Ditching Hourly

DH ### Shawn Yeager

===

Introduction

---

Jonathan Stark: Hello and welcome to Ditching Hourly. I'm Jonathan Stark. Today I am joined by guest Shawn Yeager. Shawn, welcome to the show.

Shawn Yeager: Appreciate it. Thank you, Jonathan.

Jonathan Stark: Did I say that right? We've just met, so I didn't check.

Shawn Yeager: You've got it. It's not Yeager, it's Yeager. Thank you so much.

Jonathan Stark: Okay. Oh, I meant Shawn.

Shawn Yeager: Oh, that part. That's easy, I hope.

Jonathan Stark: Just kidding. Well, thanks for joining me, Shawn. Before I have you introduce yourself, I want to let the listener know that we're going to be talking about AI and its effect on the billable hour today. So stay tuned. I think you're going to get some interesting insights from Shawn and we're going to have a little discussion. So with that said, Shawn, could you let folks know a little bit about who you are and what you do if they haven't met you before?

Shawn Yeager: You bet. So thanks again, Jonathan. Delighted to be here. So the through line for my career has largely been emerging or new technology and getting it to market in short. And so starting back on Microsoft's first browser team, which definitely dates me, and having moved through the SaaS wave, you know, cloud slash SaaS, mobile for the last five years, Bitcoin. And so, you know, sort of on my fifth wave of technology in AI and the work, though my background is a computer science, I got pulled early over into being a suit as it were, not so many suits anymore, but it has been sales marketing partnerships. And effectively, how do you take that technology and convey it and translate it into something that customers see value in, want to part with their precious capital for? And that has, as I mentioned, taken me from Microsoft to Accenture to early stage startups to my own. And that includes my work in consulting over the years. And now what I do is run a firm called Upshift and we are focused specifically on helping professional services firms cross this new chasm to understand what do you sell after AI.

Jonathan Stark: Yeah. So that's great. Perfect. So you sent over a relatively recent article this morning.

AI and the billable hour

---

Jonathan Stark: And maybe let's just tee it up there. Let's start off there because the insight, it's sort of on the one hand to people who are familiar with the weirdness and how nuts hourly billing is. It's kind of, obvious is the wrong word. It's almost like invisible, but it's there. It's like clearly there. Obviously, the whole reason why AI is bad for hourly billers is because hourly billers sell their time and AI or anything that massively increases your productivity gives you less to sell. So it's like a productivity increase is actually bad for you. But the insight that I really liked from this article was that it pulls apart two different things that used to both be wrapped together in the billable hour. So why don't you take it from there and we can just talk it out.

Shawn Yeager: Absolutely. And so, that piece this week came in response to a talk that I've been giving called 'AI Killed the Billable Hour.' To your audience, that may be obvious, and your audience is probably way ahead of this curve, but it is purposely a somewhat aggravating title, agitating title to get a conversation going with my primary audience, which, as we mentioned, are professional services firms, owners, senior partners, etc. And so with that, I'm explaining that billing by the hour, as I think I noted, will probably outlive us all, but the billable hour as a unit of measure, as a unit of value, presumably has now been exposed. And so that which was hiding in the billable hour is now out on display as AI crunches and crushes and shortens and reduces the time to completion. And I think the perceived value of that work. And so what is left, it's almost a cliche now, so many are saying it.

The judgment sandwich

---

Shawn Yeager: but I do believe it's judgment. And I, I think of, uh, I was at a fireside. I was part of a fireside discussion a few weeks ago alongside a very talented lawyer who used a couple of terms which inspired this. One is that, uh, what now is delivered as a judgment sandwich. Uh, and if, if you're employing AI, which many of your audience being, uh, software engineers and, and similar may certainly already be doing, you know, that the output is only as good as, as the direction you give it, the harness, the, the structure, the prompt, the planning, and then on the tail end, of course, you are left to determine, is this a suitable outcome? Do I, do I take the work product and, and go with it? And so there is that judgment sandwich. What is in between, um, is now largely, uh, commoditized. The other thing he said, which I thought was interesting, and this is specific to legal in his remark, but I think it's broadly applicable, is that much of this is now, well, it has been white collar manufacturing. And so across sectors, across industries, you know, um, and this is a guy who I'm sure probably charges, you know, bills at a thousand dollars an hour, which is in itself a bit, a bit ironic or, or, uh, interesting. Um, but what is, what is now left is, uh, to, to sort of break that billable hour down and to see what can be automated, what, what pieces can be put together in that sort of factory floor. Uh, but the premium

