and what they will say, which makes sense to me, is like, okay, on its face, it might seem like SEO is going to be easily disrupted by ai, but what if you have three companies in a category and they're all using similar generative AI tools to create content? Doesn't that just once again lead to a homogenization problem where each company needs to figure out how it's going to transcend the status quo, presumably by doing something more than relying on ai.

(19:12):

So even in that case, I think there will be a role for human content creators, and I think more than anything, what will distinguish the companies who are slashing their content marketing spend and replacing it with AI from those who continue working with human experts is what is their view of marketing and how do they view its role within their organization? Because if they respect marketing and they think this is a complex thing that's going to have a big impact on our business, they're probably going to continue to hire high value experts. Others, the kind who were offshoring it and using Upwork and all that. They might be the first ones to cut it and not to go back to it.

Doug Downs (19:56):

So good. I really appreciate your time today, Joe.

Joe Zappa (19:59):

Yeah, thank you so much for having me. This was great.

Doug Downs (20:03):

If you'd like to send a message to my guest, Joe Zappa, we've got his contact information in the show notes, stories and strategies is a co-production of JGR communications and Stories and Strategies podcasts. If you'd like this episode, do us a favor, share it with one friend. We're also on YouTube and YouTube music. You can check us out there, full audio episodes and we publish short video segments as well, so just search for us on YouTube. Thanks for listening.