Dan Lewis, founder of The Comms Stack, helps communicators approach AI with a builder’s mindset. On this episode of SaaS Half Full, he joined host Lindsey Groepper to talk through where AI falls short, how communicators can build their own tools without learning to code and why AI mandates often stall.

What AI still can’t replace

Dan pushes back on the idea that AI lets companies cut back on comms support. LLMs can draft talking points or an all-staff email, but they can’t anticipate the pushback a message might generate or catch risks a company hasn’t considered. That judgment is exactly what a senior leader loses when they assume a chatbot can do the job instead.

“You can’t say to your employees, ‘Oh, I didn’t think about that, let me ask ChatGPT for an answer.’ And you can’t say in an interview, ‘ChatGPT suggested I say this, and we didn’t explore the question you had.” — Dan Lewis

Dan also warned that audiences can spot AI-generated writing the same way viewers spot an almost-human CGI character. Once a message reads as generic or robotic, it becomes a liability, damaging trust with the employees, customers or partners a company is trying to reach.

Comms professionals don’t need to become developers

Dan has built his own systems to get more out of AI than basic prompting allows. Using Claude Code, he created an AI chief of staff with access to his inbox, chat transcripts, meetings and calendar, so it can track recurring deadlines and surface what’s coming up next.

He turned the same approach into a free course called the Comms Gut Check, which walks communicators through building a tool that reviews a draft for tone, jargon and audience pushback — no coding background required. 

Working with AI is more of a mindset shift than a technical one. Dan’s approach treats the tools as a way to pull together scattered data sources and turn them into something usable, the same principle behind the Comms Gut Check tool.

Why “just use AI” doesn’t work

Dan has talked to plenty of senior leaders who tell their teams to use AI without explaining what problem it’s supposed to solve. He pointed to ADKAR, a change management framework, and argued that AI adoption needs a specific business outcome to work toward, like a KPI a team is trying to move, rather than a vague push to “be more efficient.”

A few other takeaways from the conversation:

• Dan isn’t concerned about Claude’s watermark feature on its own, but he’s watching how much metadata gets attached to it, especially if it starts showing how AI was used.
• He expects some job loss in comms as AI takes over operational tasks, though he sees an opening for junior employees to build skills by working alongside AI tools.
• For newsletter platforms, Dan recommends Substack for thought leadership and Beehiiv for building a media brand. His bigger takeaway: just publish.

Dan’s conversation with Lindsey covers a lot more ground than we can fit here, including his take on where the comms function is headed over the next several years.

Episode Transcript

This has been generated by AI and optimized by a human. 

[00:00:00] Dan Lewis: I really think that it’s important for any team when you’re trying to figure out how to get your team to be AI-enabled, to be really focused on change management.

[00:00:10] that all starts with why. So if you start with the why, like here’s the key performance indicator that we’re trying to actually move and in what direction and how much, that’s a lot stronger than just say, use AI. 

[00:00:22] welcome back to SaaS Half Full, the only show serving B2B SaaS and AI marketers. I’m Lindsey Groepper, EVP at Pan Blast, and I will be both your host and bartender today.

[00:00:34] I had a great conversation with Dan Lewis, who is the founder of The Comms Stack. Now, we internally at Pan Blast, we follow and subscribe to Dan’s newsletter, always chock-full of interesting information on how the comms profession is adopting AI, um, as well as a lot of other interesting topics. So we thought, “Hey, we should have Dan on the show.”

[00:00:55] so while comms is just a little subset of overall marketing, we’re diving into some larger topics [00:01:00] around the technical requirements now of marketers, how comms leaders should be thinking about how they’re leveraging AI, and moving beyond just now using AI as prompts. How can we actually use code to make our jobs better, even if you are not technical in nature?

[00:01:15] So if you’d care to, grab a drink and join me as I speak with Dan

[00:01:19] Lindsey Groepper: Hey, Dan. Welcome to SaaS How Full.

[00:01:21] Dan Lewis: Hi, thanks for having me. Glad to be here

[00:01:23] Lindsey Groepper: I’m excited to have you. I know that you’ve talked with many of my team members at Pam Blast over the years. You and I have never had a chance to meet, so it is nice to finally meet you. Uh, I feel like in our internal Slack channels, we are often sharing little tidbits that we’re picking up from the comms stack.

