Building AI-Enabled Agencies with Greg Shove

Drew McLellan [00:00:37]: 

Hey everybody, Drew McLellan here with another episode of Build a Better Agency. Welcome back. If you’re a regular listener, if you are new, you picked a great one to join us. This is going to be a great conversation and a very different perspective on some of the AI conversations that we’ve had of late. So before I tell you about our guest, for those of you that are anxious and interested in elevating your agency’s AI knowledge, the ability to talk to clients about it, the ability to talk to teams about it, the ability to use it to improve the operations of your agency, improve the work you do for clients, all of that sort of thing. If that’s, if you’re hungry for that, remember we have a peer group that is focused on that. Every month they get together for 90 minutes to talk amongst themselves about the AI tools they’re using, the experiments they’re doing. They’re doing experiments together, but best of all, they’re building a repository of all of the prompts and all the things they’re building out in their experiments to share with one another. So if you’re curious about that, head over to the Agency Management Institute website and under memberships, look for the CIO Innovation peer group and you can learn more about it. All right, let me tell you a little bit about our guests. So Greg Shove is a serial entrepreneur. He has started and sold many businesses which I’m sure he will tell us about as he introduces himself. But over the last few years, he’s been very focused on his latest business, which is helping organizations just like yours figure out how to wrap their arms around AI and how to take advantage of all of the tools, technologies, advantages that I can give us. And more importantly, how do we talk about that with clients and how do we innovate our own business based on what’s possible? So I’m really excited about this conversation. It’s going to be a lot of blue sky, blue ocean blue everything, opportunity, conversation. And I’m hoping it really gets you thinking differently about sort of how you’re approaching AI in the short term, but also thinking about it more in the long term. All right. All right. With that, let’s welcome him to the show. Greg, welcome to the podcast. Thanks for joining us. 

Greg Shove [00:02:56]: 

Thanks, Drew, for inviting me. 

Drew McLellan [00:02:57]: 

You bet. Give everybody a little sense of your background and how you have kind of come to have the knowledge that you have today before I start peppering you with questions. 

Greg Shove [00:03:05]: 

Sure. Seven time entrepreneur, came from Canada to the US 35 years ago to go to business school. Stayed. And because someone told me in Toronto, where I was living, that if I really wanted to be in the technology business, I should go to a place called Palo Alto and a grocery store called Molly Stones and I’d meet more people on a Saturday morning doing my groceries at Molly Stones who are in the tech business than I’d meet all year in Canada. And I thought that was good advice. So I came into Palo Alto. 

Drew McLellan [00:03:32]: 

Was it, was it true? 

Greg Shove [00:03:33]: 

Yeah, it was absolutely true. I mean, I started, I got lucky. I snuck into Stanford Business School. So that gave me a good start in terms of a network and. And that was sort of the beginning of the Internet. So that’s how old I am. I’m 63. I type with two fingers. So the good news is, if I could figure out AI, anybody in your audience can. I’m not a technical person at all, but yes, seven startups, some exits, some. Some wipeouts and, you know, and some stuff kind of like TBD we haven’t figured out yet. Is that a success or a failure? This company section, venture backed. So my most recent startup, I’m the CEO, and I pivoted the business on February 1, 2023. So that was the day I played with ChatGPT plus for a few hours and thought, okay, this changes everything as it relates to knowledge, work. And I needed to pivot myself. I needed to pivot my business and kind of head in that direction. Not easy, by the way, both personally and for the team. I think the team thought I was crazy. Yeah, I bet they got sick of all the text messages and slacks and emails about AI. Again. Two years ago, it wasn’t as clear as it is now what’s happening, but it’s proving to be a good move, obviously. 

Drew McLellan [00:04:39]: 

So the company that you have now, what is the main focus of the work you do today? 

Greg Shove [00:04:45]: 

Today, the main focus is basically helping organizations become AI enabled. So almost every organization is full of people who are AI novices and likely anxious. And so the only way to leverage this and get an roi, I think, is to move everybody from that state, anxious and novice, to confident and proficient. And so my Our opinion is the faster you do that, you get people to be not necessarily AI native. I think the kids coming out of high school and college are truly AI native. But for the rest of us, I think this is going to be our first dance with AI. So let’s get it to happen quickly and let’s do it well. 

Drew McLellan [00:05:23]: 

Yeah, let’s get to competent. Right? 

Greg Shove [00:05:25]: 

Yeah. Let’s get confident and let’s be concerned around how jobs will change and how we’ll deliver value to clients or, you know, how our services or products will change. You know, let’s not, but let’s not be freaked out about it. You know, let’s be thoughtful, concerned, you know, thinking about it, but let’s not be too anxious about it. You know, there’s going to be a lot of change. It’s already happening for sure. A ton of opportunity at the same time, which, you know, we need to, we need to figure it out. 

Drew McLellan [00:05:50]: 

So when clients come to you, have they gotten over the hump of we’re ready to embrace this or do you, are. Do you still meet inside the organization? A lot of resistance. So I was just on a webinar today, I was telling you before we hit the record button about the agency edge study we just did. So we just did the researcher and I just did a webinar with, for our clients, just walking them through it and somebody said, you know, my employees all are convinced that AI is going to take their job. So are you bumping into that as well? Or is that unique to our industry in that everyone takes such pride in creating, you know, that we’re creating bespoke products and services and now, now machine learning or a tool is going to come in and make me irrelevant. 

