How can you shape your AI brand footprint?

Farra Trompeter: Welcome to the Smart Communications Podcast. This is Farra Trompeter, co-director and worker-owner at Big Duck. In today’s episode, we’re going to ask the question, how can you shape your AI brand footprint? And if you’re not sure what that term means, don’t worry, you’re going to learn all about it in just a few minutes. I am delighted to be joined by George Weiner. George uses he/him and has actually been on the podcast before. You may remember him talking with Sarah, our founder, on what data really matters back in 2019. And if you don’t remember, we’ll be sure to link to it, and you can listen to it all over again. But George is one of the most enthusiastic nonprofit tech evangelists I know, and I’m excited for today. George runs Whole Whale, a leading marketing and analytics agency, specializing in data-driven strategies for nonprofits and social impact organizations. Before Whole Whale, he was the CTO of DoSomething.org and co-founded PowerPoetry.org. At Whole Whale, he has developed products like CauseWriter.ai that help nonprofits safely create AI content in their voice, which is quite apropos of today’s conversation. George, welcome to the show.

George Weiner: Longtime listener. One-time guest, and excited to be a caller today.

Farra Trompeter: There you go, back on it, back on it. Well, a few months ago, you published a blog that caught my attention. It was called Measuring your AI brand footprint: The hidden visibility challenge. Of course, that phrase “AI brand footprint” intrigued me immediately. And I know, and I believe actually, that your team made that up. So, did you make that up, in fact, and what does it mean?

George Weiner: We, yes, did make up the concept because we needed some white space to put words behind what we’re seeing in the data and in practice for how these LLMs (large language models), these AI chats are pulling information and to put it nicely, pulling information from our sites and then representing them as, you know, as facts and information for people that are writing. I’m curious for you, though, because on the brand side, y’all are always paying attention to that client narrative. Like, what alarm bells went off when you saw that?

Farra Trompeter: Well, I think the question is like, what does an AI brand mean, right? Do we have to have a separate brand for AI? How do we manage how our brand shows up in AI? So that was what was intriguing me to want to learn more about.

George Weiner: Yeah, and it’s tricky because in the sort of, you know, pre-AI period of time, we could cultivate our brand a lot more easily, right? We have our polished materials, we have our “About” of our site, we have the, you know, three tent poles of internet placement where we talk about ourselves. Now, what we’re dealing with is this blend of interpretation that is being done by the AI that curates, pulls in, and then summarizes all of it and all of what it finds about your brand, how it’s talked about, and where it’s talked about. And this requires a bit more cultivation and intentional work around it. What’s more, your AI brand footprint isn’t just about Whole Whale, isn’t just about Big Duck, it’s about the concepts and areas where you are expert, where you talk about. And that kind of leads us back to: Why carve out some clear space for AI brand footprint to talk about it? Well, as it develops as a term, we hope that our fingerprint is on there. Our way of thinking is in there. And that kind of extends the concept of an AI brand footprint because it is an entire ecosystem inside of this large language model that it’s pulling from. And then that gets also augmented by internet search results. But it’s this giant soup that we’re now trying to play with, away from somebody may not even ever meant like end up on our website, right? And still find us.

Farra Trompeter: Yeah, in fact, I think I was telling you before we recorded this show that a few weeks ago I got a call from somebody, and we always try to find out “How’d you find Big Duck?” And usually it’s, “Oh, a friend of mine worked with you all had a great thing to say,” or “I stumbled on a blog you wrote,” or “I listened to a great podcast with George,” you know, I’m sure we’ll be hearing that in the future. But this person said, “I actually did a search on ChatGPT about leading nonprofit communications firms, and I found Big Duck,” and I was like, “Okay, guess we’re showing up in the right places.” So, that was exciting to hear. It’s the first time I’ve heard it. I’m sure it’s not the last.

George Weiner: The implications of that are very, very powerful and are not to be overlooked. We had a similar case where folks reached out to us, and it was a, “We found you on ChatGPT and decided to reach out as we went through that process.” We are in a very great moment in some respects to have had both of us, Big Duck and Whole Whale, some brand reputation over a decade of work. The links, the types of online reputation that now get baked into that, it’s going to be so much harder I think for new entrants to suddenly get into that flow that the, the building of your (see previous comment), AI brand and AI brand footprint is really going to have a tremendous amount of value. I don’t know how this will change over time, but right now we’re paying very close attention to where we show up, where we can show up with our competitors, and there’s a whole bunch of analysis that you can do on that. And by the way, it depends on the AI model that you are using and in what context. So it is not like there’s like one central like Eye in the Sky database. You’re playing this game on many fronts.