[00:01:40] So it is nice to have the founder on here. Um, since we share so much of your knowledge internally, it’s good to be helping you share it externally as well. Um, are you joining me for a drink today?

[00:01:51] Dan Lewis: I am. I have with me a, uh, a strawberry soda. Uh, it’s, you know, lunchtime, so I’m taking it easy. [00:02:00] Yeah. What about you?

[00:02:00] Lindsey Groepper: perfect for a hot day. I’m assuming it’s hot where you are. We’ve been hitting, having like a 90-ish degree heat wave here the last few days. How about you?

[00:02:09] Dan Lewis: Yeah. We have the– So I’m in the New York City suburbs, and we– the summer has been like 75 to 90, oppressively humid during the day, and then at like 8:00, 9:00 at night, you just get a torrential downpour every night. It’s, it’s an adventure every day.

[00:02:24] Lindsey Groepper: we’ve been getting a lot of rain here, too. Um, well, as my listeners always know, no matter the day, no matter the time, I will always have a drink on my show.

[00:02:32] So I am having my go-to, which is a Tito’s and soda. Although I’m on the fruit train with you, this is like a, brand is Bubbly, but the flavor is cherry. So it is a cherry Bubbly and Tito’s, which tastes pretty darn good. Um, cheers. It’s nice to meet you.

[00:02:48] Dan Lewis: Nice to meet you too

[00:02:50] Lindsey Groepper: Uh, so Dan, we’re gonna be diving into topic around the comms function today, and we’re gonna cover kind of a lot of different areas of that.

[00:02:57] And, um, while it is a, [00:03:00] uh, a niche within marketing, um, I can… I would be hard-pressed to imagine that there is no marketer, regardless of your level, um, that hasn’t been examining their comms function over the last one to two years. Um, I love the fact that comms has, uh, sort of entered the spotlight even more so over the last five years.

[00:03:19] We’ve been seeing more, like, chiefs comms officers and, um, that being really more of a business function and not ju- not just a marketing function. So I’ve loved seeing that. Um, but before we dive into our topic, I want to give our listeners a little bit of background on who you are, Dan. you are founder of the Comms Stack.

[00:03:35] You are a marketing advisor and consultant. but what has your journey looked like that brought you to today? Has your… Was your background, uh, really in marketing from the start? Was it happenstance? Just give us a little bit of your background, and then would also love to hear about, uh, the Comms Stack.

[00:03:53] What is it? Who can gain access? Where to get access?

[00:03:56] Dan Lewis: Great. So, uh, again, thank you for having me. Glad to be here. [00:04:00] Uh, my path to where I’m at right now has been anything but traditional. I started as a lawyer. I’m actually a lawyer by training. practiced for a few years, but I really found my passion at that intersection between technology and storytelling.

[00:04:13] How do we leverage emerging channels, new channels to reach audiences directly? I, uh, about 15 years ago, had the opportunity to join Sesame Workshop, that’s the nonprofit behind Sesame Street, as their first-ever social media lead. And I expanded that role into a, full digital marketing function there, uh, that included email marketing, uh, SEO and all the website stuff, and of course, just social media as it expanded, right?

[00:04:41] It started with, like, Twitter and Facebook, and then it became Instagram and, like, deals with Snapchat for custom lenses, and then ultimately, after I left, TikTok after that I was at Warner Brothers Discovery where I had the opportunity to again focus on digital channel strategy for the corporate brand.

[00:04:56] So I switched over to the communication side of things, uh, in the corporate communications [00:05:00] function. And most recently I was at Visa. I actually went into internal communications to, map out that digital channel strategy for the thirty-five thousand global employees. So how do they connect with the company via the corporate internet and internal social media platform, which, uh, my team rolled out, digital signage in all of the offices and of course again email newsletters.

[00:05:20] So I’ve really always thought through, how do you get your story, your message to the audiences that you’re trying to connect with through all the nuances of these digital channels. What do you want them to think? What do you want them to feel, and what do you want them to ultimately do given what you taught them about?

[00:05:38] I’ve also on my own, uh, been really enamored with the, uh, direct-to-consumer value that you get from email newsletters. So, fifteen or sixteen years ago I started my own newsletter. It’s called, uh, Now I Know. It’s a trivia newsletter. It goes out daily and it reaches fifty thousand subscribers. About half of them open it every day, which is great.

