Elizabeth Zurn Lujan: Welcome back everyone, that was such a great activity, thank you so much to Brent for guiding that and for his hosting throughout these last couple days, it’s been really amazing. So I am Elizabeth Lujan, I am your co-host here, and I am one of the founders of Vinrock Media, we’re a digital marketing agency, and just like you, I am very excited about the opportunities that come with integrating AI into our agency. I’ve learned so much already and I’m sure you have too. So next we have a panel session on the topic of from chaos to clarity, building SOPs with AI agents. And so we’ve got two amazing panelists who are going to be joining us. We have Robert Patin and Karl Sakas, and so they’ll both be coming in now. Hello, Robert, thank you. So Robert is a creative agency growth specialist, a three-time best-selling author, and the founder of Creative Agency Success. He helps agencies grow smarter and more sustainably. And then we also have Karl here with us, he is an agency adviser and executive coach helping agency leaders work less, earn more, and build high-performing teams. So we really appreciate the two of you coming in to share your expertise with us. So just to kind of kick things off first, I know that, Karl, you’re taking kind of a leadership approach to things, so kind of tell us a little bit about your approach to the topic today.

Karl Sakas: You know, in my work with hundreds of agencies all over the world, one of the questions is how do you do more with less, and AI gives you some solutions with that, and particularly AI agents around how do you accomplish more with the team you already have. And the things we’ll be talking about today will help fill in some of those answers. As you mentioned, my approach is especially around leadership, thinking about your human team and your AI team.

Elizabeth Zurn Lujan: Great, wonderful. And then Robert, tell us a little bit about how you approach this topic.

Robert Patin: I mean, generally speaking, I’m a very operational mind, so I’m going to approach it kind of from two different angles. First, the traditional-oriented SOP, but how to be able to create them, define your systems within your agency right now to create more clarity, and then two, how to actually augment your existing process with AI agents, with custom GPTs, with that sort of thing, that allows for you to be able to do more, act faster, being able to provide a more systematic way in the way that you operate.

Elizabeth Zurn Lujan: Wonderful, wonderful. Well, Karl, I had a question to you to kind of get us kicked off here, and I was curious, where do you see these AI agents making the biggest positive impacts in day-to-day agency operations?

Karl Sakas: I would focus on two areas for everyone who’s listening. You’re like, where do we focus? Two areas. One of them would be what are the areas that are happening most often in your agency, for instance things like client onboarding, that’s happening every time you have a new client, or client invoicing, that’s happening every month, or at least when you get started, and auto-drafts and all that, those are some opportunities. So, what happens frequently. The other category would be things that don’t happen as often but they tend to be very time-consuming, or if something went wrong it would be catastrophic. Those particulars may be unique to your agency, but I would take a look at that, what are the things that are super complicated and what are the things that happen all the time.

Elizabeth Zurn Lujan: Okay, great, so that’s really helpful to give us kind of a limited scope of where to start instead of just being overwhelmed. I can’t tell you how many times when I’ve been working with agencies, they’ll say, my goal for this next month is I’m going to make every single SOP that we need to have for our agency, and I’m like, okay, I know you’re not going to do that in the next month, right, we got to start somewhere, and that’s a great way to give us clarity on that. So Robert, in terms of actually building these AI-powered SOPs, what does that look like today, can you share a real example of how that actually happens?

Robert Patin: Sure. I’ll take a step back for a second and just start talking about SOPs to begin with, and I want to clarify something that’s really important, kind of off the back of what Karl was saying, is that I think it’s really important for us to recognize that if we put bad information in, we’re going to get bad information out. And so what I see happen in so many instances, with the vast majority of agencies that I come across, is that there’s too much variety, everything’s bespoke, we don’t have a systematic process in which we’re doing things in the first place, and that if you don’t have a process, AI isn’t really going to help you in that instance, you’re not going to create order out of chaos. You have to go and do the underlying logic, thinking, inputs, to make sure that you’re able to actually create clarity. So AI is not going to replace you fully, think about it as a junior-level team member and how can you leverage that fully, and take something and be able to streamline it.