Strategy, execution, and hidden value

---

Jonathan Stark: ...is the judgment. So, uh, when I read this, it maps to something that, that, um, David C. Baker, myself and others have, uh, pointed to, which is the difference between strategy and execution. Sometimes I call them different altitudes in the software space. If you're familiar with code, let's just say there's three basic altitudes you can operate at. The bottom one is support and maintenance. The middle one is execution or implementation. And the top one is like strategy or design, the architecture, those kinds of things. And that top level one, the one where you need all the judgment or taste, that's still where value lives. And the execution layer is getting nuked in the software space. I mean, it's actually not bad if you bill on some sort of fixed basis, not by time, but if you do bill by time, you're getting nuked. So what the connection that you made for me in that article was how it's all hiding in that hour. And it reminds me of sort of my early days of noticing this when I first went out on my own and was giving fixed prices and value-based prices. I was like, Oh, I've been giving away the, what I would consider to be the important, the judgment, the reason why I'm better than someone who has less experience. The important part, I'm charging $150 an hour and then typing semi-colons that anybody could do is I'm also charging $150 an hour. And it just sort of works out in the wash that like, okay, you know, there's probably a lot more typing semi-colons that needs to be done in terms of calendar time or clock time than the making the decision because all of the making the decision stuff happened when you, you know, previously, like you've already learned that stuff. And so it's right there ready for you to use. So it doesn't take that long because you don't have to like relearn it every single time.

Shawn Yeager: And you tell me, but I would also assume much of that was up in the discovery phase early in the project, right?

Jonathan Stark: Yep, exactly. Like really picking the architecture, the stack, the direction, everything at the beginning based on, uh, and making sure that aligns with what the business actually wants, like what the outcome, first of all, uncovering what the business actually wants, because a lot of times they just give you a punch list. Here's a bunch of things we want you to do, go do them. And I would always back when I was doing this kind of work, I would always push back and say like, well, we can do all that, or I can do all of that for you, but why do you want that done? Cause they're really not the expert at it, which is why they need me in the first place. So I would uncover the business constraints and goals and vision and everything first. And then only then would I decide like how I was going to contribute or what I was going to do to help achieve those goals. And if you're billing by the hour, there's no need to do that. You know, there's an old consulting joke. You probably heard it. Uh, it's like, uh, the client signed on the dotted line, you start billing, I'll go find out what they want. Right. And that's me. That was really a perfect encapsulation of, of it. So to bring it back, I love how you boiled it down to how it's all hiding, like in the same hour. I used to think of it as like there's strategy hours and there's execution hours and that's still sort of true, but I like the idea of thinking of it all living inside of that unit. And there's like really valuable things in there and there's less valuable things hiding in there.

Shawn Yeager: Right. And I just like pulls the lid off. There are certainly, um, I don't know that I would have historically counted myself in this category, but there are those who are excellent at context switching and, and maybe they, maybe they do have different kinds of hours, but yes, I think by and large, even in the most lofty.

Jonathan Stark: You know, sorts of environments where it is a $1,200 hour. I think it's, it's, it's hiding a lot.

Client conversations about AI pressure

---

Jonathan Stark: Right. Okay. So when you're working with clients and maybe you described this to them and they either, maybe they don't have a light bulb moment, but they do see the downward effect on their revenue. I mean, frankly, you know, it's like, we don't have a, we, uh, we're going through work faster than we can get new work because we're using these new high productivity tools. And, uh, you know, I often joke, it's like, why would you buy a faster computer? If you bought it by the hour, it doesn't make any sense. It costs you the money for the computer and you get penalized financially for being faster. So what, when you're talking to someone who's maybe on the fence, they see like something's wrong, but maybe they haven't quite, the light bulb hasn't quite gone on yet. Are they, are they, what's the conversation there when you're trying to, you know, you're in front of a room full of people who maybe are partners at big law firms? What are they thinking? How do you talk to them to turn the light bulb on? And then like, what do they do once the light bulb is on?

