Subscribe
Share
Share
Embed
Do Good Work founder Raul Hernandez on how to answer the client who asks "can't you do this cheaper with AI," why agencies should own their data and agents instead of renting a model, and four new ways to price agency work in the AI era.
Agency Forward explores the future of agencies as tech and AI drive down the cost of tactical deliverables. Topics include building competent teams, developing strategic offers, systemizing your business, and more.
New episodes delivered every Tuesday.
Chris DuBois 00:00
No one was asking for another community, but I've made one anyway. So what's different? The dynamic agency community is designed around access rather than content. Access to peers who've done it before. Access to experts who've designed solutions. Access to resources that have been battle tested. And right now, the price for founding members is only $97 a year. Join today so your agency has immediate access to everything you need to grow. You can join at dynamicagency.community. It's easier than ever to start an agency, but it's only getting harder to stand out and keep it alive. Join me as we explore the strategies agencies are using today to secure a better tomorrow. This is Agency Forward. All right, one of the things my clients bring up all the time-I say all the time, frequently enough-that I have some concern. It's when their clients are saying, "Can't you do this cheaper with AI? Are they bringing up that conversation? How do you attack that?
Speaker 1 01:10
Well, isn't the answer? Yeah, we can.
Chris DuBois 01:12
Yeah, that's what that's what I say. Like maybe that's the answer.
Speaker 1 01:14
So I think the thing that's changed, and we even see this in like with with the big four consulting groups, is that end clients are asking for outcomes, not just done work done, and because AI is able to execute a lot of things now, some of the hands. The question becomes, what are you actually doing for me, for my problem, for my business case? So yeah, it can do it cheaper and it can do it faster, but is it right? Do you align it to an outcome, and also here's the thing: these whichever model you use, use them today when they apologize to you. Oh, sorry, I screwed up. I won't do that again. But you're on the hook. Who's on the hook? You or the AI? You are. You have to look at the client, saying, "Here's why we didn't hit the number, or here's why this experiment failed. Which is fine. It's part of it. It's part of the game plan. We experiment to find what works and what doesn't work, or here's why this worked, and here's how we're going to move forward. So there's tension there that's happening, but I think that's because value and services have to move up the stack. They have to go from execution up to not just judgment of like here's my empirical experience about what to do to having a shared belief with the client, a shared belief of the future of the work that we're doing and what's actually possible, and that requires-and this is like an ethereal-everyone has to be the strategic partner. In reality, it just requires a relationship and future pacing for what you're actually working for, not just delivering assets or delivering websites, or delivering copy.
Chris DuBois 02:43
The all right, we got a bunch to get into now. I want to start though with the idea of ownership because I do think that maybe not as like you see it in the agency relationship and stuff, but it also just plays out with how people are using AI for themselves with their LinkedIn content. It's like they could sit there and think through a strong LinkedIn post and write it, or maybe even just work with AI to get a great post. But instead, they're trying to build the system to get get the content out, and then they sit there and complain that they're not getting engagement, right? Impressions are down, all these different things, but it's like, yeah, but you're just you gave this to AI,
Speaker 1 03:24
yeah,
Chris DuBois 03:24
right. So like, how do you? I mean, there's obviously a difference between the people who are going to take ownership and say, okay, well, obviously it's not working. What can I do to improve this so that I'm actually getting the traction I want with or without AI, versus the people who you know are just kind of they're trying to one prompt this. They're trying to just say like, "Oh, I'm doing everything I should be, but not getting the result. Like, what what's your response to them? Like, how do we how do we get them to see the light here and own this?
