Micah Johnson 0:00

A hundred episodes ago, we asked, What if you could save one hour a day? And that felt ambitious. We had no idea we were about to watch AI compress entire job functions into a single afternoon.

Alane Boyd 0:14

Welcome to Automate Your Agency. Every week, we bring you expert insights, practical tips, and success stories that will help you streamline your business operations and boost your growth. Let's get started on your journey to more efficient and scalable operations. I can't believe we're celebrating our hundredth episode, Micah.

Micah Johnson 0:36

A hundred episodes, Elaine.

Alane Boyd 0:38

It it went by so fast. And it's been such a fun journey from that first episode and us just winging it and trying something. And really, that is the best way to start.

Micah Johnson 0:50

I mean, I remember recording the first episode, and it's so funny because we were we were talking about how do you save an hour a day, thinking, all right, this is gonna be awesome. We can save people an hour a day. We started our podcasting journey. And if I remember correctly, even in that first episode, it's been a 99 episodes since I listened to it. We came up with so many ideas from that one episode. I think we recorded like five episodes in that day. Just I think so too. Went crazy.

Alane Boyd 1:23

We were we were crazy back then. We've learned so much.

Micah Johnson 1:26

We're old now.

Alane Boyd 1:28

We're older and wiser. But it's it is crazy to think because when we launched, we were really thinking so much about traditional automation and project management systems because those could save an hour a day. But now we're talking about saving more than six hours a day. It's so much time saved that now a lot of times you and I are talking about time saved on a weekly basis.

Micah Johnson 1:53

Geez, if we think about the last episodes, the last few episodes that we recorded, it's it's not even an hour a day. It's how do you change your whole dynamic of operating a business?

Alane Boyd 2:05

That's one of the conversations that I've been having a lot lately too. And I can tell you from the companies that really are wanting to not just get ahead, but like really utilize it. They aren't saying, how do I make AI work in the way that we operate? They're changing the way that they operate because when you figure out something that works manually, that doesn't mean that that has to be the same way it works with AI agents or automation or AI.

Micah Johnson 2:36

Yeah, I'm having those same conversations all the time. I think the things that stand out are really can you change and adapt to change? Can you iterate quickly? And can you leverage that superpower and apply it to how do I change my operations to use AI? And it's it's like you said, right? Like if you take how you've been doing business forever and just add on AI, like that was last year. That was that was chat, right? And you could just add chat right now that it's doing different things. You can't just add on stuff like cowork. You have to think about how work changes. But I thought it'd be kind of fun in this episode, Elaine, if we just revisited the timeline of these last 100 episodes.

Alane Boyd 3:26

All right, let's do it.

Micah Johnson 3:28

So I went back through with the help of our friend Claude Cowork, of course, um, all 100 episodes. And we essentially pulled a timeline of what happened over the course of AI, paralleled that to the course of the 100 episodes that we've done over what is that? We did, yeah. So two years, two years we've been doing this. We've got those two listeners in two years, Elaine. It's been absolutely wonderful.

Alane Boyd 3:58

We're big fans though, those two listeners.

Micah Johnson 4:01

That's right. But where we started, I mean, all everything we were talking about was the fundamentals of operations, right? So it's like SOPs, delegation, and project management tools, Asana ClickUp and Monday. And it's funny thinking back on that because we still could be talking about all of that.

Alane Boyd 4:24

Oh, yeah. I mean, that's the core of using like cowork and AI agents, is you still need a system, you still need to know the system, and you still need to have an SOP, whether that's a quad skill or a legitimate SOP that people follow. Like those things haven't changed, which I think is fascinating.

Micah Johnson 4:43

Yeah, I agree. And and I can't remember if we talked about this off the podcast or on the podcast yet. We've had so many conversations lately, but just this whole concept that maybe even the fundamentals that worked before computers were even popular are actually some of the stuff that might swing back into effect now because AI is so much like working with humans. It can make mistakes, it can make the wrong assumptions, it can reason incorrectly. Well, all the fundamental systems used to be around human error and, you know, the exact same issues. And then we got onto very logical constructs with computers, whereas like, if then it's gonna be a robot and it's always gonna act the same way, totally. But now we got to go back because we've got AI, and while it's a great tool, anyway, fundamentals is where we started. It actually worked out really well for the AI timeline because it was just right at the beginning.

