Find out what AI and Agents can’t fix

In this episode, Alane and Micah tackle one of the biggest misconceptions about AI agents — the idea that they can magically fix broken workflows. They unpack why automation only works when the underlying process is solid, clear, and well-defined.

They share real stories from their own experience, including how strategy and design make up 80% of the work in building automated systems, and why slowing down to document and standardize your processes can actually make you move faster later.

You’ll hear:

• Why AI agents can’t save a poorly structured workflow
• How to identify when you’re ready to automate (and when you’re not)
• Why the design phase is the most valuable part of any automation project
• How to document, diagram, and break large systems into smaller, achievable workflows
• The importance of starting small and improving over time

If you’ve ever been tempted to “throw AI at a problem,” this episode will show you how to rethink automation so it actually works — and how to build a process your agents, and your team, can rely on.

Disclosure: Some of the links above are affiliate links. This means that at no additional cost to you, we may earn a commission if you click through and make a purchase. Thank you for supporting the podcast!

For more information, visit our website at biggestgoal.ai.

Want more valuable content and helpful tips?

Explore our Courses/Programs:

• Complete our self-paced Process Mapping course

Enjoy our Free Tools:

• Free Asana, ClickUp, or Monday.com Selector Tool
• Get 25 Free Custom Automation Ideas for your Business

Connect with Us:

• Join our Community
• Follow us on LinkedIn
• Subscribe to our YouTube channel

Transcript

0:00:00 - (Alane): 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. Micah. We hear a lot of people really excited and they should be about using AI agents, but I think there's a mission perception of what they can do in regards to fixing problems in your workflows.

0:00:34 - (Micah): The people or the agents?

0:00:36 - (Alane): The people misconceive how an agent can fix a workflow.

0:00:43 - (Micah): Gotcha.

0:00:45 - (Alane): So, for instance, if the. If the workflow isn't fully fleshed out or you're having issues with the way that a workflow is executed on within your organization, the agent doesn't solve your problem.

0:00:58 - (Micah): Yeah, it's definitely. So what you're saying is a broken process stays broken no matter what technology you throw at it.

0:01:06 - (Alane): Exactly. And I think some of the pain that people can feel is they get really excited. They want to automate something with an AI agent, and then the brakes get put on because then that's when they realize, oh, I have to put in the work in order for the agent to execute.

0:01:27 - (Micah): Yeah. Yeah. I've got a lot of opinions on this and thoughts on this. I'm not going to say them all right now.

0:01:33 - (Alane): Tell me one.

0:01:35 - (Micah): I'll tell you one. Is that going along the misperception, it. Everything you see makes it look like it's really easy to build agents. And it is. What nobody talks about is how difficult it is to design systems and processes and workflows.

0:01:54 - (Alane): That's the challenge, and that is the secret sauce of your business. How you execute on the work is the thing that makes you special. So it should be the part that you spend time on designing. The agent can come in and eliminate some of the manual work or a lot of the manual work, but it's still dependent on your company defining what that is, and that's why you do what you do.

0:02:20 - (Micah): I think that's so, so interesting to think about. I. I feel like I haven't thought about that for years, even though maybe I shouldn't admit that on our podcast. But it's. It's either, you know, kind of in the back of our heads or we get caught up in the, like, what's the next thing that we can do? And what's the next thing that we can automate? Yeah, we called this podcast automate your Agency for that reason. There is so much that you can automate, but being able to sit down and say, what's our differentiator?

0:02:52 - (Micah): And tell me if this isn't where you are going, but like being able to say, what's our differentiator? It's how we execute and then what makes it special about how we execute that gets stuff done faster or better or more securely or you name it. Like, for us, we are not the fastest at building things out. Like, you could hire anybody on upwork and they're going to be super fast, but it is garbage in, garbage out. If they didn't take the time to design it, if they didn't understand how to build a test suite, if they didn't understand what you're really looking for, sure it's fast, but are you going to use it? Are you going to want to use it?