[00:05:58] and all I’m sharing is a fun fact and a [00:06:00] story behind it. And that’s really taught me how to reach people in ways that marketers don’t necessarily think about, from the owned media perspective. So I really focused on owned media over the course of my career. Last January I realized that, you know, AI is impacting all of our work.

[00:06:18] I was, focused on communications work at the time in particular. as a newsletter veteran, I wanted to just kind of meld everything together into something that was useful. So I created a newsletter called The Comms Stack. It’s thecommstack.com. It’s specifically designed for, uh, communications professionals who want to understand how AI is impacting the work they do, how they can better adopt AI, and really just what this new AI first world that we’re entering, and I still think we’re entering it, I don’t think we’re quite there yet, how that’s going to change how we operate as communicators and as storytellers.

[00:06:53] Thank you. Lots to unpack there. Did not know about your little Sesame Street roots. That’s difficult when you are [00:07:00] communicating to parents and maybe at the time were not a parent, but it definitely is how a lot of us in the comms space operate, is that we are working within industries that we don’t necessarily live or have familiarity with.

[00:07:11] Lindsey Groepper: Um, and your Did You Know newsletter, also not familiar with that. can you tell us what e- like yesterday or today’s fact was?

[00:07:18] Dan Lewis: one of them was about how Christopher Columbus was able to, because he had, at the time, more modern technology than the Native Americans that, he came across in the New World, he knew a lot more about eclipses and other things like that, and he used that to, his knowledge of eclipses to, quote-unquote, “predict” that the moon would disappear, and convinced a lot of the natives that he had come across that he had, like, godlike powers because he understood that, right?

[00:07:46] Lindsey Groepper: I love that. Thank you for sharing. Okay, let us dive in here. So originally when we were ideating on a topic, um, we were talking through challenges that comms leaders have, um, [00:08:00] in really using AI as individuals and saying, “Oh, yes, I’m an AI user,” versus actually like having organizational or operationalizing AI across the comms stack.

[00:08:10] well, I want to talk about that for sure. but I also just want to get your opinion on how, um, the comms profession has changed because, you know, on any given day, there are stories about how, or people’s opinions, on how it’s changing for the good, others that are fearful that it is gonna be taking, certainly taking jobs.

[00:08:31] and so I was of course working with, ChatGPT on coming up with some questions for this, and I really liked the phrasing of, um, a couple of these questions around the job. So I’m actually gonna look down here and read it. So if AI makes average comms work incredibly cheap, and we hear this a lot, “I don’t need a comms function because, right, I can now do it with AI.”

[00:08:54] Does being an average comm professional become economically unsustainable? And what does someone [00:09:00] have to be unusually good at to remain valuable?

[00:09:02] Dan Lewis: Sure. So, I don’t necessarily agree with the premise that, AI is going to allow us to give that average comms output, in a way that is valuable enough or even safe enough for companies to use. 

[00:09:15] the, the big concern I would have if I were a senior leader when it comes to communications, and I’m not saying a senior comms lead, I mean a senior leader that has comms support or comms under him or her, Am I getting good judgment from these chatbots? And the answer is just obviously no. You know, chatbots, LLMs are only gonna give you outputs that are as good as the inputs you give them, and they don’t have the context that you as a, communications or even a marketing professional has that you bring to the table.

[00:09:51] Whether you’re, you know, one or two years out of college or you’re, you know, 25 years on the job, it really doesn’t matter. You have a better understanding of what the business’ objectives [00:10:00] are, where the third rails are, what’s been done in the past if you have more experience, or, you know, what you’ve learned about that’s been done in the past if you haven’t.

[00:10:10] The LLMs don’t really have as much access to that, and what they’re gonna give you as a result is a very generic output. The more information you can give them, the more context you can give them, the better output you’ll get. But that requires the, let’s say, quote, “average communicator” to just pivot from, “I’m gonna write down my first draft and then revise, revise, revise,” to, “I’m going to learn how to use these LLMs as a thinking partner and a writing partner to help me train it on here’s what good looks like, and help me get there,” not necessarily faster, because I don’t think that’s what, LLMs are good for.

[00:10:44] But help me get more iterations, explore more variables, and just deal with a lot more data and give me more things to evaluate so I can bring that forward. So I think if you’re [00:11:00] a, if you’re a senior leader and you’re thinking, “I don’t need a communications function, I don’t need as much, as big of a communications function because I don’t need you to draft my all-staff email or to put together talking points for my press interview.

