Boosting Agency Efficiency: Why Focusing on Throughput Matters More Than Utilisation with Skyler Reeves

Hey everybody and welcome to this week's Agency Accelerator Podcast. Now I've recorded a number of podcasts over the years about the importance of being able to measure your team's capacity and their utilization if you want to run a profitable and sustainable agency. So when I met Skylight Skylar and he said to me that we shouldn't care so much about utilization, but rather focus on throughput and efficiency, I was really intrigued and I was keen to dig deeper into this topic on today's podcast. And, and if you're as intrigued as I am, then stick with us because we'll dive into all things efficiency and profitability.

I'm Rob Da Costa and this is the Agency Accelerator Podcast. As someone who has stood in your shoes, having started, grown and sold my own agency, I know just how it feels in the ups and downs of agency life. So this podcast aims to ease your journey just a little by sharing mine and my guests experiences and advice as you navigate your way to great growing a profitable, sustainable and enjoyable business.

So Skyler is the founder and CEO of Ardent Growth, a content marketing agency that he formed back in 2019. And before that Skyler served as a combat medic in the Iraq war and later pursued a degree in computer science and philosophy. So Skyler, welcome to the show.

Thanks for having me on, Rob.

Now that's a really varied career. So what led you from doing all of those things previously to. To jump into the agency world?

That's a good question. I think it's hard to trace back and really say like, here's what led me here. It was, I've always been interested a lot of different things. I think when I graduated high school I didn't know what I wanted to do, which I'm sure a lot of people don't. I figured, well, let me go, let me go to the military and let them mature me a bit and give me some time to grow up and learn about the world before, you know, I decide what I'm going to do next. And I remember it was in my third deployment and I thought to myself there towards the end I was like, maybe it's time to finally go get an education. And so when I got out, went to college and worked for a while as a regular employee and then realized that I needed to go start my own thing.

Excellent. Yeah, it's interesting. I did a computer science degree at university as well. It seems like a lot of us do, and then realized that I was much better at talking to people than I was talking to computers. Anyway, so today we are discussing why agencies should care less about utilization, but rather focus on throughput and flow efficiency. So I think that's a really big, bold statement. So can we kick things off by defining exactly what you mean by that?

Yeah, so utilization. And I'll preface this by saying, I'm not saying that you should completely throw utilization out the window. I think that there's just far too much emphasis put on it, especially whenever people chase it, trying to have some sort of perfect level of utilization, even though when they try to get granular with it and things like that. So I tried doing that for several years and all that really left us with was a burnt out team and me as a burnt out person, I wasn't, you know, over time my, my own performance was beginning to diminish and I thought maybe there's, maybe there's something different here that I should look into. So went to the, started studying, you know, other approaches to work and landed on the methodologies that Toyota is kind of known for around lean management, things like that, and discovered metrics like flow efficiency, whip limits, throughput, all these sorts of things and we began to put it in place. So what they are, is with, with any system you have inputs and you have outputs, right? So this really resonated with me and it might with you too. It's a same concept as a queue in computer science, right. Things can only move through the system, you know, one at a time. You know, unless you're multi threading or something like that. But you have things that come in, you work on the thing and then it goes out. In simple terms, that's, that's what throughput is, it's what's the end result. And the way that we would look at the other two when it comes to flow efficiency is flow efficiency is comprised of two metrics. It's the, it's your cycle time and you, that's the denominator and you divide that by your active time. So active time is the time that you actually spend working on the item, whether it's you, your team, etc. This isn't, this doesn't include the, the weekends or whenever you context switch and start working on something else. It's the only time that you're actually working on it or you're not waiting on a client to get something back to you. And then cycle time is the actual total time from the time that it first begins to be worked on. So let's say you're working in a Kanban board or if you're working In a list. Let's say you start your time, it starts from the time that you actually start to do it and it ends from the time that you actually close it out. So you know, when you think about a lot of the work that we do, let's say you're doing a website or you're writing a piece of content, Cycle time includes all the time that you're working on it and you're waiting on the client. Active time is the only time that you're actually working on it and not waiting on someone else when it's not in the queue. So you divide those and that's how you get flow, efficiency.

Got it? Now, you mentioned just now and you mentioned in our sort of prep for this interview about the Toyota methodology and that you're a fan of that. So I'm like me, I'm sure lots of listeners don't know much about what that is and I know you mentioned a few terms there. So can you just quickly tell us that about the Toyota methodology?