Shawn Yeager: I think a lot of it in law, law is interesting in that I think it generally enjoys protection from historically a lot of this downward pressure. And to your question, there are those who would respond, you know, we are a brand firm. We are a boutique firm. We have X, Y, and Z extremely high profile partners, shareholders, senior attorneys. And that is certainly a sensitive conversation. You know, you don't want to walk in and let them know that their baby's ugly. And I don't, but the point goes back, you know, to what you've said, I think, which is that it will only protect you for so long. And so I won't, you know, maybe regale or spend the time today walking through it. But in every of these sectors, I will point back to what many have now seen, which is Sequoia, the sort of storied Silicon Valley venture capital firm. And I wrote a piece about this titled, you know, Sequoia put a trillion dollar bounty on your business. And that is to say, they put a call out to startups for a particular fund that they have put together to invest in startups that are going after what in some is a trillion dollar opportunity. And the long and the short of it is they're funding startups to sell the outcomes, to sell the work that professional services firms do. And again, back to your question, there are those who will respond well, in the case of law, legal, we have Harvey, you know, we have Lagora, we have, you know, brilliant firms doing great work. But the point is you and your competitors are all buying the same tools, right? So it is a race to the bottom. And it goes back to your point about why would you buy a faster computer? Well, you know, why would you buy Opus 4.8 instead of 4.7, you know, which is going to burn more tokens and deliver the answer faster. So, you know, I think it is largely about their perspective. You know, are they an individual who's at the end of their career and they just want to ride out the next five years peacefully and who am I, you know, who am I to tell them not to write someone else's problem?

Jonathan Stark: Right, right.

Shawn Yeager: And so, but presumably there is someone junior or, you know, less senior at least coming up under them to take those positions, to fill those shoes, and they will have to worry about it. And so, you know, to accounting where I think they largely get it, and many accounting firms have made migrations or moves to fixed pricing, to marketing, which these agencies are being really hit just right in the midsection because the deliverables are much easier to produce if, you know, certainly maybe not at the same quality. And again, it gets back to judgment and taste and the things that you and I have talked about already. But ultimately, I think it comes down to where are they in the life cycle of the business, in the life cycle of their career? You know, are they looking at how does one bring in, as one senior lawyer told me, he is evaluating hiring systems engineers instead of junior associates at 300k a year. And so, my first response is, wow, I didn't know junior associates got paid 300k a year, first year associates, I should say first year. But, you know, he now quite shrewdly is looking at how do I instead bring in a systems engineer, build the scaffolding, the workflows, the routines,

Validating AI output

---

Shawn Yeager: All of that to be able to compress down and accelerate the muddy middle. So, long answer to a short question, it's ego, it's stage and career, stage and role. And I think it is pressure from clients, which varies by sector.

Jonathan Stark: Okay. Yeah. Quick comment on the marketing thing. And it actually addresses a broader AI concept, I think, for users, which is that if you can't, the extent to which you can validate the output of the LLM defines how useful it actually is.

Jonathan Stark: Yes. Yes. So, if it's going to output a bunch of code for a website, let's say, which I could have written myself, I just didn't feel like it. I can read it and say, yeah, I know that's bad or fix it. I know if it's garbage or not. If someone pukes out a 50-page contract or red lines an NDA or whatever, some contract, but they don't have the expertise to know if it's hallucinating or if it's garbage, then it's kind of useless because then you have to have it reviewed by an expert to validate the output anyway.

Shawn Yeager: Absolutely. And in the marketing, everybody thinks they, you know, I know what I like and I'll know it when I see it. And so you can iterate endlessly with, you know, an infinitely patient, even if it's only intern level marketing AI, that's, you know, putting out brand guidelines and colors and styles and fonts and all sorts of things. I know brand is more than that, but, but, you know, if you're doing like Facebook ads, you can be like, oh, that looks like a good ad that you think you can validate the output, even if maybe you're not that great at it, but you can be deceived into thinking that looks good. I would click on that ad. I'm just going to use that.

Jonathan Stark: Right. I suspect there'll be a little bit of a pendulum swing back when someone does spend $10,000 on Facebook ads that don't work, that they generated themselves. And then they're like, oh, maybe I should talk to an expert, but okay.

Judgment, marketing, and cost of being wrong

---

Shawn Yeager: So, well, and I think if I may, I would say, no, I think you raise an excellent point and you know, not to get too deep in the weeds, but there is a plugin called Marketing Skills by a very talented fellow, Corey Haynes, and impeccable style.

Jonathan Stark: Oh, excellent. Wow.