Speaker 1 03:53
Well, I think the thing we're talking about now is craftsmanship and quality. And I said this even before AI is that you have to think about your own thinking, and you know how many founders I've told that to, plenty. How many have actually followed through? Few. Even one of the dudes that made it to the Wall Street Journal bestseller, because it's it takes time. He had no time, granted, so it was a legitimate excuse. But everyone has no time to think about their own thinking, metacognition, and I think that's really if you're going to outsource your your edge, then unfortunately you will be replaced or augmented by something else that can that can do the same thing. The key right now isn't doing more, and I've actually stopped posting on LinkedIn consistently because of it. It's been like a sloptopia, and I'm focusing more about because I'm able to now work and literally grow and do things in the in the company for myself and for clients at the speed of my thoughts. My issue is in capacity. The issue now becomes quality and judgment based on the quality. Alignment, seeing if I can just do everything. If we had Taylor Black on the podcast, he's at the Microsoft Directors of CTO office. Like he's got an interesting title, but Taylor Black from Microsoft. He said mere impossible problems in the past that seemed to be impossible are now merely tough.
Chris DuBois 05:19
If
Speaker 1 05:19
that's true with this technology, let's just call it technology for what it is. If this technology takes something that used to be impossible, now just something that's actually hard to do, but it's still possible, I think it's it's left of just because you can do everything doesn't mean you have to do everything. It's what should you be doing in relation to not only just your state in life, but your your clientele, the impact that you're making, the value that you're creating, as well as the markets that you serve,
Chris DuBois 05:44
right? Yeah, I think discernment has become that like that thread through this thinking is a power tool
Speaker 1 05:50
now. Like
Chris DuBois 05:51
right, the because even with my own business, it's like the one of the pieces of like we figure out the my client's positioning, and then we have to start building those assets. Like, how do we get your positioning into the world? Part of that's always going to be the website. And last year, even the year before, it was like, all right, well, call your developer. Like, they have to get building on this site, right, so that we can get all the stuff we want out there. Like, as of you know this year, I've been doing all of their updates with AI. Just even if I don't have like AI building in the background, it's like I I can just talk to it and figure out what I need to do because Webflow's chaos, and it's like oh just follow these steps, and then and now I can quickly get these ideas out so we can you know more rapidly test ideas and see what's working. But I'm probably not going to go build all of these other applications and tools for my clients just because I could. So I'm going to stay focused on the positioning aspect, where I know what we need to bring to the world. Yeah, I guess. Yeah, what's your? How do you decide what you should be focused on versus just picking some stuff up because you can?
Speaker 1 06:58
I think I Steve Jobs actually mentioned this. I forgot which year at a keynote that he used to start at the technology first, and then see what the market would be for that technology. So starting with product, oh, we have this cool feature set, and let's go sell that and force it upon. Versus starting with customers, what do you need to make your life easier? And if you can sell that, then you work backwards to the technology, and I'm I'm looking at the same thing. What do people need right now? And I first solve my own problem, like just like the back context around it. I have a ton of intellectual property. I've written I don't know how many hundreds, if not five over 1000 pieces of blog posts. The podcast is like only only on 300 episodes, but it's still everything's indexed. I think about my own thinking, and I've practiced this over time. So I have and I have a process for delivering client results, and I and I like to think about how do we innovate, how do we change, what we can do to do. But I'm like, okay, I'm just getting tired of having to remember all these things. So I first started systemizing that in two different agents and skill sets, and then packaging that together. I'm like, what if I can actually create this thing that now helps me as a thought partner, but also in execution? I can give some examples if you like of that. But that first was useful to me, and then I started sharing. Hey, here are the ingredients to do something for yourself. Here's what you need to do. And people were like, that's super interesting. I want that. Like, oh, so you start with the market. What do they need? What are they looking at? What's blocking them from building this? Because I can tell you, like this is not. I've had to flex my engineering mind quite a bit, like a lot more than I thought, and it was it's it's a beautiful thing. But I had to flex it a lot harder at knowing what I actually want as an outcome, and then from there working towards here's something that I can actually serve more people, and it's still in relation to what I do. Like it's not something that's left field, and it's like it like a left field like hole over there. It's random. It's still tied to the entire process or the the client experience because a lot of the work that I do starts from the baseline of operations. Like people don't understand, or at least clients don't initially understand that growth problem usually stems from operations, and that supports their value proposition. Then it's the business model. Then from there, it's their pricing model. Then we can hit growth. Then it's go to markets, and it's like sales and process and all the stuff that comes with running a team and managing humans, and then managing yourself, which is the fun part. So, but it all stems from the operations piece. So that's how I think about building the right kind of solutions, if not internally, but also you can take it to market.