Farra Trompeter: Well, let’s talk about the front of Google for a minute, which I believe is still the most popular search engine. And I know I have seen that AI search results are often the first thing that someone sees when they enter a search on Google. What does that mean for traditional approaches to search engine optimization? What should organizations do to leverage this shift where AI and the large learning models are sort of coming into our search results?

George Weiner: We’ve been doing something with our clients called a large language model audit, an LLM audit for the content. And what we’re seeing very, very real, is the decrease in organic traffic, that traffic that people used to type into Google, which commands roughly 89% of the US search market. And that decrease is very real and it’s continuing to happen as AI answers the questions that people were asking: That was How Tos; those facts about; that article that probably did pretty well before that was maybe, you know, adjacent to your mission but did very well is now getting the axe from Google and also because of other things like ChatGPT as a search or Perplexity as a search. So, I think doing more of what you used to do is a losing strategy, and that’s where we are moving into, how you know, the AI brand footprint, but also into LLM optimization. It also goes by many other names, as people love inventing things like GEO, which is Generative Engine Optimization, and then a rebranding of SEO, which is Search Everywhere Optimization. It is all the same flavor of ice cream, though. We need to rethink and measure what we’re creating and how we’re creating it.

Farra Trompeter: Well, let’s talk more about that measure piece. You flagged in your blog post how analytics tools are often not capturing AI-generated impressions or mentions. Why is that happening, and what should people do about it?

George Weiner: It’s not happening because we’re not being given those data for a number of reasons. So, put another way, when Big Duck showed up in that ChatGPT search, you didn’t get some sort of dashboard. Your brand has been mentioned this many times. You’re not even getting that in the AI overview, which is what Google calls the AI answer that happens for some X percent of searches, X is going to increase, by the way, every single month. You’re not getting that yet. There are ways though, however, of looking at referral traffic. We talked about organic traffic, but referral traffic is another channel of inbound people that click on stuff and end up on your site that you can find in Google Analytics. You can look at AI-based sources of referral traffic and then bundle those together, kind of like putting them in a shopping bag and branding that as a whole separate channel.

George Weiner: And so we are doing that with all of our clients. Now, you can ask some interesting questions, what pages of our site are people coming to? And also, we can reverse engineer. We can just say, “Hey, I think if one person clicked from the source ChatGPT, there was probably 10 others that did not click.” What is my AI brand footprint? What and how much is my brand being mentioned, given the fact that I only get 10%, 5% of those as clicks to my site. And then you can begin to play a more accurate game.

Farra Trompeter: Well, this is great. I want to pivot up and away for just a moment because I know there’s so much you have to share when it comes to AI. Can you offer any ideas, tips, tricks about what nonprofit staff can do to proactively shape their brand footprint in AI tools?

George Weiner: Yeah, I don’t want it to be the doom and gloom, like your traffic is going to go down. What’s going to happen is your traffic is going to become more relevant and the tactics that used to work for SEO traditionally, which is the, you know, write this How To… article, write this tent pole explainer on everything you need to know about hunger, everything you need to know about animal adoption, like is not going to work because AI can answer it. Put another way. If your article can be answered as good as not even better, just as good as AI, don’t write it. If you’re using AI to generate the entire article, you’re wasting your time. So if your classic style of writing is being commoditized, being done as good as or better by AI, where do you have to move?

George Weiner: Well, you have to move to first-party information. What are the things that Farra has to say? What is the quotable thing that AI can’t make up? Or if it does, it is actually, you know, illegal, it is now guard-railed against. So, “According to whom?” I like asking that question of my team, of our clients, for whatever article we’re talking about. “According to whom is this true?” And what insight does that human, who’s quotable and you should quote and put that in there, because, then again, you’re building into the AI brand footprint. What is there? The other thing that you have access to is first-party data. What data have you surveyed, collected, or found in your backyard that you can package and graph and put out there as “Here’s what we saw. It is a small snapshot. Now, mind you, we didn’t do the double-blind longitudinal study for a decade, but like, calm down, here’s the information we found and here’s what we learned from it.” AI can’t fake that, and actually AI will reference that.

George Weiner: So I think you take the energy that used to be around, “Let me build the, you know, ultimate guide to adopting Mittens the kitten” and you come back and say, “Hey, here are stats about the five reasons why people bring back dogs, and now here are our data, and now we have authority according to our head of department.” So hopefully that gets you thinking about where to put that energy. But I am growing increasingly concerned about classic SEO strategies and putting content writers to work on a plan they’re going to lose on.