So there’s, as I was saying, there’s kind of two different angles that I wanted to approach it. Of how do we take and document our existing SOPs, our existing processes that we’re already following, and you have them, and making sure that you have them documented so they can be repeated with every single team member. And so the way that I like to do this, me being the Excel geek that I am, taking just a very simple kind of Google sheet and creating a prompt creator that allows each person to actually create a video. I think that being able to just document a quick Loom video, or a video of the step-by-step process that you’re actually taking, and making sure that you do it in a very systematic kind of format, following the traditional kind of SOP format of, what is the purpose of this SOP, any related documents, templates, things that are going to be involved, and then the actual process that you’re going to follow step by step, the frequency that’s going to be involved, who’s responsible for doing this action, and if something goes wrong, who does it get elevated to, and who needs to be involved in order to actually correct this specific challenge that you’re going to encounter. And then actually going through that in the video, and then taking the transcription to allow AI to create a documented format. The reality of any process is that some people prefer visual, auditory, and then reading, as far as learners, and if we cover all three, we are able to actually cover the gamut. And sometimes I just want to quickly get to an answer.

And then on the latter, there’s so many different ways that you can go about actually leveraging AI to make your processes faster. But as examples, it could be, how do I take a proposal, how do I take a sales call recording and generate a welcome email to a client in the onboarding process that is automated, it’s already drafted, that’s in your outbox, that you can send out, that shares excitement, shares a little bit of background, that creates this really welcoming feel to your clients, while being able to take the questionnaires that are sent, being able to then distill that into a creative brief that is done in an automated fashion, that allows for you to be able to do things faster.

Elizabeth Zurn Lujan: Great, so am I understanding you correctly that you’re still recommending that the human generate the process first, right, and then we’re using the AI tool more to follow it, so if you don’t already have your process, AI might not be the first step, it might be to actually identify what the process is and to systematize it.

Robert Patin: Okay, so I would add on to that, just I would like ask AI to yes-and you, right? So here’s what I’m doing, what am I not thinking about, what areas do I need to cover, yes-and you, not actually replacing.

Elizabeth Zurn Lujan: Okay, okay, that’s great, yeah, because for example, yeah, a follow-up email, some people might not be doing that from a sales call, so it could be an additional step. Okay, so as we’re starting to integrate these AI into our SOP systems, what KPIs should an agency be tracking after they’re doing that integration to see that they’re effective? Yes, Robert.

Robert Patin: Yeah, um, it’s a really broad one, but I would say that there’s a couple key ones that I would be looking at. Is, what are the adoption rates of the existing process and systems that I’m currently using, which is a really important one. The effectiveness of that, so how much time is being saved, so if you’ve got time tracking within your agency, now I’ve implemented this process, did it make it faster, did it make it slower, what impact did that ultimately have? And if we’re talking about processes that are powered by AI, meaning like the example that I was saying where draft emails being done, is, what is the frequency by which human intervention is necessary to actually override AI? I really think that it’s important, especially at these initial kind of infancy stages of AI, that you’re testing, because the reality is it can kind of go off in left field sometimes and do things that we don’t want it to do, make up information and do all sorts of stuff. So you want to make sure that you’re doing it with a reasonable amount of accuracy, and so testing until the point that you feel comfortable.

Elizabeth Zurn Lujan: Okay, so a lot of hands-on it still here, as we’re in this first stage. So Karl, you have so much experience, helping leaders become incredible leaders of their organization, so what leadership challenges do you see arising with AI SOPs, and how can teams and leaders adapt to that?