Shawn Yeager: He has, I don't know him. I just know his work and he has produced Marketing Skills, which is incredible. And, you know, for example, it does a brilliant job of conversion rate optimization, of cold email writing, of ads, of video. Now, you know, again, better in the hands of someone who knows to your point, but I think, you know, what that brings up is I would like and do pay someone for their pattern recognition, their experience, their insight, their judgment, ultimately in concert with doing some of the grunt work myself, because, you know, running a startup, it's budget friendly. So, I think, you know, depending on the sector, we'll see more or less of that. But, at the end of the day, I think what remains and the question that I ask many clients and the folks that I speak with is, what's the cost of being wrong? And I think you've just touched on this. So, you know, the contract versus the Facebook ad that flops, right? So, you know, I don't catch a red line or I don't respond to a red line properly in a contract that could be devastating, versus an ad that flops, you know, two very different outcomes.

Jonathan Stark: Exactly. Right. Is there an emerging framework that you or Accenture or McKinsey or...

AI transformation and commercialization

---

Jonathan Stark: Somebody is, it's kind of like the new digital transformation, right? It's like, now we're doing AI transformation. Let's do that.

Shawn Yeager: Yes.

Jonathan Stark: Okay. So what are the broad strokes there? Because for my people, it's kind of like they just decide and then they can start doing it. It's not like they have to convince a thousand people.

Shawn Yeager: Absolutely. And in fact, I would just say quickly, you know, one of the sort of points that I call out is if you're a solo or, you know, boutique, you're nimble and you can respond and you can be agile and you can try things. If you're a big four, they have collectively now put $10 billion into AI. If you're in the middle, I think that's where it's tricky. And, you know, where I focus, where we focus is in the commercialization. And so, you know, not a pitch, but a framing is to say the workshop that I run is about.

Shawn Yeager: For those who have that background, it is two things. Effectively, it is business model transformation from the Alexander Osterwalder. Apologies, Alexander. I had a great opportunity to sort of train under him, if you will, back in 2010, and I've been using that framework ever since. So, there's business model transformation, and there's customer discovery, product market fit. And so, call it what you will, but ultimately, this is about scoping, specing, designing, packaging, pricing, new offerings that you test before you go build it. And so, all that to say, you know, I'm not rolling in to say which tools to pick and which workflows to automate and how to architect these systems, but rather what will endure and survive and what will your clients value. And so, I look at this in sort of three frameworks and four arcs, if you will, and one is sort of the delivery shift. So, the two here are moving from episodic to continuous, sort of, you know, discrete projects to always-on delivery, from reactive to proactive, which is, you know, waiting for the call to surfacing problems first. And that might look like having an always-on agent that's looking for a compliance problem or brand fidelity. It could look very different in different sectors. And then, in pricing specifically, as your audience knows all too well and you've, you know, been doing now for 15 years, I think, is hours to outcomes, right? And so, from billing time to billing results, and that's the pricing shift. And then, moving up the productization ladder. So, from bespoke to scalable and a few levels in between. But it's custom work per client to codified platforms and, you know, there are no silver bullets and you don't get from A to B or A to Z rather immediately, but those are sort of the four spectrums that frame the conversation and the work that we do.

Jonathan Stark: Hmm. Interesting. Okay. So, you remind me of a side comment. It's related, a little different though, is I just watched a video that was on Lenny's podcast, which tech people will be familiar with, and it was an interview with Benedict Evans, who's been a tech analyst commentator since way back in the mobile days.

The hard parts big firms still need humans for

---

Shawn Yeager: Absolutely.

Jonathan Stark: Yeah, I've always been a fan.

Shawn Yeager: Oh, is that right? Okay.

Jonathan Stark: I'm a long-time subscriber. Yeah. Great newsletter. So, one of the things he pointed out, which I thought was really interesting, is that these big four, Bain, McKinsey, you know, whatever, they should be firing people, but they're hiring like crazy, according to him, I don't know, but I believe it. And he was like, well, you would think if you're going to get 150 experts for free from Anthropic or OpenAI, then why would you hire more experts? You should be downsizing, right? And he was like, well, the question really is, what's the work? What's the hard part? The hard part, I mean, maybe the time-consuming part was creating the 75-page deck that you're going to present at the end of the engagement. Is that really the hard part? That's the deliverable, maybe, but the hard part is going to the factory, walking the site, interviewing the stakeholders, you know, having the judgment on both ends, and that's still the hard part. And people, companies, their clients, clients of these big companies don't have a bunch of people sitting around doing nothing. So who's going to do this AI integration project or whatever project, whether it's AI-related or not, who's going to do it? And if there's a lot of consulting work that comes up, then the hard part that the McKinneys of the world can do, or are famous for doing, they can't staff up to do that. They're not going to staff up to do that. It still has to be, there still has to be a person doing big pieces of this, even if the deck gets generated in three minutes instead of three weeks. And oh, by the way, the client's just taken the 75-slide deck and jamming it into Claude anyway, you know, to get the TLDR, right?