Farra Trompeter: Yeah. Well, this sounds like a good evolution of content strategy, and I want to go back to your comments around the sort of idea of authority and quotables. I know that some people are worried that AI will start taking their organization’s intellectual property. Is that a well-founded concern? What can people do if they feel like they are being stolen from reactively, and what can they do now more proactively?

George Weiner: The answer is yes, your content is being stolen. I’ve actually proved it. I’ve done things that I have, you know, screenshotted, and here’s the truth, like I’ve written concepts, I’ve created them, and AI has once it does its snapshot, ingested it, and not credited us. So I’m really beginning to work on how do I make sure that in the, you know, everything down to the semantic markup of how I’m writing about content, I speak in a language that is clear that, as it is ingested, it is credited back to us. And that is something that is at a bit of an evolving field, but it is very real. The things that you put on your website are and have been snap-shotted and ingested and maybe not properly attributed to you. That is different than saying the AI overview. The AI overviews, right, this is the addition of links and search results into the AI, answering a question is more likely to attribute in its footnotes links to you. Like that absolutely is happening. But there’s a lot of different flavors of this, and many folks are using ChatGPT cold without search, which means you’re dealing with that snapshot of information that pulled in your data and is then sort of summarizing and you know, attributing at whim.

Farra Trompeter: So if you find out that’s happened, are you saying this is our new reality and just be aware of it and as much as possible write content, you know, according to you, and sort of adding all of that in? Or is there something people can do if they find out their content is being stolen?

George Weiner: Yeah, so it depends on the size of organization. And let’s be clear, if you don’t have an army of lawyers at your beck and call like you have a problem. Here’s what I’d recommend for the majority probably of organizations that are like, “Yeah, we’re not The New York Times and we can’t afford a multi-year lawsuit.” I have been working on a page of intellectual property that is owned by Whole Whale, has been used at this date in this way. I would recommend you look at your intellectual property that is already publicly available because it’s already been ingested, there’s no sort of toothpasting it back into the tube. So I would say, what does your intellectual property page look like? How are you linking to it? And you know, that might be one place to start because it is beyond me to say that like a nonprofit operating at under $1 million, which 75% of nonprofits do, can fight that battle. Surely you can submit these things into claims and comments, like I challenge you to get somebody on the phone at ChatGPT. Seriously challenge: like, go do it.

Farra Trompeter: Hard to get anybody on the phone anyway these days. But that’s another story.

George Weiner: Kids today don’t know how to use the phone.

Farra Trompeter: Kids today. “What are these rotary phones you used to use?” Well, let’s leave on a high note and talk about more of what people can do. So you’ve got a great suite of tools on all different things folks can do with artificial intelligence that you created for free for nonprofits at causewriter.ai/resources. Can you share a little bit? We hope people will go there. You can go, there’ll be links to the blog post George mentioned and a few other resources at bigduck.com/insights. But George, let’s go back to those resources that you’ve published. What are some examples of the tools there, and how people can use them?

George Weiner: Yeah, again, those are free tools because we want more good words in the world at Causewriter, and so those are free to use. Things like donor persona builder, right? You can go in, put in your organization, and it’s going to build out donor personas that you could then chat with. There are other things, like 990 analyzers. You put in a 990, it comes back with key information from it. There’s also things in there like a prompt builder, because a lot of what we need to do is build a prompt, right? The instructions we give the AI so it does what we hope it to do. There is a prompt builder that uses our approach that will help you create like, “Oh, I want grant proposals for my animal welfare organization. Here are some details.” It will then create a prompt that you can then use on tools like Claude, or ChatGPT, or Gemini, or other places that you are using tools. So we are constantly playing with and exploring, and we want you to think about what is possible with AI tools that have been a bit better tuned than the generalists you are handed out of the box at ChatGPT.

Farra Trompeter: Well, I love that you provide that. And if you’re out there and you’d like to learn more about George’s work or access the amazing resources his team continues to share and provide, go to wholewhale.com. George, you’ve given us so much to think about, but any other parting words of advice you’d like to share with our listeners?

George Weiner: Fear or love? When you look at your team, I think you’re being met with both of those as motivators, you deal with either differently. I’d encourage you to have, if you don’t already, an AI policy that unpacks internally and externally how you’re approaching these tools. And then internally, this sort of culture, hopefully of love and exploration rather than fear, “This is going to take my job, and I’m going to find reasons not to use it.”

Farra Trompeter: Well, let’s let love overcome all. Thank you, George. Have a great day.