Greg Shove [00:06:37]: 

Yeah. So maybe a two part question. I’ll do the first part first. Which is. So our, our clients are coming to us in one of three or four states, one of which is they’ve already deployed AI into the organization and it failed. And we can talk about why that happens. It’s usually because they deploy it like software and AI is not software or they’re about to deploy and they’re realizing, okay, I, you know, I need help. This, this is different than, than deploying an ERP or a marketing automation software or, you know, canva. This is something quite different than that. So I need help. So they call us or they’re trying to get the courage, you know, to deploy. So they’re, you know, they need maybe more handholding, more advice, more sort of executive support. So we’ll get Involved early, typically with executives, even boards. Now we’re getting invited, you know, into board meetings to say, listen, can you help kickstart this? So, you know, kind of all the above in terms of different states of adoption, I guess, or deployment. So that’s the first part of the answer. Second part is, yeah, everybody’s facing the same issue. Workforce anxiety is at all time highs and not surprising for a lot of reasons. 

Drew McLellan [00:07:40]: 

But AI being one of them, right? 

Greg Shove [00:07:41]: 

Yeah, AI is only one of them, but it’s the latest and it’s not insignificant. And you know, our employees aren’t stupid. So no matter what we say or if we say nothing, which is worse, in my opinion, about the impact of AI, it’s employees will fill in the gaps. And humans often fill in the gaps with, you know, the worst case scenario, not the best case scenario. 

Drew McLellan [00:07:59]: 

Right, right. 

Greg Shove [00:08:00]: 

We catastrophize. And the media doesn’t help. For the last two years, the media has run these wildly inaccurate stories and, you know, sort of narratives around, you know, AI will get us the cure for cancer in five years and, or AI is coming for all of us and we’re all going to be unemployed picking up UBI checks, you know, like, neither are true, but it’s what gets clicks if you run a media business. So, you know, you write the sensational stor. So it’s just like everybody’s talking about how Waymo’s are being set on fire, but people aren’t talking about how, you know, Waymo’s taken 16% market share in San Francisco from Uber and Lyft, and women riding at night prefer Waymo than Lyft and Uber. Makes sense, right? So we gotta get to the, to the reality here. The reality here is it will impact some jobs, but not as many as we think. And mostly jobs will be changed as opposed to impacted, meaning, you know, removed. And, yeah, and, and by the way, a lot of value is going to be created because the people that really figure this out will see the opportunities from it. And by the way, when you actually start using AI, you realize, well, it’s pretty good, but it makes a lot of mistakes. So it’s not taking my job anytime soon. I tell novices the best reason to move from being a novice to being an expert with AI is you’ll realize there’s no threat from AI because it’s not that good yet. It’s going to get better for sure. I think anxiety is high. I get often asked from executives, what should I tell the team we’re going to deploy AI internally. What should we say? Here’s what I say, Drew, and I think people should come up with an answer. And it should be as honest and transparent as you can be, because your employees aren’t dumb. Here’s what I say, Drew, because we have a small company, we have 35 people at section. We’re essentially an L and D, a learning and development agency. We create a lot of content about AI, about how to upskill people and get them ready for the age of AI. So we, you know, we’re definitely an organization that is impacted by AI. We could be put out of business by AI or we could be made dramatically more efficient about AI. Here’s what I tell the team, and I just told them recently, we’re going to continue to deploy AI even further into everything we do as Section. And everybody’s. All 35 of us were AI experts. We have all kinds of AI tools and AI workflows. But we’re going to go further. We’re just going to do as much as we can with AI. And one of three things is going to happen. Some teams are going to be smaller. We will be able to reduce the size of some teams. Some teams will stay the same size and do more work, and some teams are going to be bigger because with AI, those people are so much more valuable that we’re going to want more of them in the organization, not less of them. Right. And the honest answer is I’m not sure which ones yet, meaning which teams will be in which way. But we’re going to kind of go do this together and figure it out and I think get to a better place. Meaning we’ll have more productivity, more new products, more new services. I think our clients are going to be happier. So, yeah, I think you need an answer, though. I don’t think it should be. We’re not going to talk about it. We’re going to deploy AI and kind of cross our fingers. It all works out. I don’t think that’s going to work. 

Drew McLellan [00:11:08]: 

Yeah, well. And I think it’s becoming increasingly obvious that doing nothing with AI is. Is actually almost impossible. It’s not even that you can choose to not do anything, just interwoven in everything. And it’s becoming so ubiquitous in the community of profession, personal, all of that, that you kind of can’t avoid it. 

Greg Shove [00:11:31]: 

That’s absolutely right. Listen, I. This reminds me of, like talking to lawyers two years ago or management consultants when I did. And I talked about using, you know, law firms using AI or my. The partner at my Law firm that I use using AI. And of course their answer was, oh, no way, you know, absolutely not. Right. Not right. And now they’re all, you know, they see AI absolutely coming into their firm at all levels. Right. And so, yeah, this is inevitable at this point. I think we have to be optimistic. Pragmatists. That’s my approach. I’m a pragmatist. It’s coming whether we want it or not. It’s already here. Our clients are going to expect it, quite frankly. They’re going to have some concerns about it, you know, in terms of how we use it as, you know, in terms of supporting our clients, but they’re going to expect it. Why would you hire a lawyer or a management consultant? That is dumber and slower. 

Drew McLellan [00:12:16]: 

Right. 

Greg Shove [00:12:17]: 

And that’s what I tell everybody who’s in the business of selling their time and selling their intellectual capital. Like, clients are going to expect you to use AI because they want to hire the smartest person who works the fastest because that’s going to get the. 

Drew McLellan [00:12:28]: 

Best outcome right and the deepest right. That’s one of the things that our research showed. It wasn’t necessarily that clients wanted it cheaper. They just wanted more. And they realized in the same amount of time, but for the same expense, they could get more. 