Chris DuBois 09:26
Yeah, first I'm bought into that that idea. I think most positioning go to is serve with like, well, let's what do we need to say to the world? But I like inverting that. It's like, well, what are you already doing that the market wants so that we can use that, right? But I am curious. What's your what is your background? Like you mentioned, like using your engineering brain. I think a lot of people struggle with AI. Period. And like even just knowing where to start. How did you like? How did you get into this and attack it?
Speaker 1 09:55
I started as a creative first. I was actually going to be an art major and then a psychology major. Then I. That entrepreneurship was a thing, so I started doing entrepreneurship. Went to business school. Business school was easy; like it's not that hard. But I always had the engineering framework. Like my mom was an engineer, so I always have that behind, and that was my secret weapon: like pairing growth with operations. And every single agency that I've touched, any team or internal framework was all starting with operationally first. How do we take this growth strategy and operationalize it into reality? So I am an operator, and like I just think that way. But I also know what good it looks like. Like I know what I need to see. I've worked with CTOs with cybersecurity. I have a good friend at CTO. Like we chat every day, just talking about what's new, what's building, what's he doing, how he's thinking about problems. So that's my background around it, but really, you need to, for listeners, when you look at using AI, right now we're we're kind of being sold that the model is the one we have to go after because these companies need to IPO. At least right now, these companies need to pump and need to pump hard. All right, so there's fear mongering. They need to go and sell their product and say my product is better than my 6.7 iteration or my 7.8 iteration
Chris DuBois 11:07
or
Speaker 1 11:07
whatever the number is when this podcast publishes. Like they need to make you believe that you're missing out by having the best thing ever, when in reality that's just part of the equation. The real thing that matters is the harness that you build around the model, and you can own the harness. This is the most beautiful thing. You can actually own the agents around it, the context, the history, the frameworks, the actual skill sets that you load, the tools that you can, the tools you may not own, but you can connect your tools and your workflows to make the model better. Because right now, every model that you use, I don't care if you use the latest, and I've used and I pay for the latest. They're just a brilliant stranger, and right now you are renting. You're giving them your data first of all, and we'll talk about data privacy. And that's a this is anecdotal conversation, but you're giving them your data. Then you're paying for the right for them to process your data, and you're paying for them to give you your data back. Think about that. So, really, what do we own? We own the data. You should own your data. One of my clients was using Manis, and then when the whole transaction with Meta got blocked, they had to delete all that. Like Mantis had to delete all the users and have all users export their data and then re-import it back. The export is a JSON file. That JSON file has an encryption key. Who owns the encryption key? Not you, Manis. Right for your data that you, my client in particular, paid 1000s of dollars a month to use. Okay, so you own your data. Second, you have to own the harness, the orchestration layer, the skill sets, the context, the history, and the memory. You can build it. You can own it, you should own it. It is an asset, and it will make you free-not just free, but it'll give you power and control. Because there's a difference between being controlled, like what just happened, like with with a client situation with Manis, or you having options. And the third is the LLM. And here, there's you can go open, you can go close source, you can rent it, you can download it, but here you got options, and that's the beauty of it. And what that's actually what they don't want you to know that you know you you have options, you are in control.
Chris DuBois 13:12
Man, music to my ears. The every time someone tries pitching me on, like, oh, have you tried this model? Have you done this? Like I, I built what I call my chief of stuff, and instead I play on chief of staff because it doesn't manage people, but it's got all my other things going on, and just it's built with Clod. But I've got my a knowledge graph. I've got a vector store that has like everything. It's got access to all my tools, all my stuff, all my skills, certain agents to do certain work to push push things right into my Slack. But like, I've just been I keep adding to my own system like over time when I need something. I don't know that I could just rip this out right now and like hand it to someone, but it it wouldn't matter if I did because
Speaker 1 13:55
it's built for you.