[00:11:12] I can just go to ChatGPT and it’s gonna do it for me.” What you’re doing is you’re ripping out the judgment factor that a communicator will, will bring to the table, right? What you’re doing is you’re effectively exposing yourself to the negative feedback or the pushback you’re gonna get once you start talking because you haven’t explored that layer, and ChatGPT is not there to do it for you, right?

[00:11:36] Like you can’t say to your employees, “Oh, I didn’t think about that. Let me ask ChatGPT for an answer.” And you can’t say in an interview, “Oh, you know, ChatGPT suggested I say XYZ, and we didn’t explore the question you had.” You need someone who can anticipate that, and you need someone who can get you prepped for that.

[00:11:53] and that’s something that communicators are going to do more and more and more. What I am [00:12:00] mostly concerned about when I’m asked that question or what I usually say is, “Look, you can go for it if you want, like try it out.” Um, but there’s another problem that you’re gonna hit, which is that everything you c- that comes out of your mouth is gonna sound kind of robotic and generic.

[00:12:13] I think that consumers, employees, partners in, in a B2B environment, I think the people you’re trying to sell to, the company’s trying to sell to, we all have really good, um, AI detectors at this point where you see certain phrases where you’re like, I don’t see those outside of the context of ChatGPT giving me an answer or Claude giving me an answer.

[00:12:36] This person didn’t take the time to put their own thoughts in it. They’re just regurgitating what their AI is giving them, that’s an immediate turnoff. You know, I wrote something a few weeks ago about how there’s this concept in like computer animation called the uncanny valley. When you create something like a, CGI character that looks almost human, you, th- we realize that no, these are, [00:13:00] it’s really fake, right?

[00:13:01] If you do something like straight up animated, it’s like we can humanize something that’s straight up animated. But that CGI character that is almost human, it really turns us off. I think AI writing or AI slop is getting us there as well. and that’s a huge risk for senior leaders. So, my big advice if, uh, to anyone who’s getting that sort of, you know, pushback from a senior leader is, warn them about those two things.

[00:13:23] Number one, like you need our judgment and you need us to look around corners because they’re too busy. That’s not what they can focus on. And the second one is if they go with, you know, robotic sounding messaging, audiences are gonna detect that and that’s gonna really harm their reputation and harm the company’s reputation.

[00:13:40] Lindsey Groepper: So are there parts of the comms job that you think realistically in, you know, call it three years, that companies will need less people to do?

[00:13:49] Dan Lewis: I don’t know if it’s gonna be less people, but I think people are going to have to think differently and work differently. What I’ve been doing for myself is I’ve created effectively an AI chief of staff using Claude Code. So [00:14:00] Claude has access to my inbox, my chat transcripts, my meetings, and my, uh, my calendar, et cetera.

[00:14:06] Anything that I produce, I use Google Drive, but if you use OneDrive, you can do something similar. So it has access to basically all of the data I create on a day in and day out basis. And as a result, what I can do is have it spin up a calendar of what I need to do from here over the next month. If I use this for a year or two and I have repeatable events, I’ll be able to do the same thing and say, “Okay, I know that we have every quarter we do a big, um, event for potential customers,” 

[00:14:33] All of those are things that as I’m doing my day-to-day work, my AI chief of staff can now say in retrospect, “Oh, here are the different beats that you have to hit. Here are the touchpoints and milestones you have to hit and when.” we typically either don’t do that at all because comms teams are very thinly staffed traditionally, and I think potentially to a fault, or we have low-level people doing that.

[00:14:56] I think what you’ll see is that first we’re gonna [00:15:00] have some job loss there because you don’t wanna hire for something that AI can do. But I think what we’ll actually end up seeing later on is you’re going to have lower level employees whose job it is, is to basically have conversations with these AI chatbots for the operational work.

[00:15:16] Because as you do that, you learn the other hard skills that communicators need to have and the soft skills, like the judgment stuff that, communicators need to have. And that’s just, that’s actually a pretty good way to onboard somebody who’s pretty junior into, uh, the work environment, and I think you’ll get real value out of that.

[00:15:34] So I think those, you know, back in the day it was, the lower level employees were, you know, writing the first drafts of press releases and, you know, more senior ones were editing them and trying to sell them into the– trying to get, executive buy-in or executive sign-off on quotes or positioning or things like that.