Greg Shove [00:12:41]: 

They can get more. That’s right. That’s why I say some teams will stay the same size. They’ll be doing more Right. And clients will get the benefit of that. They’ll get the more right. Hopefully. Or at least some of them. Yeah. The way I think about it is we’re at a crossroads, I think individually and for organizations. Crossroads meaning over the next couple, three years, not like today. But the crossroads is there’s going to be one of two ways we use AI individually and then our teams and companies, we’re going to lose our minds or borrow 30 points of IQ. 

Drew McLellan [00:13:10]: 

Right, right. 

Greg Shove [00:13:12]: 

We are going to cognitively offload and leverage these technologies to enhance or augment ourselves. But I think we’re going to have a choice in terms of how we do it. The lazy way is going to be the way that frankly, Big Tech wants us to do, which is outsource all of our cognitive thinking to AI, basically get this huge, know, coworker, co pilot, whatever, that we just kind of like mail it in, you know, we just like, okay, you know, what if. What’s AI say? We’ll cut and paste and pass it on to the client. That’s not good, obviously. That’s what I call losing our minds. Yeah, Big, Big Tech like that. We’re addicted. We love it. We have to pay for it. Right. The other is we’re going to borrow 30 points of IQ every morning when we go in the office. And I think we’ll learn how to manage our AIs, learn how to enhance the output from AI. You know, everybody gets the same superpower. So if you’re going to take the lazy route, you’re going to be in kind of a, you know, AI passenger. All your work’s going to sound like everybody else’s work and it’ll be undifferentiated, you know, not particularly interesting, not particularly creative. But if you manage your AIs and really borrow that IQ, but add the human specialness, right. Had the creativity, add the judgment, add the intuition, you know, add what makes us unique, then your work’s going to be faster and better. But you’re going to have to differentiate it because AI is basically free. So we’re all going to use the same superpower. 

Drew McLellan [00:14:34]: 

Right. Until it’s not free anymore. 

Greg Shove [00:14:37]: 

Until it’s not free, yeah. And we know the playbook, right? Make it near free or free or near free. 20 bucks a month is near free, Right. So make it near free, get us addicted, raise the prices later. Right? Yeah. So they’re absolutely going to run that playbook. It’ll be apparent in a few years. 

Drew McLellan [00:14:51]: 

Yeah, I think so. So you said something interesting. You said that one of the ways clients come to you is if they’ve deployed AI, but it’s failed because they viewed it as software. So can you talk a little more about that? 

Greg Shove [00:15:05]: 

Yeah, sure. So, I mean, we’re used to software working the same way every time. We’re used to software doing a defined set of tasks within a domain. So, you know, email automation software, you know, does a certain set of tasks, has a certain set of features and capabilities inside of its domain to manage your email or manage your, you know, managing creative output or whatever it might be. Right. CRMs, CRMs all kind of work the same way. Some have some features, some, you know, some don’t. But we’re used to working with software that behaves the same way every day, that when you put data into it, it comes out, you know, in the right format every time. And it basically does a certain set of functions that helps us. Right. AI doesn’t behave any way like that. AI is a co intelligence that behaves erratically, hallucinates. Its performance varies not by day, but by hour. 

Drew McLellan [00:16:01]: 

Right? 

Greg Shove [00:16:02]: 

Right. I mean, it loses its memory. It has memory, but then it loses Its memory. I mean, I was working with GPT last night for two hours on a health question, you know, using GPT as a doctor, basically. And I was getting really frustrated because it would lose its memory. It was brilliant when it held its memory and the conversation was really productive. And then there was moments where you could tell it was losing its memory and I had to kind of go back and remind it and kind of retrace some steps. Super frustrating. Overall, I got great value out of the two hours spending that time with GPT. Much more time efficient than trying to chase down three different doctors or whatever, go down the rabbit hole of Internet searches, but still frustrating. Software doesn’t work like that today. Software is more reliable, more precise, more predictable. So when we deploy like software, I think we’re naive and fundamentally misunderstand what these capabilities are. So what the mistake we often see is the CTO buys it, or the head of technology or the head of operations, they buy the AI, you know, GPT for teams or Claude for teams or Copilot or Gemini, whatever it is, and they deploy it like software, meaning they turn it on, give everybody access and they run a couple training sessions or they, you know, I don’t know, they do a couple lunch and learns. Okay, not bad. But that’s not, you know, it’s not going to work. Meaning you’re relying on the early adopters, the growth mindset people, the innovators, they’re going to pick up and run with it. There’s only 10% of any organization. Drew never has more than 10% who are truly early adopters and innovators. They’re already running with it. And even if you banned it, they’d already been using it on their phone, right? 

Drew McLellan [00:17:39]: 

That’s right. 

Greg Shove [00:17:40]: 

It’s like, we know that. But if you’re really going to get to 50% adoption and weekly and daily active usage, which is transformative, particularly for language intensive businesses, for your audience agencies, we’re in the crosshairs of AI, meaning risk, but tremendous opportunity because we’re in language intensive businesses. But if you’re in that kind of business, you’re going to want to get to 50% weekly and then daily active usage within 12 months, in my opinion. Otherwise you’re missing the opportunity and eventually get to 80 or 90%. You always have a few laggards who want to stay on the sidelines. You know, hopefully they can pull that off if they really want to do that. I’m not sure how you do, especially if you’re in the creative services and Agency business. But yeah, I just think that we can’t think about it like software. We have to think about it as this ever changing set of very powerful capabilities that we need to support our teams to really figure out, use and keep upskilling. I hear a lot Drew from people like we’re waiting for this to settle down or we’re waiting to see which one’s going to win and then we’ll train people on the win or whatever. Like it’s not settling down. Right. It’s going to be. 

Drew McLellan [00:18:46]: 

We’re just at the infancy. So just have a long time before it’s a mature. 