Jonathan Stark: Right. Yeah. So I just thought, I thought that was sort of a related point where, because the next question I was thinking of is like, what, what are these organizations, you know, so let's say you're talking to someone who's in a situation like this, not a consulting firm, but like a manufacturer, let's say, or something like that.

Cost recovery versus new offerings

---

Jonathan Stark: And they are, are they coming at it from a, this is going to allow us to decrease our costs by lowering headcount? Like you said, the lawyer, you know, the $300,000 a year, first year associate, or are there, what's the percentage of ones that are thinking like, oh, this will be a great cost cutting mechanism and versus the ones that are like, this is going to allow us to deliver a product or service that couldn't have existed before?

Shawn Yeager: Yeah, great question. And I will say that though my focus is on professional services, I'm adjacent to other sectors by way of firms that I partner with who do delivery, who do the build, the build, design, build, run. And, you know, what I'm seeing and hearing from them consistently is it is squarely about cost savings, cost recovery in the beginning. And so, you know, I think stage one or early stages of this are examples of we receive, you know, we receive 50 spreadsheets a week from suppliers, you know, we as a distributor have to somehow ingest all of that. And so that's, you know, exercises like that are very low hanging fruit. I think most have not yet thought about what do we sell now that we couldn't before. And, you know, should I say this in public? I don't know. Maybe I'm early, right? You know, because I look at others who are doing the build and good on them. You know, they're going in and just building a punch list of, you know, there's one firm here in Nashville, Nashville Automation Company. And I think the tagline, I know the founders, so disclaimer, but it is, you know, make Monday suck less. I think it's something like that. Right. And so, you know, that value prop is pretty clear. And they're going in and they're just tackling these, you know, these grueling workflows and processes. So that is all to say that I believe most are looking at cost recovery, cost cutting. I think it is, you know, the leaders who are out there thinking about what do we offer now that we couldn't before or couldn't profitably, the unit economics didn't work out. And then the last point I'll say about the hiring, I'm also a big firm of a big fan rather of a company out of New York called Every.to. Great, great resource. And so as you as you would know, as a reader, Jonathan, they talk really about giving every, you know, they're a 25 person firm. So great, you know, small, nimble firm who just produce crazy amounts of content, product, etc. And so everybody gets an agent, you know, and and that I think goes back, not at the BCG, or the McKinsey level, perhaps. But the wonder, I think, in all of this is for a small to midsize firm, it takes work to get there. But, you know, I often say only somewhat tongue in cheek, what would you do with another 10 employees? Or in the case of an agent, you know, a hyperactive, extremely smart, you know, slightly less focused sort of intern, right, which is where I think we're working to build these agents out. But I think that's, that's ultimately where it seems to be headed is not so much do we do we lay off, although we read a lot of that in big tech. But what can we do now with everyone having a chief of staff sitting beside them?

Agents as extra employees

---

Jonathan Stark: Yes. Well, it's funny that you said chief of staff, because maybe I don't understand what that title exactly means, having never had one. But you also said intern previously. And to me, the mental model, this is what maybe more so with soloists, I don't know. But the mental model seems to be people trying to outsource the thinking, the high-level stuff that they're actually good at, to an LLM versus the tedious administrative stuff.

Jonathan Stark: That no one probably wants to do at all. Right. And I absolutely use that intern metaphor all the time. It's like an entry-level employee that needs to be treated with that kind of attitude, just from a security standpoint, from trusting the work output standpoint, from if they fall asleep on the job, because sometimes they just, like the server, nods off or falls over.

Interns versus chiefs of staff

---

Jonathan Stark: So they do need a little bit of handholding and management and babysitting. But it's a lot less than doing the work of collating all of those inventory reports.