Karl Sakas: So, first, a challenge that’s not arising, which would be with employees. You know, if something isn’t getting done, or it’s not getting done well, I think of a Venn diagram, desire, competence, and capacity, you need all three of them. If you assign something to someone, if they don’t have the desire to do it, they’re probably going to avoid it, if they don’t have the competence, they’ll do it poorly, and if they don’t have the capacity, they don’t have the time to do it, they’re probably not going to get it done. Now, thanks to AI, eliminate those three things, whether the AI wants to do it, I mean that gets into a question of consciousness, but okay, desire to do it, competence, and capacity, that’s all set. But there’s another thing, which is what I would describe as your values, goals, and resources, your VGR. If you’re asking someone on your team to do something, if you don’t give them what they need, they’re going to keep asking you questions about it, they’re going to keep asking, well, this popped up, what do I do, or the client said this, what do I do now? If you’re running a smaller agency, you’re still going to get pulled into that, but as your agency grows, if you keep getting pulled into questions like that, it is time to empower your team, and one of the shortcuts to do that is what I call VGR, values, goals, and resources. The idea is that you want to make sure that everyone in the team knows the values to follow, that is, how do you operate, the goals, what are you trying to accomplish, which could be at the agency level, it could be at a client level, it could be at a campaign level, it also could be for their role, what are their goals as a project manager versus an account manager versus a strategist, and the final is resources, what are the resources they have available in terms of time, money, software, what have you. You know, for instance, if you build a system that’s really great at getting things done but you’re not feeding it with the appropriate data, you’re not going to get great results, so make sure that your AI team members know the values, goals, and resources, just the same as your human employees.

Elizabeth Zurn Lujan: So it sounds again similar to what we were talking about earlier, your AI is only going to be as healthy as your organization in the first place, right?

Karl Sakas: Exactly.

Elizabeth Zurn Lujan: I’d love to kind of dive a little bit deeper into this with you, Karl, and talking like, how do you ensure that AI SOPs reflect company values and culture, I mean, how do you make your robot have values?

Karl Sakas: Yeah, well, I mean, they can’t read your mind. They can, to some extent, for instance, if you loaded your employee handbook, they would have some ideas of things, if you loaded your website. Though, I have to say, you know, agency owners will reach out to me about getting help, and they’ll say, we focus on such and such industry, and I’m like, that sounds good, that’s not necessarily what their website says, right, do things reflect that or not? But the starting point would be to get clear on your mission, vision, and values. A lot of agencies have values, they’ll be on the website, if you have a physical office you might have them painted or otherwise posted on the wall, you might have a sense of a mission, you might have a sense of a vision. The question is, does everyone on your team know what those are, the mission, vision, and values? Often the answer is no. So if you’re listening here and you’re like, oh, I don’t know if we’ve done all that, it’s okay, but now it’s time to do something about it. When I onboard new clients for my agency growth diagnostic, I ask, what is your mission, what’s your vision, what are your values, and what would your team say if I asked you what were those at your agency? And often the answer is that, well, I don’t know, we kind of have a mission, maybe it’s in my head, we kind of have some values, like we talked about it once, and I don’t know, a vision, I just want to get through this week, you know, it’s hard to think further out. That might be a sign that you’re stuck in firefighting rather than shifting to being the fire marshal, as I share in my Calm the Chaos book. So, getting clear on your mission, vision, and values, I’ve just shared in chat a template that I use with my clients on determining your mission, vision, and values, that’s a direct link to the Google doc, there’s no opt-in required or anything like that, you can download it yourself. And it’s going to take some time, I mean, to Robert’s earlier point about kind of putting the bad data in, before you start automating and AI-ing things, you probably need to have some very human conversations about what you want, what you’re trying to accomplish. It’ll feel like it’s slowing down the technology, but this is a vital human step before you proceed.

Robert Patin: I’m also curious, Karl, on your take, because, interestingly, last week I was meeting with a client on-site, and in the leadership conversation there was a lot of conversation around AI and implementation of AI, and one of the challenges that we were facing as a leadership team was a discussion with the team actually actively fighting, I mean to the point that there was conversation about the ethical component of AI and the green component of AI, and largely, these are conversations to be had, but the basis of it, to me, was fear amongst the team, and am I going to be replaced, and there’s a lot of resistance in actually leveraging it, there’s a whole lot of unknown in the space. And what are your thoughts on combating that challenge, because that’s a significant one that so many agency owners are going to face?