Greg Shove [00:18:50]: 

Yeah. And if you wait for it to kind of mature out a bit, I mean, kind of, you know, wait for the next 10 years, this thing, you’re right. This is just getting started. And the models, it does make it harder. Listen, let’s agree that the AI industry hasn’t, has not made it easy on the humans. They keep releasing new capabilities. They keep changing the names of these things like they’re, you know, it’s a mess. Talk about needing some product marketing help. These guys need some product marketing help. But listen, a billion people will be using GPT by September is the estimate of this year. It’s 10% of humanity. This product is already. Think about it. This is a hard to use product that behaves erratically and hallucinates and makes mistakes and already a billion people use it. It’s crazy. 

Drew McLellan [00:19:34]: 

Yeah. 

Greg Shove [00:19:35]: 

It’s not software. Don’t deploy it like software. Deploy it like what it is, which we’ve never done before as a species. And we’re deploying this co intelligence and we’re going to hope that we thrive while using it. That’s the goal here. 

Drew McLellan [00:19:47]: 

So for agency owners that, and I think a lot of agency owners have, they have a task force that’s playing with some tools or whatever, but they haven’t really wrapped their head around how to move from the novice to competence. If somebody’s listening and they’re like, okay, I know I have to do this, but I have no idea how. Would you guide them into creating a plan, a learning plan, an exploration plan, whatever you want to call it, to get from 0 to the 10 to the 50, you know, and you’re saying that boy, you got to be at 80 or 90 within a year. 

Greg Shove [00:20:30]: 

Yep. 

Drew McLellan [00:20:30]: 

What does that look like? What does that learning path look like? 

Greg Shove [00:20:34]: 

Sure. So let’s zoom out for a sec. I want to introduce a framework. Drew, I think that any agency owner or leader can use or any frankly leader in any, any organization. It’s called oat. Oat Optimize, Accelerate, Transform. Optimize means optimize my internal operations with AI, make myself more efficient, make everything we do internally more efficient. The result of that will be gross margin expansion in the, you know, certainly in the short term. Right. You’re going to become more profitable as an organization that’s optimized. In my opinion, a hundred percent of organizations that have knowledge workers sitting in front of computers can optimize. Accelerate means can I meaningfully change my product or service by injecting AI into my product and service so my clients do see the impact of the AI and meaningfully change it so that they will buy more, pay more or renew at higher rates. Right. Be happier customers. The third is Transform. That is the most aggressive operating mode or strategic mode which is I have to transform or change my business dramatically and quickly because I’m in the crosshairs of AI. The example I use always is Chegg, the homework helper company that did not transform is now out of business essentially or the company for sale for scrap. So don’t be checked. Right. Transform if you need to. But the way I think about it is as a leader, bring a framework like that to the whole team and say listen guys, we’re going to optimize, Accelerate and maybe we’ll transform or not, maybe we don’t need to, but we’ll optimize and accelerate. So inside of Optimize, that means every workflow or process here will examine it and see if AI can make it faster, better, cheaper. The way you do that is you give everybody access to a paid LLM GPT. Gemini, Copilot, it doesn’t matter. I don’t think really we have our preferences and if you’re a Google shop you want Gemini. Maybe if you’re a Microsoft shop you want Copilot. Give them paid access to the premium version of that. Hook up company data and use the enterprise version. And if you’re not a big company, use GPT for Teams or Gemini for Teams. It’s an extra five bucks a month per user and your default default settings are none of your data and none of your clients data is being shared with the, with the AI model which is what we, you know, just take care of that right away. It’s easy. Do not be worried about data privacy. We, we’ve been unnecessarily concerned about data privacy. I think it’s been exaggerated the risks of kind of data leakage. So turn on good AI for everybody. Allow people to use company information and client information. Not pii, but other company and client information. You get more value from AI when you work with data that’s meaningful to you. Right. And finally, have a non dumb AI policy. A non dumb AI policy is one that encourages people to use AI. And the way to think about this is this way, if you have an AI, sure, you need an AI policy, you need rules and guardrails, I get it. But if you stop there, all you’re going to do is discourage usage because you scare people. If I’m doing it the wrong way, I’m going to get fired. Right. Plus people aren’t sure, especially in agencies. Am I cutting quarters? Am I cheating? Will the client find out? You got to deal with all that stuff up front. So we talk about AI policies and AI manifestos. Non dumb AI policies have the policy and they have the manifesto. The manifesto is the point of view in your case, in your audience’s case, of the agency owner, what does he or she think about the use of AI? Spell it out, encourage it. It’s not cheating, it’s not cutting corners. You don’t hide it from clients. Right. You celebrate success with AI, you share lessons. When AI doesn’t help, that kind of idea, I think that’s the minimum sort of plan, I guess. And then you got to decide carrot or stick. And it’s probably a bit of both. Meaning we’ll encourage people and incent them to use AI. And occasionally, if we see the adoption slowing down or not even not really getting anywhere, we might have to go a little more stick. Toby Lucky, the CEO of Shopify. That’s a big company, 8,000 employees. You can tell what happened. He got frustrated over the past 12 months and just said, okay, you know, I’m going to lay down the law here. Every performance review of every employee will include an AI proficiency measurement. So we’re not going to give people good performance reviews unless we can determine their AI proficient. You can’t ask for more headcount, human headcount, unless you can prove that the AI can’t do the work. So they’ve laid out a manifesto, a point of view that’s really pretty direct around. They’re going to become an AI first or AI enabled organization within 12 months. And they should. They’re a software company, right? 

Drew McLellan [00:25:13]: 

Right. 

