Lifting the Fog on AI | Katie Robbert
10.31.23
Ann Kraus: Hello and welcome to Simpli.fi TV. I'm Ann Kraus. Our guest today is Katie Robbert, she is the CEO of Trust Insights. Katie is an authority on compliance, governance and change management. She oversees the growth of CEO Insights, manages operations and product commercialization, and sets overall strategy. Her expertise includes strategic planning, marketing, operations management, organizational behavior, market research, and analysis. Prior to co-founding Trust Insights, she built and grew multimillion dollar lines of business in the marketing technology, pharmaceutical, and healthcare industries. Katie, welcome to Simpli.fi TV. Katie Robbert: Thank you so much for having me. I purposely made sure that my bio is a tongue twister for everybody so that we can maybe try to skip over it and just get into the conversation. So when people look at it, they're like, "Yeah, no, I'm good." Ann Kraus: Yeah, yeah, well, yes you did. You did well on that one. I do want to talk about AI and I want to talk about AI with marketers, specifically how it's used in agencies. But before we get into that, I'm going to ask you to share the acronym that you have for the different types of AI FOG, OG. Can you explain that please? Katie Robbert: Absolutely. My co-founder, Chris Penn, he's a data scientist, and when I talk to him, we often get into the very technical weeds. And so he's told me multiple times that generative AI is not the only kind of artificial intelligence, that it's the other two ones that I can never remember. And so I was like, "Is there an easier way for someone who isn't a data scientist, someone who isn't technology-first to remember so that you can have an intelligent conversation about artificial intelligence and still know which pieces you're talking about?" And so for the life of me, I can't remember what the proper terms are, which is why I created FOG, which is also sort of a riff on my foggy brain. I can't remember what they are. So FOG is Find, Organize, and Generate. And so the three parts of artificial intelligence are, first, the AI needs to find data. And so this is where when we talk about the large language models, what are they trained on? They have to find information. And so systems like ChatGPT are trained on the internet. And so it found information from the internet and said, "Great, tell me all of the things that you know. Let me, O, organize them into things that I understand." So conversations about Katie Robbert go over here and conversations about Ann go over here and conversations about Trust Insights go over here, so let me organize and categorize these things. And then the last is G, generate. So now that I've found the information, I've organized it, the user's going to ask me a question and then I can go into all of my buckets of information that I've organized and give them an answer. I can generate information for them, which is where FOG came from. Because again, I can't remember what the proper terms are for the life of me. There's a lot of things I remember, old song lyrics, Miss Mary Mack from being a kid, choreography to NSYNC, but I don't remember the three parts of artificial intelligence. Ann Kraus: That's awesome. I love NSYNC. So when it comes to AI and when it comes to utilizing FOG and agencies and everything, where do companies and agencies in particular, where do they go wrong when they're introducing new technologies like AI? Katie Robbert: They go wrong 100% of the time by picking the tool first. So they pick a solution before they really understand the problem that they're trying to solve. And this is where I've developed what I call the 5P framework, which is purpose, people, process, platform, and performance. Because what marketer doesn't love, alliteration. Again, something easy to remember. But by walking through that framework, the goal is to understand the problem that you're trying to solve first. And so that is 100% of the time where agencies and companies go wrong when it comes to the conversation about artificial intelligence. And so, this whole calendar year, the conversations have been overrun with AI and these tools and these startups and these shortcuts, and "Will I lose my job?" But the conversation very rarely goes to, "What problem do I have within my team, within my company that AI could be beneficial to?" It's usually, "Hey, here's a tool, let's bring it in and then retrofit it into our processes and our culture." And so it really needs to be the reverse. We need to first figure out what is our mission, vision, values, what is our company culture, where are our teams, the people, where are they struggling with the processes that they've built. Do we even have them built? Are they documented? What platforms currently exist in our tech stack that could benefit from having yet one more tool for our teams to master? And then what is our performance outcome? How do we even know we were successful? Is it that we spent money, we were more innovative? How do we measure innovation? And so there's a lot of things that have to happen before you even get your hands on a tool. Ann Kraus: That's interesting because it is right now the talk of, "Is it going to take my job?" We've all gone into doom and gloom rather than into, "What could this possibly do for us?" Katie Robbert: It's a relevant conversation, but it's also a premature conversation. Ann Kraus: Okay, interesting. All right. So what are some of the suggestions that you have for agencies to use AI? What do you suggest they do with it? Katie Robbert: So there's really