[00:15:51] I already think we saw a lot of that job loss pre-AI, where more senior folk were writing the press releases and doing a lot of the… Or the, [00:16:00] or you were outsourcing the media research to companies that, uh, do it for you and, you know, probably can do it better than somebody junior. 

[00:16:08] Lindsey Groepper: what you walked through with what you’ve built in Claude is definitely more, you know, more advanced, right? That is not a skill that everybody has.

[00:16:15] But how technical are they gonna need to become? Do you feel like building, you know, lightweight internal tools will just become part of the, the job description and part of what you need to be good at your function?

[00:16:28] Dan Lewis: I think short term it certainly helps to, uh, in the very least not be technophobic about it. most of the things that I work on are just, you know, natural human, natural English language conversations, uh, very human-ish, but talking to an LLM. I know the technology side of things, but I’m not a web developer.

[00:16:47] I’m not a coder. I’m not a software engineer or anything like that. I speak the language, but I can’t write it. So I don’t think you have to be all that technical. But that’s me and I’m, you know, a data point of one, and that’s very short term. I think [00:17:00] over time it won’t feel very technical. Like the idea that, you know, you or I can, you know, set up this conversation through like, uh, uh, recording platform like Riverside or even send an email or, do a screen recording or a screen capture or share a file digitally.

[00:17:16] All of those are second nature to a lot of us now. And I think 15, 20 years ago they weren’t. You know, we– I think we’re seeing that as, you know, technology becomes more and more accessible and as people grow up with it more and more, it’s less, less scary. think LLMs, because they’re designed to just be a natural language conversation, I think the adoption curve will ramp up very quickly.

[00:17:43] Lindsey Groepper: The real barrier is more curiosity than anything else. I think you really need to be, curious about how you use these tools and say, “Oh, can Claude do this? Can Gemini do this?” And you’ll find more often than not that it can. It can do things that you didn’t [00:18:00] expect it to. Yeah. It’s interesting. I recently had this conversation with my parents because they were asking about AI, and they’re also, like, heavy consumers of news and get, like, deep into, you know, specific threads and channels, and they’re, they’re asking just opinion on AI. And I, much of my answer echoed the same, is that many of the things that 15, call it 20 years ago, if you would’ve told, described how we live today and what we’re able to do today, I would’ve said, “Absolutely not.

[00:18:30] I don’t want to live in that world. That’s scary to me.” And we’re, we’re just in that, that now with AI. And I do, I agree with you that however long that timeframe is, that in 10 years, 15, 20 years, this is just going to be how we live, and we will also be like, “Ugh, in that… Don’t you remember that 20 years ago where we were like, ‘There’s no way that that’s a possibility in this world?'”

[00:18:49] I echo much of that as well. something that you recently did was, uh, teach comms professionals on how to build a comms gut check tool. [00:19:00] Tell us about what this was, um, and what you maybe f- see it unlock for these comms professionals that was, like, well above just how to create good prompts.

[00:19:11] Dan Lewis: Yeah, absolutely. I actually had an interesting conversation with somebody about that yesterday. So, I created a series of courses. It’s– They’re available at build.thecommstack.com if you wanna take a look. Each one is basically a recipe. Uh, I give you the prompts, you copy and paste it into Claude Code, you follow the prompts, and at the end, you get something that’s built.

[00:19:31] so just like you would go into the kitchen and with, you know, the, your ingredients list and your instructions, and you end up making yourself like, you know, dinner for the family, it’s the same real process of it’s like a cookbook for vibe coding, I guess, but for communications tools. So the, the free one, the intro one, is called the Comms Gut Check.

[00:19:49] It walks you through step by step how to create this little web tool where you paste in a draft communication and hit a button, and it analyzes it for, [00:20:00] tone, jargon, potential audience pushback, where you can simplify, and then gives you a version, that takes that feedback into consideration, and it’s just a, a revised version.

[00:20:11] it’s generic, right? It’s not, mapped against your voice or your brand principles or anything like that, but The idea behind it is to show how easy it is to build something that solves an actual problem for you in Claude Code.

[00:20:24] I wanna enable people to build their own solutions using AI, to think about AI more than just that magic search box where you say, “Oh, tell me about this thing,” and it gives you an explanation.

[00:20:34] and more than the, uh, “Here are like 10 disparate thoughts, give me a first draft.” Those are great, like AI is very useful for that, but The ability to use it as a thought partner that can actually code things is really powerful because all code really is, is taking a lot of data sources, connecting them up in weird ways, and running some things which to you and I might look like [00:21:00] magic to get outputs that you never expected.