Greg Shove [00:25:14]: 

Yeah. I think at some point the leaders got to step up and decide how to catalyze this change. 

Drew McLellan [00:25:19]: 

So we need to take a quick break, but when we come back, I want to ask you, there’s so many Tools out there. So many tools out there. How does, in our case, an agency, do we need to know all of them? How do we know which ones to use? Like when we’re, when we’re starting to do this optimization, how do you know or how do you explore or how do you investigate the right tools for the task? So let’s, we’ll come back and talk about that right after this break. Hey, everybody. Just want to remind you before we get back to the show that we have a very engaged Facebook group. It’s a private group just for podcast listeners and agency owners that are in the AMI community. And to find it, if you’re not a member, head over to facebook.comgroups/b a b a podcast. So again, facebook.comgroups/bab podcast. All you have to do is answer a few questions to make sure that you are an actual agency owner or leader, and we will let you right in. And you can join over 1700 other agency owners and leaders. And I’m telling you, there’s probably 10 or 15 conversations that are started every day that are going to be of value to you. So come join us. All right, we are back. So I asked you before the break, 8 million bazillion tools and there’s three new ones every hour, it seems like. Right. So as we are looking at, because for agencies, it’s not just about our own internal efficiencies, as you say, but it’s also about how do we change the way we serve our clients. So there is a fear around. I’m going to hook myself up to the beta video rather than the VHS machine, and we’re going to be deep into that and then beta is going to go away. So how does a company, how do agencies vet the tools or know where to lean into? And how many do we have to have in our toolbox? Is it a 300 tools? Is it, you know, the top three or five is plenty. When you’re guiding clients through this process, how do you help them build out their tool set? 

Greg Shove [00:27:26]: 

Yeah, sure, let’s. I think that’s sort of an easy answer in some ways, which is start with the basics, which is everybody gets access to a premium LLM. Some people call them frontier models. That would be GPT, Gemini, Microsoft Copilot, maybe xai’s Grok if you want unhinged AI. I don’t want unhinged AI inside of my company, so. 

Drew McLellan [00:27:48]: 

Right, right. 

Greg Shove [00:27:49]: 

Not a grog fan. Yeah. So I mean, that’s first the premium version, so the paid version and Start there one, maybe two at most. At Section, we love Claude for its thought partnership, its kind of thinking style, its writing style. We love GPT, Swiss army knife of AI and we love Perplexity for its research kind of Google replacement that’s probably not surprising to most people. Those are, those are obviously three common models. If we had to get by on one, we’d probably get by on GPT. And we use, we use GPT for teams. Start there. Two other aspects to the answer. Your AI champions or your AI Tiger team or SWAT team, whatever that, whatever you call that team, pick a small group of people and give them more latitude to experiment. Don’t have experimentation happening across the company. Get everybody proficient with the one or two sort of, you know, everyday tools you want to use. Take your AI champions, your specialists and give them a little bit of budget and let them go play a little bit more and see if they want to bring something back to the rest of the organization. And then the final step is go deep on the high return use cases. So as an example, adsection, we do use synthetic video. We use avatars now when we create some of our learning content. And we knew we were going to go deep there once we began to experiment and we chose what we thought was the best tool at the time. Six months in we switched horses. We didn’t change the use case but we changed the tool out. We might change again in six months. Right. That change and the cost of that change is worth it because that’s a high value use case for us. Right. And so there’s probably half a dozen to 10 of those where it could be audio, it could be video, it could be presentation software, like you know, gamma, something like that. It could be design, you know, AI design software. Yeah, it doesn’t matter. But pick your sort of top use cases where the specialized AI software is going to make a difference. Give it your best shot, like pick what you think is the best one. Try not to negotiate 12, you know, try to avoid the 12 month contract, you know. 

Drew McLellan [00:29:53]: 

Right, right. 

Greg Shove [00:29:54]: 

Do the monthly deal kind of thing and just get going. I do think that people are trying to over optimize here. Don’t over optimize around the tool. Over optimize if you will, around the use case. Prioritize your use cases. The way to think about this is there is hundreds of ways to use AI inside a typical agency. No shortage of ways that AI can be helpful. So your job as a leader and or your AI SWAT team is kind of open that aperture, Brainstorm list All those ideas, potential use cases, and then to close that aperture, meaning pick 10, run that for 90 days. What you’re going to do is you’re going to throw out some of those, they didn’t work. AI wasn’t that helpful. And then the next quarter, add more 10 onto that list in terms of use cases, not tools, but use cases. 

Drew McLellan [00:30:42]: 

So when you, when you’re working with clients, what’s the way they, it doesn’t occur to them to use AI that you open their eyes to, and they’re like, oh my gosh, I never occurred to me to use it that way. But now I’m going to use it that way every day. What’s the thing we’re missing? 

Greg Shove [00:30:59]: 

Yeah, what we’re missing is we’ve never worked with software again. It’s not really software. We’ve never worked with anything like this before that has such a big broad range of capabilities. Right. It can draw and it can do financial analysis. 

Drew McLellan [00:31:12]: 

Right. 