Karl Sakas: I definitely hear that at agencies I work with as well, where often the owner or owners and the leadership team are very much on board with applying AI, but frontline employees may be concerned. Now, of course this has been a concern throughout all of capitalism. I had a client, you know, pre-AI, their paid search team was really overloaded, and they brought in contractors to help, as freelancers, to help reduce the employee team’s workload, and the employees panicked, they thought the contractors were there to replace them. That was never the plan, and it wouldn’t have made financial sense, it’s like they’re paying twice as much per hour to the freelancers generally as they were to the employees, it would not have made any sense, but employees took that perspective and they kind of ran with it. So it is for sure a legitimate concern about what are you going to do. I do see better results when agency owners and leaders talk with the team about what they’re trying to accomplish and what success looks like. You know, if you’re telling the team that success is you’re able to work on more strategic things while custom GPTs and AI agents and other things are taking care of the less fun part of your job, that we’re not doing this to replace you, we are hoping to make your work easier, that’s a bit different from, we’re here to take all of your jobs. But some of that is kind of built into capitalism generally.

Robert Patin: Yeah, I would also say one of the things that I really like to do, from a pragmatic standpoint here, is that there is a lot of unknown, and there’s the principle of kind of like failing forward, and the reality is, in testing of AI there’s been so many pieces where I have absolutely failed at it and it didn’t work, and I tried, and then this didn’t work, and I tried another idea, and actually just kind of sharing that experience that you have with it, as a leader of the organization, to share, hey, I’m testing something, it didn’t work, here’s what I did, here’s what I’m experimenting with, that allows for them to feel that failure is actually kind of part of the process to success, and so that it doesn’t seem quite as scary as well. I think people will also see that AI is not ready to take an entire human’s job for most agency roles, right?

Karl Sakas: I’ve built tools to make my life easier as a coach, as a consultant, as a trainer, and it does, can they completely replace me? Not yet. I mean, I’m working on it, but it’s not there yet. Part of it is just differences in, going back to data sources, I did a survey with Databox, the analytics firm, and one of the things we were looking at was comparing profit margins, like, what do the most successful agencies do that’s different to market themselves than others, part of it is making time for the self-marketing. But one of the things we wanted to look at was what are the profit margin differences, and people could link their QuickBooks, they could link Xero or other things, what we found though was we had to manually ask people their profit margins, because QuickBooks API-wise and Xero API-wise report net profit margins differently, and so we would have had to manually calculate things to compare, if people even had it loaded correctly in the first place. So, you know, we are a ways away from AI taking over. But certainly using AI as a helpful colleague, a sounding board, a writing partner, the analyst so that you don’t have to do quite as much in Excel, that’s where it is now. Now, where will things be in 10 years, who knows.

Robert Patin: The analogy that I’ve liked to use recently is like the age of the dot-com boom, right, where e-commerce was really taking off, everyone was saying retail’s dead, we’re not going to have stores down the street anymore, and I’m pretty sure that Target still is a mile away from my house. And while there’s a lot less being sold in retail and e-com is definitely a predominant source, the amount of packages that arrive at my front door is proof of it, but it’s been that way, we’ve been talking about it for nearly 30 years.

Karl Sakas: It is fascinating to see what Amazon will deliver, you know, if you’re near one of the warehouses, same day, or, you know, at like 4 a.m.

Robert Patin: Yeah, yeah.

Karl Sakas: I mean, for instance, you know, some stuff is not surprising, toilet paper, you know, cold and medication. I did notice that there’s the book Fair Play about delegation in relationships, and you basically get a card deck, you can divide up who’s in charge of the trash, who’s in charge of medical appointments, who’s in charge of this or that, that, it turned out, was available for delivery as early as 4 a.m. the next morning.