So I'm no expert, there are plenty of other experts on this, far more than me. All I know is we began to apply it and our life got better. So that's where I'm kind of speaking from here. What I would say are kind of the key components that stuck out with me and that we've really began to adopt is we started working from a Kanban board. So a key thing is you visualize your work through, through some sort of flow. And so you have columns or stages, however you want to do it. If you want to do it in a tool like Trello or a tool like Asana, or even on a whiteboard, doesn't matter. You can do it on a piece of paper. It's just you visualize your system. It's what are the stages that you have to go through to work this. And so that's one key concept is actually being able to look at your system because that's how you're able to identify bottlenecks, understands where, understand where constraints are, where things get hung up and how you might need to either re engineer your process or think about how you might want to reapply or reallocate resources to be able to free something up. So that's one of the core concepts, is the ability to visualize your work. Another one is this concept of WIP work in progress. So you want to limit that and ideally there's no perfect math on what should your WIP limit be. And I think where a lot of people go wrong is they try to reduce it too fast. Whenever they first begin to adopt this, it's not realistic. But the goal I think that you want to aim for is you. In a perfect world, you would only be working on one thing at a time. Now, that's not always practical, right? And so what we try to aim for here is just how can we reduce the number of things that we're starting and increase the number of things that we're actually finishing? Because when we start too many things and we're trying to bounce back and forth between everything, in the long run, you end up with projects that just kind of go stale, right? And then you gotta, then you gotta spin them back up, get them started again. There's lots of switching costs associated with that. So WIP limits is another, is another key one is try to reduce the amount of things that you actually have in progress at any one time. And then that's where other concepts like throughput, flow, efficiency, all those kind of go into it as well. But another metric would be what we call work item age. And that's just how old is that actual work item? How many days has it been in progress? We usually keep an eye on that. And there's some graphs and things that you can use to identify when something is beginning to reach that point that it's taking longer than what a piece of work would normally take you. That way you can triage it and try to get it finished and through the pipeline as fast as possible.

Great. So some things to pick apart in that I think, and I've spoken about this before in terms of productivity and time management, that context switching is a super waste of time. And smarter people than me have worked out that when you context switch, it takes you 23 minutes to get back into the piece of work that you were doing, especially if it's like a kind of real thinking piece. But of course, in agencies, you know, account managers and so on, have got multiple clients to manage and they can't always, you know, tell a client, I can't deal with you now because I'm dealing with this thing. So how do we apply some of what you said into the sort of reality of busy agency life where people are being forced to multitask, if there is such a thing, and therefore they are being forced to context switch. What, what changes as an agency need to make in order to start implementing some of what you've just outlined?

Yeah, so what I would say there's the way that you apply this can be different for various types of work Right. So the way that you might apply limiting the work in progress or focusing on one thing at a time and trying to reduce context switching is going to be very, very different for, say, someone who's writing a piece of content for a client or designing a webpage for a client or editing a video for a client. Right. Versus an account manager who is really there to juggle a lot of those different pieces. Right. That said, I think one thing about context switching that we should think maybe a bit deeper on is what really constitutes context switching. Right. It's not switching between apps, it's not switching between, you know, different email conversations. It's. It's more like what's the nature of the work that you're doing? So if you're managing clients and you're triaging messages as they come in, you're replying to them, you're maybe adding things to a plan. Right. Even when you're moving between clients as you do that, if it's the same type of work, it's just, it's for different clients, but the context itself hasn't changed. Right. You're still applying the same types of skills. It's very different if you were trying to do that and then also have to go do research to understand the problem that they're asking you about and develop some sort of plan of attack against it or, you know, strategy or something like that. Those are two different contexts. And so I would try to separate those then when it comes to how do you manage this with clients when they're asking for something. I remember there was a. There's a sales coach. He's actually Mark Robert, if anybody's familiar with him. He's like a third employee at HubSpot. When they first spun things up. His dad wrote ahead a blog years ago and I was reading his old post from like 2006 or so, and he was talking about this as well from a sales context. And it was that you don't. I know it's easy. Sometimes I feel like we have to drop everything when the client wants something. And I wonder at what point we convinced ourselves that was true. And have we ever tried to say, hey, listen, I know you would like to meet right now. I'm booked until say, 5 o', clock, but I can meet with you after that. Or if that's too. That doesn't fit your schedule, we can meet tomorrow morning. What would you like for me to do? And what ends up, you know, what would end up having, I think in the long run is either a, they would say, okay, five will be fine with me. And they're appreciative that you maybe worked a little bit later to work with them. Right. They'll remember that or they'll say we can do tomorrow morning, but they'll still remember that you offered to meet that afternoon. Right. So I think it's about giving your customers choices and letting them know that you're willing to be flexible. But they're people just like us, they're bouncing things too. Right. They understand we just, we convince ourselves, I think sometimes our mind that we can't tell people no or they we can't be firm with our own time. Because also remember, whenever you say yes to one client, you're saying no to another one too.