[00:21:02] that’s something that for communicators and for marketers and for other people who are not developers unto themselves, that’s a sea change. And I want to make that something we can all think about.

[00:21:13] Lindsey Groepper: Yeah, I like that.

[00:21:14] I did want to ask your opinion too, because this just came out yesterday, Claude just announced the, the watermark saying that, uh, things were processed through Claude and people are kind of like losing their minds.

[00:21:26] Curious to get your two cents on what that means.

[00:21:28] Dan Lewis: I don’t think it’s gonna be that big of a deal. I think, to go back to my earlier point, if we were transitioning from typewriters to word processors and somebody said everything with a word processor is gonna do is have a watermark. Well, they kind of do, right? They have file data, like, uh, metadata attached to them.

[00:21:42] But no one’s like, “Oh my God, you used WordPerfect,” now that that exists. But you know, no one says you’ve used Microsoft Word for this. we’re past that point. I think if anything, if I were a senior leader at a large organization and my, um, enterprise element of [00:22:00] choice were Claude, and I were finding that people, that people’s work weren’t– didn’t have that Claude watermark, I’d be like, why not?

[00:22:06] you can probably, you know, take, take 30 seconds, fix the typos, you know, maybe restructure some of the thinking. Maybe you can catch things you haven’t. or even s-ask something as simple as, “Given what you know about the work I do, what are five risks that I should be thinking about as part of what I’m about to put out into the world?”

[00:22:25] Like you should do that. It’s, you know… You’d ask the person next to you to give it a, a second look, but they don’t have time and they’re expensive, and I can see the concern, I really can, but I think that they’re, they’re gonna be so ubiquitous, people are gonna say, “Yeah, of course, everything has the watermark.”

[00:22:42] Lindsey Groepper: Dan, you and I are on the exact same wavelength with that. I was literally woke up in bed this morning thinking about it because there had been chatter with our team yesterday, and then it’s like all, you know, all over LinkedIn with, you know, mostly marketing folks because that’s my world. Um, and my thought was the same as yours.

[00:22:58] I was like, “Who cares?” At some [00:23:00] point, everything is gonna have AI touch it. 

[00:23:03] and it’s not saying it was written by Claude, it’s just that it was processed in Claude, which could be any number of things. But yeah, my, I, tend to agree with you. It’s like, I don’t think it’s gonna matter at some point. 

[00:23:12] Dan Lewis: but that said, like I do think we should be concerned about how much metadata is attached to that watermark. So if the question is, was Claude used in some– at some point along the chain to, you know, make this in some way, shape, or form?

[00:23:26] Fine. If it’s, here’s how it was used in specific ways, that’s probably not great. And I am concerned that that level of metadata could be available at some point. I just think also that’s just, bad from a n- from a, in a lot of different ways. Also, in part because I think we shouldn’t get to the point where people say there are good ways to use this and bad ways to use it.

[00:23:51] And if the, if that watermark or metadata says you use it in a bad way, then, we don’t want to work with you anymore. Because For some people, [00:24:00] you know, I mean, you can think of like people who are neurodivergent, for example, using it in like one way might be a way for them to be more expressive and, and be more, participatory in society than for other people.

[00:24:12] And each, like LLMs are flexible enough where each person should really find their own path forward using them, right? Uh, and I, so I do hope that the watermark is just a binary yes or no, not a, “Here’s how this person used it,” because that gets on the creepy side for me.

[00:24:31] Lindsey Groepper: Yeah, I would agree. Um, well, Dan, this has been awesome. Is there anything that we didn’t tackle that you wanted to make sure that we got to today?

[00:24:39] Dan Lewis: yeah, I can think of a couple things if you want. Uh, so one of the big ones is, um, when talk, we talk about, uh, teams adopting AI for marketing and communication use, and you hit upon this already, like there are a lot of barriers to entry. But most of us think that, like, what should I be using it for and how should I be using it?

[00:24:59] Like the what and the how [00:25:00] are the barriers. And I think what I’ve seen when I’ve sp- and I’ve spoken to a lot of senior leaders about this, the mandate to use AI is almost always just use AI, right? Sometimes it’s use AI to do this, like run a draft through it to catch typos or, as I said, ask, you know, ChatGPT or Gemini or whatever to identify potential risks that you didn’t surface.