0:03:37 - (Alane): Probably not. I mean, we, we kind of did that early on where it was like, great, that's what you want to build. We're going to build it. And we didn't value it ourselves. On working with the client to define what their process was to really get the outcome that everybody desired. And we had to implement that. That. And when you say we're slow, it's not that we're slow, it's that we are so methodical when we are working with our client to really understand the workflow, how it works.

0:04:06 - (Alane): Because if you get to the end and it's not what you wanted, you wasted weeks of time to go back and then change what you want and to get a new outcome and desired outcome. So it's that time that is actually the most valuable. The quickest part to execution is when all of that work was done and understood.

0:04:29 - (Micah): Yeah, yeah. In fact, I would even say 80% of the work in building automated systems these days is design.

0:04:37 - (Alane): Yeah, I would agree. Because there's some of clients that I think about that we might have spent six weeks planning with them and development time was only two weeks.

0:04:46 - (Micah): Yeah, there's a lot of scenarios like that. So we actually are fast.

0:04:50 - (Alane): We are. And it also depends, and this is why I wanted to bring this topic up is when a client comes, or, you know, not even a client, just if somebody goes, I want to automate this process. Can you define and articulate that process in every step? Because if you can't, you're not ready to automate yet. And sometimes it does take a third party to help you articulate that. I mean, Micah used to do that with me early on and go fan base, our previous agency there was because I had been with you from such an early beginning and understood Every single client. I knew what 350 clients, what they purchased and what package they were on to.

0:05:36 - (Alane): Like, everything. And then when you told me we needed to hire and hand off invoicing to someone new, I could not tell you how to do that. I couldn't tell the person that we had hired how to do invoicing.

0:05:53 - (Micah): That was a fun afternoon.

0:05:55 - (Alane): Oh, yeah, yeah. Lots of delightful conversations. But I needed a third party, you coming in and working with me to be able to articulate it. It wasn't a fast process. I had to think of. There's a lot of if thens, then buts kind of things that happen in a process. And if you can't articulate it, it's even harder, which means you might not be part of the entire process.

0:06:22 - (Micah): Well, I would even argue that's where companies like ours actually provides the most value. It's not so much in the building and the deployment, it's in the strategy. So somebody comes to us. And I mean, your case is a great example. Even if it's internal. A team leader, an executive says, I've got to get this off my plate. I've got to delegate. I've got to build a system. Whatever it is, it starts with a desired outcome.

0:06:52 - (Micah): Right. So in your case with this story, the desired outcome was, I need somebody else to do invoicing. We've got to delegate this. Great. How do we get there? I remember that afternoon, paraphrasing it. It was a lot of. I don't know, I just do it.

0:07:15 - (Alane): I'm. I'm a delight to work.

0:07:17 - (Micah): You are. So I think that's indicative of so many leaders, though, and so many managers and so many people, is you do the same thing over and over, or you're looking for a desired outcome. And it could be that it's so ingrained in what you're doing that it's. It's hard to articulate it even if you can do it.

0:07:35 - (Alane): There's another part to this. It's. It is easier when it's one person that does the process to get. To get to an end result. I will tell you the other thing that I see. When a company wants to go automate something, it is. It doesn't involve one person executing on that process. It could be four different team members. So the way that the work is done is inconsistent.

0:07:59 - (Micah): I'm not saying it could be four different teams.

0:08:03 - (Alane): So there's a couple of different scenarios. But the one scenario that I want to focus on right now is you have for salespeople how they go from initial sales call to sold client can vary. When do they send the proposal? What does their proposal look like? How do they collect the information? What software platforms do they use? Because it's amazing to see how many companies are just the wild, wild west of software platforms and their org. Different. Different topic, but that.

0:08:35 - (Alane): So then to get a process articulated and to automate that process, you need each team member to say what their process is. And then you need the company way of doing that process. That takes a while to do.

0:08:50 - (Micah): Yeah, I mean, we could do, and probably should do a whole episode just on standardization, because that is so. Like, this wild west is so difficult. Does it work to grow fast 100%? Does it work to scale 0%?