Chris DuBois 13:57
Yeah. Um. So like, let's let's talk harness, and what are how do you approach this to help someone get exactly what they need built into their own system, so they have control over it and can build what they want.
Speaker 1 14:13
So, I built this specifically for for me service businesses where the owner operator is still like in charge of important decisions and is able to like own revenue and go to market. So I just build it for myself, and I saw that other people had the issue. So then I productized it and I preloaded it with a few specific agents, namely a security one, because like you just think because you're open, like you're grabbing open source code, it's legit. Most of them are malware. So that's one. And the other thing too is that you can instill your own value system into the rules that the system abides by, and I personally believe that humans are great. That is controversial nowadays, Chris. I believe humans are great, and I think that we should be in control. And I believe that the agents or whatever you're using as the orchestrator should shuffle ideas and information. And decisions to you so that you can decide which to point, like which direction to point your your power. And I call like the the nine frameworks of human flourishing, because I believe that this technology should enable humans to flourish and help other humans flourish in the market economy. So there's the rule system. There is obviously the the simple primitives that I did invent. It's like the agent system, the context, the memory, the state. So all those things that you can have, but I also want to focus on where are you hosting the data? Do you want to keep it local on your computer, get some backups to it, or you can host it on the cloud? And then you also have to think about what what data do you feel comfortable providing or giving to the models. Now, I was at a dinner at Napa at a conference, and in front of me was a researcher at one of the leading labs, great dude. And then to the right of me is a lawyer. And we were chatting about like that little toggle that says you can't train on my data. There is so much sophisticated legal tech into that that it actually makes it nearly impossible for them to train on your data. Take it for a grain of salt. That is an anecdotal conversation, but it is important for you to know what data are you comfortable providing the closed source models, and we've seen this like one of the guys here in San Diego, Vistage chair, and he has a group of people. One of the dudes is is a lawyer, and he says his biggest competition is Claude for a legal firm, and we have seen Claude create applications that could compete with some of the users. So just just keep that in the back of your mind. Like, what data are you comfortable providing? Which data should you never provide? Which data should be stored locally, and you can run a local model on? And being able to decide what you're going to run through, what the connections are, and what is the end outcome? What's the goal? What does good done finished to look like? And are you a good judge of what good done finished is? Because just because you can take the model or take this thing and say, "Hey, make it better, and you trust because you're using the best model that it's going to make it better, it's better to its standard, not to your standard.
Chris DuBois 17:00
I will say, even so, I was running an agency before coaching, and one of the things that we made sure every single process had was a what done looks like and what good looks like, and it had to be built into that process. Otherwise, it was not a process we were using. Like we needed a great example, we needed to be able to point to those examples and show why, so that a new team member could come in and understand all of these things. Now, this was before AI, but when I started more
Speaker 1 17:28
powerful,
Chris DuBois 17:29
right? But when I started talking to other agencies, I realized like nobody's doing this. Like even even with just their people to make sure they understood what good looks like, it's like they just weren't doing it. And so I think there. It's like now that we have AI, it's like people skip that step. There's a
Speaker 1 17:46