Shawn Yeager: Absolutely.

Jonathan Stark: Right. So, I think it's a failure mode for someone who's a high-level strategic thinker to try and outsource high-level strategic thinking to any of these models because it seems like they are trained on such a large pile of data that it's almost for sure that they're going to kind of regress to the mean.

Shawn Yeager: Absolutely.

Jonathan Stark: And strategy is almost like not allowed to do that, basically. It's like, okay, you have to come up with some high-level approach to winning something by applying your strengths to weaknesses. It's almost certainly going to be something no one else would have thought of or committed to, which you're not going to find by summarizing the internet. By definition, it almost has to be novel.

Jonathan Stark: So just to wrap that, put a bow on that, the chief of staff thing, I don't actually know what a chief of staff does, but it sounds like an executive level position. But if it's an executive assistant, then okay. I'm not sure if the metaphor applies or not. If that's true, then yes, I would agree. But if the chief of staff is someone that is applying a lot of judgment and has a lot of autonomy to make decisions on their own, I don't know if it's my favorite metaphor, but I've heard a lot of people using it.

Shawn Yeager: No. And I think I should be clear. I think where we are is largely, well, on a spectrum, I think we're closer to intern. I think the promise, and I think that there are some who are pushing it closer to that chief of staff in that, as you allude to, they are a more trusted confidant. They have more leeway. They have more authority. They have more budget. They have more resources at their disposal. And I think just in my own personal experience, again, many people will say this, but it is...

Scaffolding agent workflows

---

Shawn Yeager: It's all about the scaffolding, right? And so you can pull me back from the weeds here, but the more time I have invested in skills and plugins and, you know, once again, Every.to, the way I found them was their compound engineering plugin, which for any software engineers, you know, I only play one on TV anymore. It has long since, you know, gone away as my career. But enough to know that I can now, with artifacts like product.md, design.md, the seemingly innocuous, you know, shout out to those who've been using Markdown for 15, 20 years, it's paying off, right? But the point is, those seemingly innocuous artifacts now, back to your point, are the guidelines. They're not infallible, but they're the guidelines that allow you to bake more of the judgment in, you know? And so I won't sort of, you know, go into a whole lot of examples, but I have just seen that the more I invest in that for certain projects, the more predictable outcomes, and I do now have processes running overnight. Now, I will say, you know, I think OpenClaw and phenomenal breakthroughs by Pete and the team who did that, it's largely a LARP, you know? I think spinning up a company, you know, there's another project called Paperclip, brilliant. I think it's a fascinating look at where things are going, largely a LARP.

From chatbots to delegated workflows

---

Shawn Yeager: But I think it is possible to have very specific focused agents that run as a GitHub workflow, as an action, you know, and they do these targeted things. And in aggregate, you get some pretty great, great outcomes.

Jonathan Stark: Let's drill into it for a second in case, because I know people listening probably, I don't know how deep you are into the OpenClaw space. I'm somewhat deep into it. And I'm a nano-claw guy myself. I have landed there.

Shawn Yeager: Yeah.

Jonathan Stark: Okay. So, you know, Pete's design goal at the beginning was for it to feel magical, not to be secure, not to be able to create lots of other things. So that wasn't a core design decision. So he wanted it to feel like magic and boy, did it when I first tried it. And the one thing I'll say, but I am, you know, OpenClaw is, the concept of it is going to get copied and refined, and you've already listed some, you know, CloudCore came out like two seconds later.

Jonathan Stark: And one thing I'll point out to people listening, if they're sort of surprised that there's so much talk about AI and it's not reflected in their experience of it, and they're kind of like, why is everyone thinking this is so great? Like, yeah, it can help me like craft an email to a tough client or, you know, like my accounting firm or something, but it doesn't seem that great. And I think, I think to those people, I would say that, that when people are saying AI, they're talking about wildly different things.

Shawn Yeager: Yes.