Robert Patin: Yeah, I’m sure there were some marriage fights, relationship fights, that needed that rush delivery in order to have a good day the next day.

Elizabeth Zurn Lujan: So, everything you guys are talking about is so inspiring, I have kind of a practical question, well, it’s practical but also very values-oriented. So one of the core values of my agency is that we invest in relationships, and that, to me, has been really kind of our center core value, because it really has been what’s generated our revenue, it’s been what’s kept our company healthy, what’s kept us with incredible partners. And so as we’re looking at that and looking at how these AI SOPs might impact client relationship, client delivery, onboarding workflows and things like that, I’d love to kind of hear, maybe Robert, your take on that value, and how you can see that impacting AI SOPs, client delivery, client onboarding, client relationship, all of those pieces, and then Karl, I’d love to hear from you on that as well.

Robert Patin: I mean, I think it depends on how you approach it, right, because it can actually further deepen relationships. So here’s a practical example. I’m a big gifter, bit of a love language of mine, I like to really pay attention to what the person enjoys and send a gift that’s really personal to them. And I sent a gift to a client recently that was very purposeful for her, and it was powered by AI, and she absolutely loved it, felt very seen. I took some detail of conversations that we had, things that she enjoys, and asked it to do some research, give me 20 different options of things that I could potentially buy. It actually didn’t give me the exact product that I ended up buying, but it led me down a path that resulted in me finding the thing that was actually really perfect for her, and she absolutely loved it, she felt again very seen and recognized. So it actually deepened our relationship, and it allowed for me to be able to move faster, something that would have taken me several hours, probably walking my dog to be able to come up with the idea, and I may not have even gotten to something quite as meaningful as what I ended up choosing. And so it can actually deepen the relationship if you use it appropriately.

Karl Sakas: I am a fan of gifting as well, I’m generally spending at least $2,000 a year on postage in my business. Yeah, it, I mean, that idea of showing someone that you were thinking of them, I was on a trip to Australia several years ago, and I noticed that right across from my hotel was the Australia corporate office for one of my clients, biggest clients, so I took a photo, sent it to them, and was thinking of you. Will AI know everything about us? I mean, certainly things, you know, if we’re talking about like brain implants that could know everything we’re thinking or imagining or what have you, then sure, but is AI necessarily going to know, oh, hotel company logo sign, client, biggest client, let’s take that photo and send it to them? Eventually maybe, but right now that still requires humans. I agree that relationship building is very important, although I’ve designed my work to be remote, I love meeting with clients in person when I can, was in New York earlier this week and met up with an agency community leader, finding those opportunities is important. So I don’t think AI will eliminate that, but as Robert said, you can find ways to strengthen things. I mean, even not as AI related, but even things like Google Alerts, getting a notification about some news for a client and you can congratulate them, I had one case where a past client, we’d worked together before, they got acquired, I heard through the grapevine they had gotten acquired, and I sent them a note, hey, congrats, and they’re like, are you some sort of ninja? But the news had just happened that morning, but I happened to be plugged in to hear about it. So whether it’s kind of ninja, you know, it worked. I have had more than one client say I am the most organized person they’ve ever met, but technology helps, and my team supporting me on that, and yeah, using AI tools as well. So, agreed on relationships.

Robert Patin: And I think the key thing for everyone listening, you can’t hide behind technology, you can’t hide behind emails, texts, and so on, those have their place, but there still is value to having a conversation.

Karl Sakas: Agreed, at least for now.