Alane Boyd 5:46

I noticed this uh kind of coming up lately when I've seen people at events that have known me, you know, since we started this business, and they're like, You have a new business now. And I'm like, no, no, no, I have the same business because these core things have not changed. And why we started this business was the biggest goal to get you to overcoming the obstacles of growing and scaling your business. That is still the exact same thing that we do. And when we started it, there was a focus on traditional automation and project management systems. Those things still apply to AI and AI agents. We just evolved our offering and how we solved that problem, but our business is the same.

Micah Johnson 6:29

Totally, totally. I I would even argue that having a project management system is even more important than before because when you use tools like cowork, you need a place for it to actually create tasks and you need the systems to have the accountability. You can't just delegate stuff to AI and then move on with your day and nothing happens from there because literally nothing will happen.

Alane Boyd 6:53

Nothing will happen. Micah, are you finding this when you're talking to people? I this year, especially since cohort came out and people are seeing the demos that we do. I notice that when I talk about consistency now and talking about why having templates for the work that needs to get done, the tasks, like new client onboarding, for example. What are all the things that need to happen for new client onboarding or a new service that you sold a current client? I, for the first time this year in seven years that we've had this company, people are like, I heard you. I I hear the need for consistency and how important that is. And I am working with my teams to put that in in each service offering, in each part department. But no one before I said that, they would always say, Yeah, but yeah, but you know, this, this, and this. Now it's like, oh, I'm hearing this thing of how important it is. I want my company to get there. I want us using AI. And it's a lot of times they're not wanting to replace their team. It's because their team is exhausted. They want things to be working easier and they see that opportunity with AI to help them, but they still have to have those fundamentals in place, or AI is gonna still have a complete chaotic end result if you don't have that consistency there.

Micah Johnson 8:17

Yeah, I would say we started seeing this at the end of last year. And for for me personally, like this is so exciting to see because, well, I'm a geek and I get excited about systems and processes and running stuff without having headaches and putting out fires all the time. And it was always, I wouldn't say demoralizing, but it was frustrating that you'd talk with a prospect or you'd talk with a potential client. And it was kind of like, yeah, yeah, yeah. Okay, cool. Like operations, got it, but what about sales and marketing? And you could, you know, you know how this story goes, Elaine. We've lived this, like we've seen so many clients and prospects go through this as well. Like you can take it only so far with sales and marketing before you're scared to do sales and marketing because you're worried that the next sale or the next marketing campaign is gonna topple the whole house of cards. I feel like AI and this, like we started seeing it when everybody was just chatting and it's like, oh, cool, let's do some AI automation stuff. And it got people a little bit more excited. And then, like you said, since January, people have gotten really excited about the whole, all right, how do I retrofit my operations to you? Like people want to talk about operations, they want to streamline their business, they want to be doing this stuff now, where before I was like, I'll worry about that later.

Alane Boyd 9:45

Exactly. Yeah, they're worrying about it now, which is great to see because people are feeling, I think it's more urgent now because if other people are figuring that out because of how fast AI can execute versus a human, they are gonna see that pain move forward where they're not gonna be able to keep up.

Micah Johnson 10:03

So we're getting off track a little bit, but we we started this podcast pretty much at the beginning of main AI going mainstream with chat. We could kind of see where this was going. Obviously, we saw, you know, room and automation and different things like that. By our 40th episode, Elaine, I find this interesting, it's kind of late, early 2025, late 2024. We started talking about AI agents. And we actually did an episode 41 was what are AI agents? And I mean, honestly, we're still answering that question because the term AI agent has been thrown out for anything that is everything related to AI.

Alane Boyd 10:46

Mm-hmm. It's really confusing people because they're seeing the word agent when they're in ChatGPT, they're seeing the word agent when they're in gamma, they're seeing the word agent and clickup. And then we're talking about AI agents across systems. So not living in a particular system, but how does it connect with, you know, four or five software platforms that you're already using?

Micah Johnson 11:10

Yeah. By episode 55, we started covering N8N. And we went on our own journey with N8N. I mean, we took that, we were using, we've obviously used Zapier in the past. We used Make a lot, and we started looking at N8N and going, how, how quickly, or we saw how quickly we could build AI agents, and then we started ripping them out for ourselves. And we're we're always our own guinea pigs first, in case any listener doesn't already realize that. We test everything on ourselves first. How can we achieve this? How can we do this? And then we talk about it on this show, and we obviously implement it ourselves. But yeah, episode 55, we introduced the concept of N8N as an alternative to make and zappier. It's just continued to grow from there. But what here's what's blew my mind when I was doing this, Elaine, episode 60, we're already comparing well, what is Chat GPT versus cloud versus perplexity as an actual business decision that you need to make to start using one of these today? That's not like looking at this timeline and how quickly, and I think that's what I wanted to convey partially in this episode is two years. Two years from literally, oh, AI seems interesting. What is this thing as a side note to why haven't you rebuilt your entire organization around AI yet? Two years, 24 months.