Jonathan Stark: And one of the huge differences, there's probably two, well, maybe three. Okay. I'll give you three, dear listener, if you haven't experimented much, I'll give you three big up levels that will improve your experience so that you can kind of like, perhaps get a sense of like what people are talking about when they think OpenClaw or some competitors, why they think it's so cool. First if you're not paying for an account at all, and you just use OpenAI or, you know, Cloud once in a while, it's probably not that great. So once you have a paying account, if you're paying, then it gets better right away. And it really doesn't have that much context about you if you don't use it that much, but if you use it a lot, it has tons of context about you. And if you're a guy like me who has 10,000 markdown file, literally 10,000 markdown files on their website from 10 years of writing, and you give it access to that, it can produce incredible things like, you know, what have, what have I not written about that I talk about all the time on my podcast, just boom, giant gap analysis that it would take weeks or maybe never, no one would ever do. So it can do some really, really incredible, useful things. And then the level up from that is when it can write, write files, it can move things around, it can interact with APIs on your behalf, not saying buy things yet, I'm not there yet, but when it can move files around and do all of this like digital filing for you, it's like a boom, another huge level up. Maybe especially from people like me who don't have employees or assistants or anything like that. And I can't just say, Oh, take all of these files and rename them such that the file names are in reverse date order and then upload them to my transistor website. Cause now that we've got all of these podcast transcripts that we never had before, like I would just never do that. I'm not going to hire someone to do that, but if I can just type into telegram what I just said, and then it's just done, it allows you to do things that, yeah, it's just wildly different than like asking ChatGPT to write a book report for you.

Shawn Yeager: Absolutely. It's kind of the reference to the one shot versus a workflow. Right. And so again, perhaps seemingly trivial, when we have wrapped and you're good enough to follow up and send me the YouTube link, I will drop into Claude Code. I will use a slash skill slash press. I will give it the link. It will build the page top to bottom, everything, including the transcript, including the nice overlay that sort of blocks the, you know, the preplay on YouTube, all that stuff. Why? Because, you know, I'm a nerd and I like things to look the way I like them to look. But the point is that was a very specific bit of work, as you've illustrated, that I, you know, I don't want to move a div tag. You know, I've given up on that long ago, but I can point a single skill at a single link. I do the same thing for research, right? I'll take the latest McKinsey report, use a single slash skill, and it goes into a thesis file. You know, it will flag me when something, quite frankly, you know, goes against that thesis. I, as you, with your longstanding business, have certain hypotheses about the way things are going to play out, and I want to be challenged and I want to refine and sort of sharpen the tools. So, you know, research is another great example. But to your point, it's a whole other level than, let me jump into ChatGPT, which is, you know, now a great, but, but really effectively used by most as just a replacement for Google search.

Jonathan Stark: Exactly. Right. So if you're, if you're just operating at that level, there are a whole other worlds that people are, they're in people's minds when they're thinking about, you know, AI, when they're talking about this stuff, it's like, it's much more like having an employee. It feels like even though I'm telling it to, it's essentially automating things for me, it feels exactly like delegating. It doesn't feel like I'm automating anything. It's really, really different. So where do people go next? Like these bigger, I'm, I am really curious about bigger firms because sometimes I do get these, you know, people who've been on my list and they have a friend who's at a bigger firm and they email me and they, or whatever, and it's just not, that's not my, that's not my lane. I want to get into leadership things, I don't want to get into culture things. That's that's a different ball of wax. So when someone is going through a company that size is going to go through a transformation like this, I agree that it makes tons of sense that like 95% of them would be like selling it internally as like cost reduction or cost, what do you say, recovery. So like what happens? Like what do they do? They hire you or or someone like that to come in and just essentially it's like a straight up consulting gig where, but with this AI transformation concept and you try and try and deliver like one working proof of concept to demonstrate the savings and then build from there. That's what I would guess.

Shawn Yeager: And I think for a lot, you know, I would not, I would not advise against that, right? I think hands on keyboard, you know, this is the new equivalent of hands on keyboard, which is to your point, moving beyond ChatGPT into a terminal and using Claude Code or Codex and getting a sense because it is massively more powerful.