Greg Shove [00:31:12]: 

It can be creative and it can be a doctor. Right. It can be a therapist and it can be a, an accountant. Like it. Again, it’s, it’s kind of crazy. We’ve never had this capability. So a couple little hacks. Put a post IT note on your monitor and write ask AI on the post IT note and just stick it on your monitor. Like, we’re not AI native, so we have to remember and remind ourselves all the time. You know, could AI be helpful here? So, you know, for the last 18 months, two years, I used a Post IT note. January, I tried something else. Beginning of this year, Drew, I tried something else which has been really successful. Costs 100 bucks. I have an AI monitor on my desk. So in front of me, just off to the right here, is a monitor. It’s on the vertical horizon instead of horizontal. It’s got one browser open with three tabs, Perplexity, GPT and Claude. My favorite AIs are open every day and so I can see AI all the time. It’s in my line of sight. It’s not on my main monitor. It’s just, but you know, it’s, it’s ever present, right. I work in that monitor with my AIs, and then I cut and paste that work over to my, to my main workspace, right. It’s, it’s helped a lot. That reduced even further the friction for me to remember to use AI and to have a kind of an AI workspace. I’ve just bought one of these limitless pins. I’ll probably lose it in a week. So I don’t know, but my plan is to run a one month experiment. Are these kinds of AI technologies useful in terms of creating less friction or reducing friction? Rather, in terms of how we access AI and things like meeting note takers and stuff like that. But listen, we’re going to have to remind ourselves, the kids coming out of college, they’re AI native. I just hired a kid out of Columbia, political science degree, just graduated, got his degree two weeks ago and started with us on Tuesday, kind of thing. He’s a representation of what Sam Altman said. For people under the age of 25, it’s an operating system for their lives. 

Drew McLellan [00:33:05]: 

Yeah. 

Greg Shove [00:33:05]: 

For everybody else, we’re having to remind ourselves and kind of teach ourselves how to use AI. This kid, he just joined our team. He’s going to run circles around us because he’s AI native. Like, he, he. It is an operating system for him. Every question, every task, every project, he just right away thinks, you know, AI first I’ll go to AI and AI will get me started. AI will give me a V1. So, you know, the good news is, as these kids join our companies and join our teams, they’re going to show us how to kind of be a native. Right. And how not to have to think about it that much. We have to get ourselves there, I think pretty quickly. It’s harder, you know, when you’re older. 

Drew McLellan [00:33:44]: 

And we have to get out of their way and let them teach us. Right. I mean, absolutely. It’s also not the business model to have the, the new kid or the intern teaching the CEO how to, how to do new tricks. So we have to be open to that as well. 

Greg Shove [00:33:58]: 

Yeah. Listen, the thing I do also want to talk about, Drew, that’s coming, is business model innovation, which either will be disruption for some or opportunity for others. Right now, so much of our conversation about AI is focused on kind of technology innovation and disruption. Right. And how that impacts our products or services. So that makes sense. You take a new technology, you realize, okay, how do I use it, how do I work with it, how do I integrate it into my products or services that my clients buy and use? That is always what happens first. But when you think about the companies that really get created in these moments, or the ones that really thrive, incumbents that thrive, or new companies that are created, they’re the ones that take the technology innovation and they bring a business model or kind of revenue innovation. So, you know, Amazon, aws, even Snowflake, and Software, Airbnb, Uber, the list goes on. They took the Technology innovated dramatically on the customer experience, improved it dramatically, and then they innovate on the business model. So this is what’s about to happen. I don’t think there’s enough conversation going on yet. I think it’s about to start around how should I change my business model and not, frankly, as an incumbent, get caught flat footed, right, by the change in the business model. 

Drew McLellan [00:35:19]: 

You don’t want to be the taxi company. 

Greg Shove [00:35:20]: 

Yeah, you don’t want to be the taxi company. You want to be. You don’t have to be the one to move first. What I tell incumbents is, I get it, you have a good business model and it’s working. You don’t want to necessarily blow it up. This idea of disrupting yourself, it’s easy to say in business school. It’s hard to actually do. It’s kind of scary, right? So you don’t need to necessarily go first, but you absolutely need to be ready. If that new business model starts to stick, then like in software, Snowflake’s model is, you know, pay per consumption, not pay per seat. 

Drew McLellan [00:35:50]: 

Right. 

Greg Shove [00:35:51]: 

If that spreads and starts to stick and clients want it, you better be ready to switch to it or, you know, you’re kind of in trouble. Right. The obvious example of knowledge work is, is lawyers. And we’re all been talking about this, when will it happen and who will lead in terms of charging for outcomes or charging for output versus charging by hour. And obviously that can apply to any kind of agency or professional services firm. But this is what is going to happen, in my opinion. I do think the new winners or the incumbents that thrive will be the ones that when the business model switches, they’re able to make the switch at the same time. Because nothing hurts an incumbent more than the business model disruption. They can handle the technology disruption because they can catch up and improve their product or service with the new technology. But when someone comes along and says, you know, you used to pay $100,000 for this and now we can charge you $10,000 for this, that’s when the incumbents are like, oh, shit. 

Drew McLellan [00:36:45]: 

Right, right, right. 

Greg Shove [00:36:47]: 

So one of the reasons I want executive teams through agency owners and leaders to get really steeped in AI right now, like really using it themselves every day, is I want them to appreciate the power of these technologies so they can figure out the business model impact. All right? So if you really understand these technologies and how clients are going to be expecting us to be using them and delivering our products and services to them, you’ll begin to Appreciate. Well, wow, how might I charge for this, right? I mean, every AI native, you know what’s happening. That’s why every AI native competitor is sitting in the Silicon Valley garage, right? Two guys and a dog. And now it used to have a PowerPoint, now they have AI. So they’re sitting in the garage, right? Three guys and a dog and access to GPT. And they’re building these companies and they’re thinking about a new AI product or service and they’re also thinking about how will we charge for it. That catches the incumbents completely flat footed, right? The business model innovation. So I just want us and your audience to begin to have that conversation as well internally. 

Drew McLellan [00:37:50]: 

So if that piques people’s interest or fear or both, how do they begin to explore that? So yes, they have to get more proficient at AI, but understanding the tools and what you can do with the tools and the efficiencies that come from that are very different from saying, here’s how we’ve always done these five things. I have to come up with a new way to do these things. So how do you guide your clients? Because I’m sure you’re saying we have to get to the transformation stage. We can’t just do the optimization, we have to get all the way to the T of the oat. So how are you helping them view the world differently so that they can see the potential model shifts one of three ways. 