Alane Boyd 12:33

I I mean, like when I think about when I first started talking about AI, which was four years ago, actually, because I I think about um pivotal moments that got me huge speaking engagements, and that's what I anchor back to. So four years ago is when we started using AI, and then to the point of even in the last 12 months, what it's looked like for us to use AI versus the three years previous to that. Change is so fast. The just not even the change, just the the pace at which you can automate things with AI is is incredible compared to what we did those three years prior.

Micah Johnson 13:12

Absolutely. I mean, leading into this year, right? Leading up to our hundredth episode of this podcast, January was when Claude Co-Work was released. By mid-February, you and I had retrofitted our entire business around cowork because again, the guinea pigs that we are, we actually went in and said, Well, how the hell can we use this thing? And the second we tried it and started applying it, well, then it's like, okay, how do we expand this to other members of our team? And man, we have not looked back from that. And we also can't stop talking about it on this podcast because it is groundbreaking. Like I personally think it's at least as big, maybe even bigger, than going from zero computers in an organization to having, you know, computers and spreadsheets and documents and everything. And I I have this conversation a lot, but it's, you know, imagine a company where one person is using a computer and everybody else is in paper. You don't get a lot of gains, and that one person is like, oh, that's the computer guy. Once you switch the entire organization to using the computer and everybody has that standard, which we're all used to now, right? Then you can operate a completely different way. And that I think is exactly what we're living through again, right now. So we talked about AI agents a lot. We we obviously talked about N8N. We talked about agents inside of tools. So, like ClickUps agents and Asana's agents, and Monday had some AI stuff going on. But what rapidly changed, especially as we got into the uh episode 90s, was cowork. And all of a sudden, now we have to talk about things like MCP servers. And you think about all the money that all these companies dedicated towards building their own agents. And meanwhile, all of us on the outside of those platforms are going, well, can I connect Claude to it? How do I connect Claude to your platform? I don't want to use your stuff. I want Claude Cowork to go in and build stuff in your platform. And they rapidly had to shift, but also all of the companies that are using that rapidly have to shift.

Alane Boyd 15:24

Yeah. I mean, I I think about my day and how it really is where I'm operating from co-work, and that's going out to the other platforms that I use instead of going to each one, which doesn't just save time, but uh this is what I find really interesting when I look at conferences because it's like, oh, there's an AI talk. We're gonna check the box. We are covering AI in a session. And teams think of this way too, like, oh, well, we did an AI training, check the box, we're good. It's really not like that. Using AI is a piece now, and when we say our operating system, it's because we're operating from AI in every part of what we do. It's not, oh, I'm doing this over here and I'm gonna use AI as a part of it. It's really the central part of what we do.

Micah Johnson 16:12

Yeah, I think, you know, at least this is where I can see it going, who knows where it will end up with where the providers are taking it, or do we start buying our own hardware to run our own models, which are all possibilities. But you're not gonna run a company these days without email and at least Google or Microsoft, you know, as at your docs provider, your spreadsheet provider, your email provider, right? You're just not gonna do that. And we're already in the stage where you're gonna need that AI provider as part of it to run your company. And then you get into skills and all the stuff that we've been talking about in our recent episodes, even within the last month, the last two months, so much is changing and so much, you know, is just being developed. But it's kind of like we said early on in this it's the companies that have that superpower to be agile to change quickly. If you're sitting around right now and going, well, I'm waiting for this to slow down, you're gonna keep waiting because that is not happening.

Alane Boyd 17:18

It would be a pretty sad day if you think that it's gonna slow down because I want it to keep changing and keep evolving. I I can't imagine not having cowork.

Micah Johnson 17:28

Yeah, I I know I've said this many times recently too. It's like if if we for some reason lost the ability to have cowork, I I think I would have to pick a different job. I started to do that.

Alane Boyd 17:40

I think I'd be a gardener. I would just sit outside and I don't know, grow tomatoes and get a tan. Yeah. So that AI is not gonna change in my day.