cluster. It's a cluster, and this actually hinders. If you're doing that right now, I get it. Do your operations, and this is not like two dudes saying, "Hey, you got to get your SOPs in line. But yeah, this is two dudes saying, "You got to get your SOPs in line if you want to make more profit and create new value. Because one of the key so I think there's four new ways to to price and create value with AI specifically for service companies. The first one is margin. Just increase your pricing or keep your pricing the same, reduce your cost of of delivery, and then just enjoy those fat margins. And I'm seeing that more than I thought, even for non like service companies, I guess one of the guys he sells infrastructure and with a real estate play, but he's doing a margin play, but he's radically increasing prices because it's it's in a hot market. But I'm seeing that surprisingly more than I thought. The second way to price is the volume play based on consumption. So services start to act and behave somewhat like SaaS. Like the more that you use this, the more that you do. For example, I used to work with one of the one of the clients, great clients, one of the top real estate lead gen agencies in the country. Their biggest competitor was pay per lead, like pay per lead, pay per outcome. So that model isn't new, but now it is in terms of how you can are able to. I'm not sure if the word SaaSify, but be able to create your services as a consumable, as an outcome-based thing, not just like here's here's a brand deck or here's a you know a thumbnail. Like that's not really what this is about. It's more about how can you go closer to done, finished outcome, or go deeper and wider with the client. The third one in this one, if if you don't have your like we were talking about two duos talking about getting your SOPs in line. If you don't have that, you can't leverage the third way, which is the IP way. The IP play is where you're actually leveraging your methodology, and it becomes your operating system. And that is beautiful because that gets you to augmentation, and we'll talk about that in just a second. And then the last way is value capture, where you're able to take co-risk with clients and go further with clients based on a shared goal and a shared outcome, and because you're using the agents and the skill sets, you have a little bit more control than you would in the past to influence that outcome. And I can share some examples there. But to tie this together, what does AI actually allow you to do, or what is super intelligence, or whatever we call this? Technology. Just call it technology, because in the past, like if this was 20 years ago, we'd be talking about the model of your computer. Just imagine. Let's replace AI by saying, "What does your computer allow you to do?
Chris DuBois 20:11
Is
Speaker 1 20:11
Chris going to do better than this other person because he's using a MacBook Studio, MacBook Studio, or like, or than someone else using a MacBook Air? Like we're thinking that way. It's archaic, okay? It's technology
Chris DuBois 20:24
because
Speaker 1 20:25
you know the first agency I started, I was in my MacBook Air 11 inch. If you remember that sucker, like a little 11 inch MacBook Air, and we were doing it. That got me out of finance and banking. Started the first agency. We started growing. I think we had 19 clients on a little MacBook Air, 11 inch. Right. So is is it the technology that changes it, or how you use the technology? It's how you use it. But anyways, this technology takes you from automation, which has been around longer than you and I have been live, like automation itself. Now we can have agentic automation, sure, but it really moves us into augmentation to be able to do more with less, or go further with the current resources that we have, but most people stop there. I'm pushing and I'm preaching to go to net new value, just because you have augmentation. Like I'll give you an example. When I do strategy with clients, it would take 20, 3040, hours, even with AI in the past, to do like a proper growth strategy for a client, like properly done, and with AI, like been 2223, yeah, I would do research, get some insight, maybe get some ideas, and I would still take 20, 3040, hours to do the strategy, but I could, you know, probably pump out 60 to eight hours of work. Cool, same time input, more value or more you know quantity mass outputs, and still deliver a good quality product and deliverable to clients. Now with augmentation and the way that I've orchestrated, and I'm using the IP play for my personal services, like the what the services that I deliver for clients, I'm able to take those 20, 3040, hours and condense it to five deep hours of thought. Not it's doing the work for me, but I'm directing the work based on what I know, and like I'm terrible at teaching right now because there's some things that I just know that I it's like the known what is it the conscious like conscious competence unconscious competence where I just I just don't know why I know but I know that this is what we need to do this how we need to take the direction here's the price point here's the offer here's how we're gonna go to market with it, and I direct AI to do that, and it's able to take all my 1220, 3040, workflows and execute for me, and deliver something that I know has quality checks because I built in different like systems to have quality checks, and that's the IP play, and that is augmentation, but that's very powerful, and if you can't document how you do things, you can't get to this play. It's a very powerful place to be in.