Elizabeth Zurn Lujan: I hope it doesn’t replace human-to-human connection, because otherwise I’m not sure what is the point of existence at that point, but anyway. One theme that I’m taking away from some of the conversations I’ve heard over the last couple days is, it’s this transition from thinking of ourselves as workers doing the heavy lift of, this is so hard no one else could possibly do what my agency can do, and saying, actually, probably a lot of agencies could, you know, lots of things are becoming easier to do with AI, and they will continue to do so, but no one else can have the relationship that you can have as a person. So my hope for AI is that it really does bring us more into realizing our value is in relationship and in finding those moments of connection, rather than just creating a whole bunch of deliverables and work, right, and just being the hardest worker, or even the most results, right, we still want that, we still want great deliverables, we still want great results, but it’s still got to be this human thing, that’s what sets us apart. So, this is not on the approved question list, so I’m gonna, but you guys, I just love hearing everything you say, so I wanted to get your feedback on this idea of using AI to challenge and improve ourselves, to fill in the gaps in our agency of what we’re not thinking of. And I was curious if, first Robert, in any of this AI SOP process implementation, did AI help you identify any gaps or anything that you needed to additionally say, oh, we should be adding this into our process, or, hey, there’s this whole other opportunity we’re not seeing, has AI brought anything like that to the table for you?

Robert Patin: Frankly, I’m sure that it has, I don’t know that I can think of an exact example where it has done that specifically. I mean, I do use it, as I was describing earlier, as kind of the yes-and, and just like, here’s what I’m currently doing, what am I not thinking about, or how can I expand upon it, like that’s how I use it practically on a daily basis, if not an hourly basis.

Elizabeth Zurn Lujan: Tell me about the prompting that you do for the, what am I not thinking about, how do you craft a prompt around that?

Robert Patin: I mean, it depends on the specific scenario, right, so if I’m looking at it from an SOP perspective, I might take that transcript from the video that I was describing earlier and take that transcript, upload it, and say, you know, here’s the existing process that I’m following right now, I’m encountering this problem that is coming up on a regular basis, which is foundationally how I like to look at the creation of process, and when do I need to be creating something, is, I’m encountering a problem, I like to only solve things once, I don’t want to solve the thing over and over again and kind of live in Groundhog Day. So, how do I even expand and think about what are the potential solutions to this problem, and that it might actually spark an idea that allows for you to solve it fully in perpetuity, or it might actually help you be able to uncover something that you would have never recognized in the first place. So that would be an example application.

Elizabeth Zurn Lujan: Karl, any thoughts on that, about, I mean, the hope for AI is not that we just keep doing the same thing we already do faster, right, it’s beyond that, so what do you see in terms of crafting any ideas about crafting prompts that could be helpful for us?

Karl Sakas: I mean, ideally it’s faster and better. I am a big fan of asking, you know, I’ll share and I’ll say, what questions do you have, and then keep asking that, do you have any more questions, I’ll answer, this is more in like a ChatGPT setting, any more questions, and eventually if the AI runs out of questions to ask, it’s probably ready to figure out what to do next. So, you know, that can work. I’m also a fan, for both humans and AI, of a concept called intelligent disobedience, this comes from training service animals, service dogs. The idea is that, if you know, say the owner is like, let’s cross the street, and the dog can hear that a car is coming, they will be intelligently disobedient, they will refuse to cross the street, they know it’s dangerous. Same thing is true when I’m hiring team members, where I want them to push back, if I’ve suggested something that they know to be contrary to my goals or contrary to something else, I want them to say, that’s a bad idea. I had that recently where I was considering a new initiative and I asked the team what they thought, and they shared a number of concerns, some that I hadn’t thought about, you know, I as an owner was being overly optimistic and they were giving me a very helpful reality check. So, intelligent disobedience, I talk more about that in my fourth book, latest book, Calm the Chaos, on 10 ways to run a better agency, you know, I wrote that around intelligent disobedience for human employees, but it could work for your AI team as well.

Elizabeth Zurn Lujan: Yeah, that’s great, because there is this trap of AI that it really wants to create something that you like, so it could be the ultimate yes-man of just kind of becoming a limiting factor. So I love that idea of that intelligent disobedience. Kind of transitioning a little bit into some really more practical elements, Robert, I would love to hear what AI tools you use for SOPs and what your experience has been with them.