0:09:04 - (Alane): Yeah.

0:09:05 - (Micah): When everybody is doing everything separately, you can't automate, you can't train, you can barely hire. And anybody that leaves is leaving with really good information in their heads. And then going back to the original topic of how much work happens upfront with designing a workflow. And so we've established that, hey, you have a desired outcome. You've got this concept that you're going from point A to point B.

0:09:37 - (Micah): But in order to define this, you can't just throw an agent at it. You can't just say, hey, agent. Hey, AI. Just do this and hope that it works, because it's not going to. It can get down to incredible nuance. Like, what's the forms options on a field? On a field options on a form.

0:09:58 - (Alane): Mm.

0:09:59 - (Micah): What's the columns in a ClickUp list or Monday board or Asana project? What data points do you wanna see in the outcome?

0:10:12 - (Alane): Yeah.

0:10:12 - (Micah): In what format?

0:10:13 - (Alane): So a lot of times, whenever, like Micah, just like you said, bringing this back around is when we start in talks with a client, they're always thinking, what can I do ahead of time? Or what can I be doing in between? And the most most valuable thing that you can do, if you're looking to automate something, is to start documenting what that process looks like or putting together the standard operating procedure for how something works.

0:10:40 - (Alane): You know, you have some ideas, most likely, of what you want to automate. You need to know what that process is. Pull in the team that does that process. Get a whiteboard, take a picture, get Miro or Lucid and start putting pen to paper and start writing this out. Because an agent can follow a process, but it has to be defined for it to be built and to execute on every single time.

0:11:07 - (Micah): Yeah. And even when you're. I would say an agent can follow a process. But it is even better if it's following a specific part of a process and doing its role. And the only way you know what that should be is if you diagram or document or list it out.

0:11:28 - (Alane): There's so many micro processes that happen within a process. We have. We. And we just. That's what we started doing. Instead of tackling these giant processes for ourselves, we just did little micro ones. There's some that are so simple but save so much time. So we just said let's just put a chat interface on top, query it and. And get something that can be a fast way to execute on just something really small without maybe we don't have the whole process. Maybe as an AI agent right now.

0:12:01 - (Micah): Yeah.

0:12:01 - (Alane): And that way actually execute on something.

0:12:04 - (Micah): Absolutely. I mean I love that as kind of wrapping this up. Alane is. It's. It's going to start with a big idea. Everything starts with a big idea. This is my desired outcome. This is what I want to try to get to. And I'm going to go from point A to point B and it's going to be awesome. Hey everybody. This is going to be this epic thing. But then you try to take that on. Like we've talked about this in other episodes where it's like writing all the sops at once.

0:12:31 - (Micah): That is a never going to happen undertaking. Start with two SOPs, the biggest thorns in your side today.

0:12:39 - (Alane): Mm.

0:12:39 - (Micah): This is exactly the same. Don't start with a hundred step giant system that's gonna solve all your problems. Start with two miniature workflows that's going to solve the two biggest problems within that process and then iterate.

0:12:58 - (Alane): Mm yeah. Like I remember the very first one that I automation that I wrote in my life, all it did was from pipedrive. One deal that information went into Asana, into a description. That was it. That was the whole automation. Because then everything was there in one place where the team did client work for client onboarding. And then from there then we could add new stuff. We could add. Okay, well we also had a form filled out. Okay. Where does that information go? Okay, well we also create this Dropbox file or whatever it was at the time.

0:13:42 - (Alane): We then built on top of it. And that's what you can do with automation. It can build. It's not. You build it and that's it. It can never change after that. You can add on to it.

0:13:51 - (Micah): Absolutely, absolutely. And that is a lot easier to design those little ones than it is to design the big ones.

0:13:57 - (Alane): That way you can get to automating your work faster. 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.

0:14:21 - (Alane): Join us next week as we dive into more ways to automate and scale your business. Bye for now.