Chris DuBois 22:43
Yeah, someone years ago, I'd heard him talk, and they kind of reframed AI from artificial intelligence to augmented intelligence, and that did it for me. It was just like, oh, I get it now, like how we should be using it. So it's interesting to hear you talking about it again. It like reminded me of like, oh yeah, that's how we should be using it. I want to hit two things. So on the on the SOP thing, just quick to go back to it. I think a lot of people are also just hoping that like if they have an SOP, it's going to be what what does it for them? For all of the listeners, I've probably harped on this before, but like, if you haven't gotten a great result from that SOP, then like, don't use that SOP. Like, go get a great result, then create the SOP around the result, then look at automating and and finding ways to incorporate the system. I see a lot of agencies jumping right to oh let's just have AI do this with us. But if if your SOP is built to get a mediocre result, you're probably just going to have more mediocre results. And so, like, how can you improve this and actually do something better? But I want your take on because you you started getting into this, but like I guess the phrasing would be like services as a software, so we're just inverting SaaS.
Speaker 1 24:05
Yeah, somewhat, somewhat. Yeah,
Chris DuBois 24:07
yeah. How do you? So, like, I guess, do you have any examples of like companies you've worked with where that's kind of been the play of like, how do we optimize this so that your services become so like streamlined, I guess that it becomes easier to manage those margins and improve those.
Speaker 1 24:28
I mean, I think so. If we're looking at the volume play, yeah, I think the the realistic lead gen agency would be a great example because they ran everything strict and they were really focusing on dialing in processes and having as much automation as possible. Where it is a paid subscription to work with them, like it's not a cheap subscription, but it still behaves almost like a SaaS. So I think that's that's important, and I think that that is valuable. But I actually don't think the play is to think about your agency as a software, even though I know who was it, Sequoia that did that paper in March that the what is it the. The inverse of SaaS, like services or software as service or services as software, is a 6x multiple or something like that. Like because you're able to charge more than software, I think the key thing has to focus on end outcomes. And I think this answers this is a good closed loop to your your initial question of, you know, if if a client's asking for a discount because you're doing things quickly, like shouldn't you just give the discount? And the answer really is, well, depends. What are you actually delivering? Where is your value actually in the process with the client? And that's that's different per industry. That is different per your competency, and that's different depending on the actual service, like the the vertical, the service that you're providing, not just like what does done good look like, but even for net new value, how do you how do you help the client go further? How do you help the client go deeper in what they're doing? How do you help the client go wider with what you're doing? Like I'll give you a perfect example for what I've what I've personally been doing. A lot of the clients when we're working together, sales is obviously a big thing. Pipeline is a big thing, and I would refer it out. I would talk to this legion agency, or talk to these, or here use this software, do this to the point where like I need to figure this out. And with net new value, and I've actually done this by hand, and it took like labor and toil to figure this out by hand. Obviously, on the computer, figuring it out and leveraging technologies with waterfall systems to not just find oh quote unquote leads. Anyone can find a lead. You throw a rock, you find a lead. Okay, you throw an ad for three bucks, you get a lead. It's not hard to get a lead, but people want. I want leads that buy. So the key thing is, well, how do I actually get someone who has the right intent that my offer is relevant to them, that I'm actually able to add value to them, and that my outreach is actually is valuable in itself, not just like a distraction, not just like something like, "Hey, I just saw that you posted on LinkedIn, buy my stuff. So reverse engineered that, and I work with a global manufacturing company on it actually, and they're not selling cheap stuff; like they're selling five to $20 million products or services, really documented that process, and now that is an anchor to one of the services that I provide, because it's actually delivered real world results, and I can help clients go further. But I'm not a lead agency. I don't focus on leads. It's just something you need naturally. It's like air. You need air to breathe. So here's the thing that I can help you go further. So the excuse to to not do this, at least for my clients, like it's not an excuse of opportunity. It's an excuse of capacity. So let's dive into your capacity. So that's just an example of finding and going deeper with clients for value versus looking at here's the thing I deliver. I can do it faster and cheaper for you. That's not where this is at. The key thing is how do you help the client solve their problem? You can help them solve faster. You can go into greater depth with that, or you can help them in more ways. Once you solve that problem, what other problems arise?