Robert Patin: So that kind of covers the gamut really. I mean, as I was describing earlier, if I’m looking at just creating an SOP, that my go-to would be to start with is a Loom video, I’m a ridiculously heavy Loom user, and so creating a Loom video that goes through those components that I was describing, of what exactly is the purpose of this SOP, who exactly is going to need to be involved, the frequency by which it needs to be utilized, any existing templates, that sort of thing, and taking the transcript of that to allow for me to be able to document it. But if I’m going to expand beyond that and actually start to look at how do we actually create additional systems that allow for me to be able to create AI-powered process, I’ve been dabbling quite a lot with a platform called CrewAI, that allows for you to layer different AI-oriented platforms, and it’s quite nerdy and a lot of fun, and I’ve been really playing with it. There was, frankly it was, it’s been several months, like three or four months ago, I was laying on the couch one Sunday afternoon and got an advertisement on Facebook for an application called Coefficient, and it was describing that it could connect AI to Google spreadsheets or Excel, and I’m an Excel geek, like absolutely an Excel geek, and I was so excited and went immediately to my computer to go play with it, I need to see it. It didn’t actually work of what I was wanting to do, and then ended up figuring out a way to go about doing it, and then ended up like walking outside like incredibly energetic about it, and being able to actually just think about how do you take existing platforms like Zapier into a Google sheet that allows for you to be able to reconnect it back into ChatGPT or into Gemini. There’s examples within existing platforms, like you see constant adaptation and addition of AI-oriented pieces into existing platforms, I really am enjoying what ClickUp is doing with its AI bots as well. I’ve built probably, at this point, far too many different custom GPTs for specific purposes, examples of clients taking their entire strategy method, specific to their industry, of how they go about managing specific metrics, to what exactly is the client having as a challenge, what are the potential mechanisms and service offerings that could actually combat or help solve that client’s specific challenge, taking recordings of client calls that actually kicks out directly what would need to be done, or a custom GPT that would even help you draft a proposal, that here’s the different pre-written language that I have for different services, that allow for you to inject components of what was discussed with a client. I mean, the possibilities are endless.

Elizabeth Zurn Lujan: Yeah, as you say that I get really excited, but also I get, I seize up a little and I get a little bit paralyzed, like, this is a little too much, right, and that takes us exactly back to the topic we’re talking about, from chaos to clarity, right, and it’s such a balance of how do we make this a tool that’s something that’s actually useful for us. So I’d love to kind of go back to this bigger outlook on everything, and Karl, I’ll put this question to you first, and then Robert, I’d love your input. So, as you’ve seen agencies overwhelmed by too much process, how do they fix it? You know, somewhat with AI, but the problem is the chaos, right, the problem is not that we don’t have enough tech tools, right, the problem is the chaos, so how do we fix that?

Karl Sakas: Your point about, you know, we don’t have too many tech tools, or there are too many, I have an article on project management software for agencies, and it is approaching a hundred different options, and I keep discovering new ones, so, like, if you’re thinking of building a PM tool, there are plenty out there, you probably shouldn’t build a new one. But yeah, I mean, just that sense of overwhelm. One thing to consider, if you step back, do we need an AI solution, or do we just need to automate it, or is it some combo, or maybe there’s something that you’re doing now that you should just stop doing. For instance, one of my first jobs after college was as a sales and marketing analyst for a small mutual fund in New York, and one of my responsibilities was to create market share reports every quarter that would go to the CEO, the COO, and our chief investment officer. And I inherited this process from a previous team member, and she was like, here’s the process, it’s going to take you about two weeks, so like 80 hours of work every quarter. So I followed the procedure, I followed the SOPs, and I felt at the end of it like I was done, I’m like, this isn’t a very good way to do this, there’s got to be a better way. And so the next quarter I made some tweaks, and then the third quarter, before I dived into just doing it, rather than say fully automating, or, you know, we didn’t have AI options at the time, I went to the stakeholders, the C-suite team, and asked them, I said, I’m thinking of ways to streamline this to get it to you faster, right, that was the key word for them, getting it faster every quarter, and basically asked them, what do they like and not like about the existing reporting, was there anything they wish they had, anything they didn’t care about. It turned out some of the things they didn’t care about that my predecessor had just been building every quarter, and no one looked at them. I also followed up with the data provider, and it turned out that we got custom consulting time with them every year for them to pre-modify the data before it got to us, so actually I wouldn’t have to do as much analysis, they would pre-analyze a lot of it. And ultimately I went from a process that took two weeks every quarter down to just two days, from 80 hours down to 16.