Greg Shove [00:38:39]: 

The first way is an oldie, but a goldie. Michael Porter’s five forces classic strategy framework for business. Been around for 30, maybe 40 years. I don’t know how long it’s been around. Basically get your management team in a room, watch a couple of videos on YouTube about Michael Porter’s Five Forces, or frankly ask AI to explain it to you and run a Five Forces workshop internally. Even if you’ve never done it before, it doesn’t matter, meaning you’ll figure it out. It’s not rocket science. It is very illuminating. It is a classic and useful strategy framework to kind of expose what happens to supplier power, customer power and so on, right? The different stakeholders inside of a business or inside of an industry when something changes. So that’s the first thing we say, get everybody in the room and you know, we’ll help if you need us, but you can do it yourself. That’s number one. Number two is a similar exercise but different, which is think like an AI native startup. So you know your clients, you know what they buy from you, you know what they pay for that. If you were two Guys and a dog and sitting in a Silicon Valley garage kind of thing. And you had, and you had, all you had was AI. You had inference, you had access to the best AI tools. What would you build and what would you charge for? Think like an AI native startup, Think like a startup, think like a challenger, don’t think like an incumbent. And the third way is ask AI. Yeah, I mean, I’m serious. You’re the CEO, like, sit down. Use AI as a thought partner. It’s the number one undervalued use case of AI for executives. AI is an amazing thought partner for any leader. And I tend to find executives and owners and you know, leaders think, oh, AI is not for me, it’s for everybody else. You know, the copywriters are going to use AI, but I don’t need. Yeah, no, that’s. AI is an amazing thought partner. Any medium to high stakes decision, I think it’s irresponsible. I tell leaders of public companies, publicly traded companies and their boards. It’s irresponsible for you as a leader that you’re not asking AI for assistance when you’re making any medium to high stakes decision at home or at the office. Medical, parenting, whatever. At home or at the office, anything. Go to market questions, competitive questions. And this question, what I would do is write a great prompt. If you don’t know how to write a great prompt, that’s a good thing to learn. Write a great prompt and have a long conversation with your AI about what an AI native challenger would do and how as an incumbent you might need to react or what you should do kind of thing that’s absolutely going to take you an hour and it’s going to spark a lot of good ideas and kind of conversations with your management team. So that’s what I would do. And again, you don’t need to make these moves yet, but you can just begin to get your management team and thinking about it. Right. And what would you do if this happened? What would happen if I’m a law firm and an AI native law firm approached my client and said, we have a new service, a new product and this is what we charge for. How are you going to react? Yeah, you know, lose the client. That’s one option that could be okay, you know, you might decide, no, we don’t want, we don’t want that business. It might be, no, we have a similar offering and we can do that. Give us a month and we can turn that on. I don’t know. Right. You know, different ways to react. But again, don’t, don’t get caught flat footed. Think like a challenger, I think is the way to think about this, Drew, and kind of be ready. The last thing I’d say is run some numbers. Like, don’t be so scared of this that you don’t run the numbers. I find leaders often avoid the questions that they’re afraid of the answer to. So when you’re afraid of the answer, you don’t go get the answer because you’re kind of afraid of the answer. So go get this answer. And the answer is, what would happen if we had to change our pricing with our clients? Run some napkin math. Don’t spin up a whole huge effort, but run some napkin math with your best clients. Okay, here’s my best client. What happens if an AI native competitor went and charged something differently for the same thing that we’re offering that client? If the retainer went from 15,000 to 5,000, but you know, the client was paying for outcomes, whatever that would look like, what would that math look like for my agency? Just like sketch that out a bit, go to a whiteboard and kind of figure that out. You might be like, hey, that’s cool, I can figure that out. Or I might be like, holy shit. Like, that wouldn’t be good. 

Drew McLellan [00:42:52]: 

That’d be amazing. 

Greg Shove [00:42:54]: 

Or be amazing. Or I want to lead. I want to be the one. Like, my take is, but I’m a startup guy. My take is, is if I’m an incumbent. Incumbents have such advantages. Incumbents have brand, incumbents have clients. Incumbents have trust and reputation. Incumbents have inside knowledge. Right. And like, we have so much advantage as an incumbent. Just you be the leader on this change. Don’t. Don’t let someone do it to you. Right. Again, pick your moment. I think you can create new products and services with this stuff and expand. 

Drew McLellan [00:43:23]: 

No doubt. 

Greg Shove [00:43:23]: 

It’s not all about losing revenue. I think it’s about growing revenue. 

Drew McLellan [00:43:27]: 

So in the confines of what you’re willing to share publicly, what’s the most surprising decision AI has helped you make in the last three months? 