[00:25:25] I really don’t like that bottom-up approach, though. I really think that it’s important for any team when you’re trying to figure out how to get your team to be AI-enabled, to be really focused on change management. Like, I think the ADKAR process is really great. Um, if you’re not familiar with that, it’s just, it’s a change management framework.

[00:25:44] but that all starts with why. if you’re not articulating to your team, here’s why we’re investing in this, here’s what we’re hoping to unlock. Like, is it we want to ultimately get down to a four-day work week, and we think we can do that, that’s a lot [00:26:00] more inspiring to employees who are often overworked and would probably like that fifth day, than we’re just gonna make you three times more efficient, whatever that means.

[00:26:08] Or it might be the, we want to hit this goal for our company, and we think we can get there if we are able to turn out, 20 different iterations of our marketing materials for 20 different subsets, sub-sectors of our target pop- or, you know, our ideal client profile, right? If you can take that and do that with AI, that’s great.

[00:26:32] You can’t do it by hand, like it just takes too long, right? So if you start with the why, like here’s the key performance indicator that we’re trying to actually move and in what direction and how much, that’s a lot stronger than just say, use AI. And it sounds obvious, and I think it is obvious, but I think that most companies are really focused on, we need to figure out how to use this new tool because everyone else is, and that’s what is buzzy.

[00:26:57] Instead of just getting down to core principles and saying, [00:27:00] “These are our business objectives, and if we want to re- achieve these business objectives, we need to use AI to do it. So, here’s the objectives and here’s where we want you to focus on accelerating.” Like find two or three specific things and start with that manifesto, right?

[00:27:15] Lindsey Groepper: Yeah, I like that idea of, of flipping the model, and starting with the goal in mind instead of like just use it. It’s like for what and to what end? 

[00:27:22] well, this has been a really good discussion. There’s actually a lot of different areas that I wanted to dive into as well, but we only have a finite number of, uh, minutes to do so, because I, uh, also wanted to pick your brain on, uh, different newsletter formats. But, is that something that you have tackled in a, uh, column yet with Comstock?

[00:27:42] Dan Lewis: You know, I don’t think it’s right for my newsletter to talk about newsletter strategy, but the, the short version is if you’re looking for a thought leadership, Substack’s a good platform if that’s what you’re trying to do. If you’re trying to build more of a media brand, I like Beehiiv. That’s what I use for everything.

[00:27:58] I think Beehiiv is just a much [00:28:00] more flexible and, solid platform. But really the important part is just publish, right? I think a lot of us are really apprehensive about taking our thoughts, putting them in something tangible that people can react to. know, we’re afraid of negative feedback and even embarrassment.

[00:28:16] People are pretty forgiving unless you do something really stupid. I haven’t done anything, I hope, really stupid yet. So, instead of agonizing over, how to do it or what to do, just, Nike said it best, just do it, right?

[00:28:29] Lindsey Groepper: Yeah, absolutely. We just do it and then hopefully end up as big of a brand as, as Nike, right? And just focus on storytelling. thank you so much. Um, as I end every episode, I do the same exact way with the same question, which is, do you have a signature or favorite toast to send us out?

[00:28:49] Dan Lewis: Oh, I don’t actually. Um, so I’ll defer to you if you have one. How’s that?

[00:28:54] Lindsey Groepper: it’s funny ’cause I asked this question and, uh, I have the, uh, regular just [00:29:00] cheers. Um, but then one of the old tried and true is, “Cheers to you, cheers to me, friends for life we’ll always be. And if you should disagree, then screw you and here’s to me.”

[00:29:11] Dan Lewis: Love it. A pleasure

[00:29:13] Lindsey Groepper: That’s, that’s more of my, uh, my non-professional bar cheer, but it’s tried and true. Uh, but no, other than that, I just say cheers, which is funny that I asked this question. So that works for me. Dan, thank you so much for joining me. I really appreciate it.

[00:29:28] Dan Lewis: Thanks for having me

[00:29:29] Lindsey Groepper: All right, you bet. 

[00:29:30] Thanks to Dan for joining me on SaaS Half Full. It was a great conversation. We ran out of time. I had a lot of different things I wanted to pick his brain about, but alas, we have a finite amount of time for this interview. Hopefully, you took a couple things away, including one of the random facts from his newsletter.

[00:29:47] And until next time, bottoms up