Robert Patin: Yeah.

Karl Sakas: Yeah, so, ask yourself, do you need AI, do you need automation, or should you stop doing this thing altogether?

Robert Patin: I think the core of what you’re saying really is defining the business challenge effectively, really, what exactly is the challenge, what exactly is the opportunity within this challenge, and how does it actually fall into the hierarchy of priority effectively. I mean, the reality that there is endless possibilities, that there’s endless different applications, and that’s kind of the plight of the entrepreneur effectively, right, like, we see opportunity in so many different areas, which then ends up creating shiny object syndrome, and then the next kind of layer of it is, what exactly is the risk factor of doing it, so like, what is the likelihood of success, what exactly is that going to create, what is the risk to clients, what is the risk to the business, what is the risk to the culture, oh, I think I’ve lost Robert.

Elizabeth Zurn Lujan: Same, same here, right in the middle of, oh, shoot.

Robert Patin: Oh, am I back?

Elizabeth Zurn Lujan: Oh, you’re back, you’re back.

Robert Patin: At what point did you lose me?

Elizabeth Zurn Lujan: You were talking about, it was just like a sentence or two back, sorry, go.

Robert Patin: So, what I was saying was kind of risk assessment, of really understanding what is the risk to clients, what is the risk to business, what is the risk to culture.

Elizabeth Zurn Lujan: Oh, the internet does not want us, the technology does not want us to hear what Robert has to say, it’s too powerful, I guess. Well, we are coming toward the end of the session anyway, and so I know it’s about time to wrap up, so I’m sad that we’re missing those final thoughts from Robert, but Karl, I, yeah, Landon said in the chat, Robert broke the internet. There you are, okay, whatever you’re trying to say, Robert, the internet does not want us to know about it, apparently not, it’s too hot for us to hear, apparently.

Robert Patin: And each one of us have a different kind of risk profile, effectively, is what I was kind of getting to, and knowing whether, if that worst-case scenario were to come to pass, is that something that you’re willing to allow to come to pass, and then from an opportunity perspective, is defining what the opportunity is to the client, is that going to have a significant amount of impact to them, what does that actually have to my business, to the agency, what does that have to culture, what does that have to the team, both from like a personal perspective as well as a work-life perspective, and really going through, thinking about each one of those different components, and is the opportunity actually worth the outcome, is the direction actually worth the amount of time investment it’s actually going to take. And to me, being the numbers person that I am, everything kind of comes to a calculus of investment versus reward, and is that actually going to be worthwhile. And then once I have an understanding of, yeah, okay, maybe it’s worthwhile, but what is going to move the needle in my business and what I’m looking to accomplish in the short term, what is the most important thing for me to focus on. And the further along in business, in my career, I’ve gotten, the more and more that I’ve realized that the less I do, the faster I move. And so actually focusing my attention a whole lot more, and the same is true for your team, is that yes, there are endless opportunities, but that doesn’t mean it’s today’s opportunity.

Elizabeth Zurn Lujan: Well, Karl and Robert, those are great words to kind of close us out on. I just want to thank you again so much for your time today, I learned a ton, and I’m sure everyone listening really appreciates it. So thank you, guys. And coming up next, we have a session on how AI is driving DigitalOcean’s revolution, so we’ll see Brent Weaver back here to host us for that here in just a minute or two. Awesome, thank you.