Greg Shove [00:43:37]: 

I’ll tell you right now. I have prostate cancer, which is a downer. And no one wants to talk about cancer, but I had lung cancer three years ago. I got prostate cancer. Now cancer seems to run in my family, and I have been working with AI to help me sort out sort of what is my protocol to slow the progression of my prostate cancer. And I have access to amazing doctors, and I pay for access to amazing doctors, you know, above and beyond what my insurance handles for me. And AI has been more time efficient and giving me better advice faster around. You know what I should be doing? I’ll answer your question a little bit differently, Drew. I’ll tell you about the most surprising use case that we discovered about 18 months ago, but we still use it every quarter. And then we just did it recently in April. So every board meeting I take the board deck. So the, you know, the content that I’m sending the human board before the board meeting. We typically send it a couple of days before the board meeting and we feed it into AI, our favorite AIs and we basically ask the AI to play the role of a board member. So we first did this 18 months ago. We’ve done it every board meeting since. Here’s what happens. AI replicates 90% of what the board tells us in the meeting. So we don’t use, yeah, we do not use meeting note takers in our board meetings. That’s for governance and legal reasons. But I have my chief of staff, you know, join the board meeting and take great notes over two hours. What all these humans, you know, all the advice we got from these humans. And I’ve got a great board, probably one of my best, best boards ever in terms of a startup. And I’ve got guys are on public company boards and I’ve got this guy, Scott Galloway, he’s kind of an opinionated podcaster. He’s on my board and I’ve got a good board. AI, Claude and GPT basically replicate 90% of what the board tells us and the 10% where AI doesn’t and the board don’t match. The humans are saying different stuff that the AI is not saying and the AI is saying different stuff that the humans aren’t saying. But 90% of the time the AI can tell me what the board’s going to tell me. So am I replacing my board? No. Right? We need boards or we need advisors, we need human coaches and mentors for sure. And by the way, boards in terms of startup, boards write checks, boards invest in companies. I need the board for that. At least, at least for now. 

Drew McLellan [00:45:41]: 

I won’t write you a check, right? 

Greg Shove [00:45:43]: 

Not yet. Not yet. Not yet. Yeah, we’ll get there. But for any, anyone again, making a medium to high stakes decision or preparing for a medium to high stakes meeting, you know, with a board, with a doctor, with an investor, with it, with a client that really matters, you know, role play this stuff, role play before the meeting or before the decision with your AI. And if you’re doing this for the first time especially. I mean, you know, I’m old enough. I’ve been around the track a few times. So, you know, board meetings aren’t as difficult for me as they used to be. But if I’m a first or second time entrepreneur or been in this situation for the first time, then yeah, why go in unprepared? You know, role play with AI. 

Drew McLellan [00:46:21]: 

It’s great. This has been fascinating. I feel like we could talk for another four or five hours, but this has been fascinating. So, Greg, if people want to learn more about the work you do, how you might be able to come alongside them and accelerate sort of their understanding and adoption of what AI can do for their business, what’s the best way for them to follow your thought leadership, to connect with you, to explore a partnership with you down the road? 

Greg Shove [00:46:47]: 

Yeah, sure. Sectionai.com is, is one way. Email me. Gregctionai.com would be great. Love to hear from you. Or yeah, you can Find me on LinkedIn. Any of the above. We’d love to hear from you. I will say one thing, Drew. We are making our AI coach. We’ve built an AI coach called Prof. AI Prof. AI helps people learn how to be more proficient at AI. Companies have to pay for it, but individuals get it for free. It’s so core to my personal mission. And the mission of Section is to get as many people in the AI class as fast as possible so they can take advantage of this moment in terms of the age of AI. So we’re going to make Prof. AI free for any consumer, high schools, colleges, nonprofits. You know, we want more people ready for what’s happening. And that means get, get proficient, get confident with your use of AI. So prop AI for consumers. It’s free. And you get a LinkedIn badge as well to prove to your employer, prospective employer, that you’re AI ready. 

Drew McLellan [00:47:45]: 

This is, this has been fascinating. Thank you. Thanks for sharing your expertise and for your enthusiasm. I think one of the things that everybody needs to hear is all of the things that are possible as opposed to all the things that are scary. And I think you helped a lot of people today. Go. You know what? A, I can do this and B, I need to do this. And I’m kind of excited to do it. So thank you for that. 

Greg Shove [00:48:09]: 

Oh, thank you. It’s been great, great conversation. 

Drew McLellan [00:48:10]: 

Okay, guys, this wraps up this episode. Tons of takeaways from this episode. You know, I am a firm believer that every episode needs to inspire you to go do something, to take some action. And so lots of Action items out of this podcast for you. And so get started. Roll up your sleeves. 

Greg Shove [00:48:27]: 

And. 

Drew McLellan [00:48:27]: 

And I think one of the most important things that Greg said that I want you to hear is many of you have a task force or a team of people who are exploring AI, but a lot of you have stayed sort of at arm’s length from it. You know, like, you’re thinking it’s not really CEO stuff, it’s really practitioner stuff or producer stuff, and that is just not the case. And I think Greg gave you a couple of great examples of how you should be leveraging it. But if we don’t understand it, we cannot lead the way. And we cannot be thinking. Our employees are not going to be thinking about how disruptive this is to our business and the business models. That’s us. That’s our leadership team. And if you don’t understand the tools and the power behind it and what. What it is doing today and what it’s going to do, then you cannot understand what it’s going to do to our industry. And so if nothing else, you need to dive deeply into that so that you can anticipate what your clients are going to need from you in the future so you can get ready to do that. So all of that, start paying attention, start digging in, start getting your hands dirty in it. I want to make sure that I thank our friends at White Label iq. As you know, they’re the presenting sponsor of this podcast. They make it possible for me to bring folks like Greg to you every week to get you thinking differently about your business. So head over to whitelabeliq.com ami, check them out, and if nothing else, send them a note on LinkedIn or on their website and thank them for being a sponsor. Let them know that you appreciate the fact that they bring this show to you. And last but not least, I love these conversations. I love learning. I learn right alongside you. So I’m coming back next week, and I hope you will, too. I’m grateful that you spend some time with me every week, and I will keep bringing you folks like Greg so that we can hang out together. Okay? All right, I’ll talk to you next week. Have a good week. 

Drew McLellan [00:50:11]: 

Come back next week for another episode designed to help you build a stronger, more stable and sustainable agency. Check out our workshops, coaching and consulting packages, and other professional development [email protected].