AI adoption inside firms

---

Shawn Yeager: And as you've alluded to, within the confines of what you allow gets access to your local file system. And so however you choose to do that, by all means do that. The trick I think often is you've got 30 employees. I'm working with a financial services firm right now and about 120 employees and 30 or 40 of them have various tools and they're seeing benefits. They are not, however, because of various complexities and financial services, which is its own thing, they are not yet sharing that product.md, that design.md. They are not, uh, you know, they're not submitting PRs to a project. They're all just sort of experimenting. So the trick in all of that is it, it hampers or blindfolds your experience of what's possible because there's just only so much one individual is going to get done, you know, either on their machine or in a VM or however that looks. So, you know, first get hands on, get some individual experimentation, and then within a particular function or department, um, pick a project, you know, pick one that is, and, uh, shameless plug, you know, I do at upshiftco.com have an assessment, which among other things sort of exposes how, or rather measures how exposed a firm is to the billable hour. And one of the other, you know, components of it is, all right, so if that is true and if there's a lot of billable hour, you know, uh, exposure there, where is most of the activity and does that activity require senior involvement? Now if you're a solo operator, you're in everything, right? So to your point, this is for, this is for, uh, small to midsize, but you know, if you've got that exposure and it's a lot of your work, but it doesn't take senior, pick that, right? Take one of those projects and see what's possible. And then ultimately, you know, what, what I'm delivering with Upshift is a two-day engagement, a one-day workshop, and it's get the senior partners in the room, work through service design, work through pricing, work through everything that at the end of the day delivers two to three offering briefs. And those briefs are packaging, pricing, positioning, the first three moves, you know, who are your existing clients that again, to those who've done customer discovery and product market fit work would recognize as let's go test, you know, and you can, you can zoom out even further and just talk about pain. But if you're an established firm, you're probably thinking already about what's the solution. So you can go test that, see if it's a fit and then build.

Jonathan Stark: Okay. It's, I am curious for a little bit more detail about the state of affairs when you typically enter into one of these companies. Is it, here's what I think you're implying and you can correct me where I'm wrong. It sounds like leadership was like, everybody gets a cloud account or everybody gets an open AI account. You're allowed, you get this much budget or whatever, it's a $20 a month account. Just do whatever you want, play with it. So like basically individual at the individual level, silos of expertise, but growing or not around usage of the tools, is that more or less?

Shawn Yeager: It is. And I think it's twofold. I think, I think what's driving that is one is positive, which is, I think let's get, you know, again, let's get hands on, let's get familiarity. The other, I think often is a response to an uproar from the, from the employees saying we are at a distinct disadvantage because we're not using these tools.

Jonathan Stark: Really? Not, not so much the reverse, not so much the employees being like, I hate it. It's killing the planet. We don't want it.

Shawn Yeager: There are those, there are those, you know, and I think not to pick on anybody, but if you go into design, you know, if you go into sort of the design function arm department, you're going to hear, get the slop away from me, you know, I don't want anything to do with it. Um, if you're perhaps in, you know, sort of a go to market function where it's look, I need to source leads, enrich that data, I need to automate a lot of the outbound cold outreach and all these things, it's, it's phenomenal. So depending upon who you ask and the other piece, just to wrap up to your point is there's a lot of, we need an AI strategy, right? Which, which I would argue is a bigger, faster version of, we need a mobile strategy. We need a cloud strategy, you know, we need a crypto strategy. We need a, and so, you know, they all have their own flavor, but I think the, the friction, the challenges ultimately back to your question, uh, it is we're on the back foot. We got to do something, uh, or a bit of an uproar from the employee base. And my position, again, not always the most popular is to slow down, step back, think about how you're actually going to commercialize this, not just cut costs.

Jonathan Stark: Yep, exactly.

Closing remarks

---

Jonathan Stark: Well, this has been amazing and I would like to let you know in case you didn't see it, that on the cover of the Financial Times, I think it was a few days ago, there was a pretty big article about how AI was killing the billable hour. And it was citing sources from large legal firms. And I mean, that's about as mainstream as it gets as far as I'm concerned.

Shawn Yeager: Thank you for that. I'll go find it.

Jonathan Stark: Yes. I'll send you, I took a photo. I almost said screenshot. I took a picture of it because I get the actual paper one, so it doesn't, I don't like stuff beeping at me when I'm trying to read. So, this has been fantastic, Shawn. Thanks so much for joining.

Shawn Yeager: My pleasure, Jonathan. Thank you.

Jonathan Stark: Why don't you give everyone that link again so they can go check out that assessment or maybe forward it to someone that they know is in a bigger firm that might need your assistance.

Shawn Yeager: Absolutely. And it is upshiftco.com. Upshift is the business name, upshiftco.com is the URL.

Jonathan Stark: Fabulous. Excellent. Well, thanks again.

Shawn Yeager: Thank you, Jonathan. Pleasure.

Jonathan Stark: All right, folks. That's it for this week. I'm Jonathan Stark, and I hope you join me again next time on Ditching Hourly.