Chris DuBois 27:45
Yeah, like that. I think there's also that layer of, and I guess I'm curious on your thoughts on this, like kind of showing your work, because when the when the client does think you're just using AI for everything, even if it was no, I actually did this manually and say it up overnight and just like crank this out because of the speed. Now everybody's questioning. Well, did you did I actually get something from you? Right? Like, did did you have to provide some strategic thought in here? And when you can show your work, like there's this perceived value. I guess that comes with effort.
Speaker 1 28:17
Yep.
Chris DuBois 28:17
And so, how do yeah? What are your thoughts on that?
Speaker 1 28:22
I think that if you have to show your work and the client is questioning that this actually come from you, you're still answering. You're doing the wrong work. Is this good? Like my clients know I use AI. They don't care. They actually love that I use AI. The key thing is because here's the thing: everyone has access to a computer. Everyone has access to AI. It's dirt cheap. That's the other paradox too. It's dirt cheap, and then they're wanting to charge you more for it. Anyways, it's dirt cheap. So if they say, "Well, I could have just done this, then you're doing the wrong work. There has to be something unique of what you do. So I wrote about this, I think, in either March or April or or February, but the services stack. Like I think services have to move up the stack, and the first layer, the bottom layer, is execution. Like this is already replaced with AI. A lot of executions be done with AI, and that if your agency is only an execution, then yeah, you're smoked, and that's why you're getting these questions. You're working on the wrong problem. Then the next layer on top is like template strategy, and I'm not going to name names, but like you know, there's people that sell their license or their methodology, then you just use it for them, and you're a certified guide for that methodology. AI is going to do that, and I think AI has taken that over and it's butchered the that offering. But then the the middle layer is okay, judgment. AI can take away my judgment, but when you turn on thinking mode and then it thinks for like 20 minutes on something, did it replace my judgment? So judgment still is something that we can own, but it's more nuanced in terms of guidance versus just like obeying the the agents or the AI to guide you. The fourth layer is transformation, and this is twofold because you have you have to focus on keeping the client accountable. And keeping yourself accountable to an actual transformation, so you have to sell the end state, but also what does that actually look like, and how are you going to hold accountability? But the fifth layer, and this is again not to not to reiterate what we mentioned earlier, but people always talk about you've got to be like the strategic partner for the for the clients. You got to be that person that's right next. Like like sure, but how do you get there? You get there by creating a shared belief about a future that doesn't exist yet, and both of you seeing that future and co-creating it together, and that's the difference between humans and technology. Humans exist in atoms, and we also have non like non atomic particle, like something else that helps us exist in that potentiality state, which isn't real yet. AI and all this-it's all bits. Yeah, we'll have robots, we'll be atoms, but it can't share, hold a future belief, and actually know it. It can conjure a belief, but it can't hold it with you and make me feel an emotion where I work with it to build that future belief on a shared reality. I know some people are dating the RAIs. I know some people are looking at like, oh, but I believe my guy, my little Chat GPT is telling me I'm the best. But yeah, whatever. But you know what I mean. So it there is that layer where you have to hold and distill that and keep the human relationship. And if you do that and you focus, oh, what's the work that we're providing for you? What are the actual problems that we're facing? What else can we solve for you? You do those things. They'll never ask that you use AI for this. They'll say, "How can we get more? How can you do more for me?
Chris DuBois 31:31
Yep, agreed. Okay, let's do 8020. Where where should people start with, or maybe not even start, but just focus their attention. So I'm guessing most people have already started somewhere. They're playing. They're
Speaker 1 31:47
dabbling. Yeah,
Chris DuBois 31:48
yeah. Where? So where should agencies focus their attention right now to make sure that they're setting themselves up for the best future?