good use cases if you're trying to get up to speed on it. The tools that are out there, they're really good at transcription, they're really good at summarization. One of the newer use cases that has become a little bit more publicly available is the ability to actually do some analysis. And so, a lot of us are using tools like AdWords and Google Analytics. We're putting everything into sheets and we're putting everything into dashboards. But then it takes a lot of human computing time in order to make sense of the data. And something that these tools are starting to be able to do really well is you can give the tool a dataset, give them some context around it with your prompt and say, "What do I need to know? What is this telling me?" Now with the caveat, with the disclaimer that it's only going to take you so far because the human understand nuance of why you made a decision. So one of our clients, for example, they have a seasonality where everything kind of drops in August. Unless we're telling the machine, the AI, that this is part of the analysis, it's not going to know, it's not just going to assume. So it's a powerful tool to optimize your process, but it's never going to fully replace you. Ann Kraus: When it comes to AI and your 5P framework, if you could give us those five again, how does that framework, the 5P framework, fit into our discussion of AI? Katie Robbert: The 5Ps are purpose, what is the question you're trying to answer? People. Who needs to be involved? Who ultimately benefits from this information? Process. How are you doing the thing? Platform. What are you using to answer the question? And then performance. Did you answer the question? How do you know you were successful? And so it's a fairly straightforward framework. You can expand it to be really in depth or you can literally just answer those five questions. And what I found is even using it in its most simplistic form of just who, what, where, when, how, can help you sort of reframe and slow down the conversation about artificial intelligence of like, "Well, wait a second. If I introduce ChatGPT, who on my team needs to know how to use it? Who even needs to be able to access it? Do I have enough licenses or tokens? Do I even know what the process is that they would be undergoing to use it to be successful? What are they using it for? Are they just playing with it and it's just research and development, or is it actually replacing some of our content generation?" And so using the 5Ps, I found, is a way to slow people down and think things through a bit more of you're introducing ChatGPT or any other generative AI tool, what is the problem you're trying to solve by introducing this tool? And a lot of times it's, "Well, I want to be competitive." Okay, competitive to who? How so? What does that mean? How many pieces of content do you want to generate? Where do you feel you're falling down on your competitors? And so you start to really unravel the why before you get to the what. Ann Kraus: That's really interesting because I know that we have all seen this shiny new tool, just like we saw with the internet in general, or with a mobile phone or with, excuse me, social media. And now this is that shiny new tool. So we're all like, "Oh, cool, what can it do?" And you're kind of asking us to, "Hey, let's take a break here folks, and let's see how we're supposed to be using this." Katie Robbert: Yeah, it's a protection against resources, your money, your people. AI is a culture shift and it needs to be treated as such. Ann Kraus: Yeah, that's a very good point, culture shift. Okay, that's awesome. I don't think that you have come into all of this lightly in what you're talking about. So is there a podcast, do you do a lot of reading or listening to podcasts? Is there anything you want to share with our viewers? Katie Robbert: It's interesting. The stuff that I tend to listen to is less technical. So I'm a big fan of Everybody Writes by Ann Handley, versions 1 and 2, because what I really like about her material is it's about storytelling and engaging your audience. And so as you're having these conversations, you really need to understand your audience and how to tell the story in a way that they're going to be open and receptive to it. So if you're teaching something very technical, how do you wrap a story around it so that it resonates? And so I'm a big fan of Ann. Ann Kraus: That makes a lot of sense, especially after what you were telling us about with the 5P's that that would wrap into a story. And if anybody wanted to actually get to chat with you or wanted to check up on anything else that you were doing, what's a good way for people to learn more about you? Katie Robbert: You can visit our website. It's trustinsights.ai. Pretty straightforward. We have all of our contact information there, all of our services, white papers blog, we run a free Slack community so you can find all of that information on our website. Joining our free Slack community is probably the best way to actually interact with my team on a daily basis. It's free. Anybody can join. Ann Kraus: That's awesome. Okay, great. Thank you, Katie. This was a great conversation. I really appreciate you being my guest on Simpli.fi TV. Katie Robbert: Thank you. Ann Kraus: And thank you all for watching. Simpli.fi TV is sponsored by Simplifi, helping you to maximize relevance and multiply results with our industry leading, media buying, and workflow solutions. For more information, go to Simpli.fi. Thank you so much for joining us. I'm Ann Kraus, and I will see you next time.
More Simpli.fi TV Interviews