Micah Johnson 17:52

But I mean, you you get spoiled in a way because you see what it looks like on the other side of that fence is like, geez, do I really want to go back and do all this stuff manually? Do I really want to analyze all these data sets manually or pay somebody to do it for hours and have a worse job than is happening? No. The you know, labor that we have to do as quote unquote work, especially in agencies and service and office kind of stuff, like that's just now really shifting away from the time-consuming monotonous stuff and into the judgment calls, the knowledge, all of that. That's where the value is and that's where the work is. There are, you know, Elaine, we were talking about this, well, a lot lately. Um, but there's stuff that hasn't changed either, even through all of this.

Alane Boyd 18:43

Oh, yeah. I mean, good data is an important one because whether humans do it or AI, if you don't have good data, it's gonna just pull crap.

Micah Johnson 18:54

You know, it's funny, and I know we did a podcast on this recently as well. Good data was always important, but you could get by because you could throw enough people at it. If you have AI starting to eliminate that and you have to keep up with that baseline that keeps raising because people are using AI to do stuff faster, your best ability is to do these fundamental things better. So if your data sucks, you're just gonna be spending more time correcting the AI, or you just have to correct your data you once or build systems to correct your data as you go to give that clean context to get a better output.

Alane Boyd 20:11

I I think another one is you you have to know your workflow. You have to be able to articulate it. It's so often that's like somebody will be like, Yeah, I know it's gotta go here, here, and here. But then it's the nuance. You and a lot of people don't know that nuance because they've been doing it for so long, or it's in each team member's head and they haven't gotten all five people in there to agree on what that is. So a consistent workflow that you can articulate.

Micah Johnson 20:40

Yeah, I see that a lot where people do their time-consuming, monotonous parts of their job, and it's almost like their subconscious kicks in and they don't realize how many judgment calls and decisions they're making to get through that part of it to get to the you know, asset creation. And so if you're trying to leverage AI to create assets, you have to build the system, like you're saying, Elaine, you have to be able to articulate all those different pieces to get to, okay, if this, then here's how we'd handle it. If that, here's how we'd handle it. Or if you can't figure it out, there's all these ways to design it. But if you can't articulate any of that, then you can't build an SOP. And if you can't build an SOP, you can't build a skill. And if you can't build a skill, then you have no repeatable process for AI and you don't get any of the benefits.

Alane Boyd 21:39

Yeah, I mean, exactly. And one of the other things we've been talking about this, Micah, for 15 years, and it has not changed, is working with software that has an API. Because even if you don't have an MCP, which is the new modern way that AI talks to software, you can still use an API. And I mean, even earlier today, I was saying, well, cowork doesn't have a native connector for that, but you could get it this far and then call an AN workflow that could do the rest of it and pull it into your software. That is still, you still want to work with software that has an API because that is how software communicates with each other.

Micah Johnson 22:18

Yeah. I mean, I don't know about you, but I find myself if I look at a piece of software that has to relate to a task I'm doing or helping out a client and looking at what they're trying to achieve, and we hit a roadblock with a piece of software where it's like, we're gonna have to manually go in and do this. It just Just feels like the wind is taken out of my sails because wouldn't it be easy if we could just connect Claude up to it and have it do this piece for us or have it send something to N8N. But we can't. So now we still have to do it the old school way. So yeah, fundamentally, APIs, webhooks, MCPs, gotta have it.

Alane Boyd 22:59

I I mean that goes back to where we started this conversation too. And that you might have to change operationally. What the software that you're using, if it is outdated, if it doesn't have an API, I mean, I would advocate for changing because you can say, even though it takes work to identify a new software, to train on a new software, I I'd put my money on that that time is going to be given back to you tenfold, a hundredfold, on having a software that could be part of an automated process rather than sticking with something outdated that isn't going to work with other technology.

Micah Johnson 23:38

Yeah. You know, one thing that I think has changed that isn't, you know, necessarily technology, but because of the technology is when AI first came out, we talked about this on our podcast in some of the early episodes, too, Elaine. You still see headlines like this, but AI is going to take over 100% of jobs. Nobody's going to have any work. They're going to be our AI overlords. And, you know, AI is powerful, but it is a tool. The more you use it as well, the more you integrate it, the more you realize there is no way in hell you should just let this thing run anything without human intervention. And we've kind of ridden through this journey and evolved with this journey of people coming to us saying, Hey, how do I automate my entire business? Guess what? You can't. How do I feed everything to AI and let it make all my decisions? Guess what? It can't. And then there was kind of this perceived notion, again, makes great headlines, but automation should equal a hundred percent of a workflow. Right. And I think the thing that has changed in that perception, and I think cowork helped a lot with this, is that it doesn't have to be a hundred percent, nor should it be a hundred percent. Can it be 50%? Can it be 80%? Can it be 90%? Can it be whatever percentage is getting to the important judgment calls?