Speaker 1 31:57
So I wrote this on Substack. There's there's seven pieces to your business model. You have to start with first principles. You have to start with first principles. I'll say it again. You have to start with first. I'm actually talking to the guy, the one of the groups that that bought EOS, and like we're chatting similar conversation. Like, well, EOS is a first principle, and AI is not going to like change first principles. Like gravity. I hope this is video, so you can see that. But I just dropped my mug. Gravity is is a principle. It's going to exist regardless if I want it to or not. You need to start with first principles in your agency. That begins with your value proposition. So there's seven parts to your business model. Go over them briefly. Your value proposition, then your key customer segments, your ICPS, your key channels, and how you're reaching them. Your key partners and how you're reaching them, your profit model, what's your actual pricing model or packages or offers, then on the flip, what are the key partners you're working with, what are the key cost structures that you have, what are the key resources that you need? That is all going to change. That is all changing right now. If you focus first, what is the value that I'm providing? How does this change the way that I operate, the way that I sell, and the way that I deliver on that value is changing. But you have to start first there. I have a pre-stack and podcast about that specifically. But start first with first principles. What is the thing that my clients need? Where are they going? Where is their industry going? What are the problems that are going to arise? How can I meet those problems ahead of the curve and start talking about that now? Versus, and I see this issue a lot with thought leaders too, and something I love. They're just anchored in their current belief, anchored and not changing their opinion because that's what I mean. That's what you're supposed to do sometimes if you're a thought leader, but not changing with how value is going to be created, because value in itself is always-it's a first principle. There's just new ways to create that value, and peripheral ways that you can extend that value with clients. That's what's changing, and that will change pricing structure, cost structure, key partners, key channels, ICP. You might expand the TAM when you do that, and that's where you need to start. Once you realize that, then work backwards. It's like, how do I capture that, or what can I build now to capture some of that right now? Or if you're just starting, well, I'm on the tech side first. What do I do now? Like, well, first of all, did you automate everything? Great, automate. Well, you have to not everything. Where are you with augmentation? Have you found net new value? Have you had your oh crap moment? You need to have that moment with AI. I know there's a lot of people in some of the conferences that I've been at. These are world leaders, legit world leaders, and they theorize on the techno. They haven't used the technology. They theorize it, and I'm encouraging them use it. You got to have that moment. Got to have that moment of what is possible. And you're like, that's an aha moment. You can only have that. Like, how can I teach you, Chris, what it feels like to drive a car? I can't. You have to go drive a car and go to the back back back road. Do 67 or 7080, miles an hour in a nice, beautiful back road right away from the beach may may not have done that, and it's exhilarating. It's fun. I mean, if you're like me, I guess I'm crazy. But how do I? You you have to understand that what possibility, and then from there work backwards. And this is where the net new value lives.
Chris DuBois 35:19
Awesome. Well, with that, let's let's land this thing. Two two more questions for you. First one: What book do you recommend every agency owner should read?
Speaker 1 35:33
That's a hard universal question. I'll just tell you what's helped me. There's a few over here besides my own, but the one that I think stands out is Zig's book called "See You at the Top. Changing the way that you think, changing the way that you perceive yourself, perceive your circumstance, perceive your opportunities will change how you operate, will change how you execute, and it will change the outcomes that you produce in life. I can tell you, like we have a whole story on that, but that bookcast transformed me in more ways than not, and it's not what you get from the book; it's the ideas you get when you read or listen to the book.
Chris DuBois 36:08
Awesome! And then the last question: Where can people find you?
Speaker 1 36:14
Anywhere that you have a computer and internet's at do goodwork.io. That's do goodwork.io.
Chris DuBois 36:21
Awesome, and we'll get your substack, everything, everything else, your podcast linked up in the show notes, so that everybody can can start learning a little more about AI and how they should actually be running it. Let's do it,
Speaker 1 36:33
and I have a lot of open source material too.
Chris DuBois 36:35
Awesome, thanks, man. Thanks for joining. Appreciate it,
Speaker 1 36:39
Chris. I appreciate you.
Chris DuBois 36:44
That's the show, everyone. You can leave a rating and review, or you can do something that benefits you. Click the link in the show notes to subscribe to Agency Forward on Substack. You'll get weekly content, resources, and links from around the internet to help you drive your agency Forward.
Transcribed by https://otter.ai