Alane Boyd 25:06

If you're that you want great team members that have good judgment.

Micah Johnson 25:12

Ah, yes.

Alane Boyd 25:14

Bringing it back to what hasn't changed, Mike.

Micah Johnson 25:16

We did a whole episode on good judgment.

Alane Boyd 25:20

But that is that you've always wanted great people with great judgment on your team, but that is what really hasn't changed because AI can take care of the task doers, is the people that have judgment that can take a strategy from point A to point B and empower that idea that that you want to keep.

Micah Johnson 25:42

Yeah, I I couldn't agree with that more. And I think that's kind of where I was going with it is from a from a leader and a decision maker perspective, the goal really should be what percentage can I productively give to AI so that my good hires with great judgment and awesome strategy and just really good acumen on knowing my audience and knowing the business and knowing all this, can do even more and not get worn out with all the tedious, you know, mediocre work that needed to get done before, but now doesn't. Mm-hmm. All right. So we could try to forecast the lane, but I don't even know a hundred episodes from now, when we're talking about our 200th episode episode, I can't even say it. I'm so flabbergasted that we could even talk about it. Where are we gonna be at if we went from AI as this side combo to, you know, it is completely changing the definition of work. What happens in 100 more episodes?

Alane Boyd 26:49

My only like true prediction is, and I've and I've said this and cowork validated it for me, is that automating work between systems is gonna is going to become even easier. So, you know, already we can use cowork to prototype things in N8N. So that facilitation between software and having agents do things, I think is gonna become seamless for those non-tech people.

Micah Johnson 27:14

Yeah. I I would say one prediction that I would make is along that, I think we're gonna see a lot of consolidation of tools. And I think a lot of it is gonna be into providers like Anthropic. And we're gonna see, you know, we might not even need NATN at some point in the future because it's not a far stretch to think that all the connectors and co-work, we have already a team or an enterprise account. We already have users, that's already an organization. What we're really just missing is some of that um, you know, connective tissue on the automated workflows that doesn't go to AI. And I think a lot of internal tools are going to be created in the next 100 episodes of this podcast. I think we're gonna see a trend in that as well. It's very capable to go from an idea to an internal tool prototype to then take that one step further. I think we're gonna see a lot of opportunities for jobs on that as well. Who's maintaining all these internal tools? Who's maintaining these workflows? Who's maintaining the skills? Who's standardizing all of this? Who's the system designers behind all of this? That's a different way to think. And it's it's pushing all of us. I think we have lots of conversations around that. Those would be my, I think over the next hundred episodes, we're just gonna see a lot more of that and probably a whole bunch of other stuff we can't guess.

Alane Boyd 28:40

Oh, yeah. I mean, I I think there's a lot of um opportunity with video too, just editing video, though, those kind of things where there's some tools out there and it it certainly helps, but that is one of the most tedious pieces of what some people have to do is like, okay, clipping this and editing this and and all those things. And it's gonna be amazing to just be able to tell an agent what you want and it produce it.

Micah Johnson 29:07

Yeah. One thing I hope we see less of is the AI videos. I hope we see less of those, the AI avatars, the just direct AI to publish with no human intervention, with no heart, with no soul written into it. I hope we see less of that.

Alane Boyd 29:26

Me too. Well, Micah, I've had a blast doing our hundred episodes. I can't wait till we're a hundred episodes from now talking about our 200th episode was changed because it's been amazing to have this space to share our experiments and our ideas on running and scaling a business.

Micah Johnson 29:46

Yes. And thanks everybody for listening to 100 episodes of this. We couldn't do it without your support. Well, we could. It just would be really boring.

Alane Boyd 29:55

We wouldn't have those two listeners. Yes, that's right. Thanks for listening to this episode of Automate Your Agency. We hope you're inspired to take your business to the next level. Don't forget to subscribe on your favorite podcast platform and leave us a review. Your feedback helps us improve and reach more listeners. If you're looking for more resources, visit our website at biggestgoal.ai for free content and tools for automating your business. Join us next week as we dive into more ways to automate and scale your business. Bye for now.