Exactly. Yeah, yeah. No, it's so true. And my world this week has been a bit immersed in this topic because I ran an over servicing workshop for TOGGL yesterday and I've been writing an article for somebody else on client service in 2024 and what we need to be focusing on. And you know, yeah, I mean, like you said, when did we forget that we actually can, you know, have an equal partnership relationship with our clients and push back? And you know, how many people are revisiting the kind of proposal or the scope of work that they wrote with a client where it outlines what their service levels are. And you know, the example I used in the workshop is that if you were sending a letter in the post or the mail and you bought, I don't know they have them in the States, but you bought a second class stamp which is like the cheapest stamp you can buy and that's going to take three to five days to get to your client will or to the person you're sending the letter to. Well, you can't expect that to get there the next day. And if you do want it to get the next day, you pay expedited delivery. And of course that costs you a lot more money. And exactly the same should be true in the agency world. And we should never let our client's urgent become our urgent because often that's happened because of their own inefficiency. They're leaving things to the last minute. So getting all of this right in at the beginning in the scope of work and making sure the client, you and your team understand the service levels is definitely going to support you in stopping context switching. And on that note, with this approach, do you feel the way agencies price or create their scopes of works would change with clients or would it be the same as it was.

Yeah. So it's definitely changed for us. So a few things to touch on there. One would be in traditional Kanban methodology there's not really classes of service but some people have added them on. So the concept of an expedite, the concept of a standard or an intangible, we've often found the intangibles are really applied to their. Our own internal work. So we have to balance external work with us actually doing what we need to do to grow and sustain the business as well. But when it comes to scoping, so here's been the greatest like I guess unlock for us is prior because I did, I did a lot of selling. I'm, I'm the founder, I still do the selling. We've been building our product, it's a software product for the past three years and treating as a tech enabled service. And so what would have to not happen is I'm, I'm selling something to somebody. They ask how long it's going to take and I tell them and I'm just making that up half the time right. Of like how long it will actually take because I don't, I'm not going into the project management tool and looking to see what's the current workload looking like right then and there. I'm on the call selling the deal. That's what I tell them on the call. It's what's going to go into the scope of work when I put together the, you know, our sow or something. Right. And as we tracked these metrics there's this, there's what you can do called a Monte Carlo simulation and that's a. It can, it can get overly complicated but there's tools out there that help you, to help you do them. So it's not like you have to. All you have to do is just put the right variables in and it'll calculate it for you. But what a Monte Carlo simulation does is it gives you a, it's a probabilistic forecast of what's likely to be true giving the historical inputs that you give it. So it changed the way we approach scoping work from oh, I think this will take us six weeks to now we can run a probabilistic model that runs over a million simulations looking at how long historical work like this has taken us. And it says you have an 85% chance of completing this within the next 21 days. Right. And so. Or the inverse would be, let's say you already know what the. If you need to get something Done within. We actually just had a project we were doing for large influencer client here in the US who wanted, I think it was something like 100 articles that we were producing for them over a 90 day period. And they weren't just like your regular, these were in depth, very thought leadership, driven from her voice type of things. And they asked, can we hit this? And I said, well let me run the model. And so there's two types of Monte Carlo simulations. One is how long will it take you to do X? What's the probability of that? And the other one is given X amount of time, here's the likelihood you're going to be able to complete X amount of items. And so they completely changed the way that we scope work. And I've actually found that clients like it more because we're not giving them these hard dates. We're telling them and frankly to some extent being able to kind of show them this and say, hey, here we ran this Monte Carlo simulation, they don't really know what that means, but it sounds fancy, right? And we're like, yep, there's a pro, it's a probabilistic model. It shows us here we got an 85 chance of having this done by X date. Is that reasonable to you? Does that, does that fit your, you know, what your needs and they're fine with it and they actually love it. And because it's, to them it feels more real than just giving them a hard, you know, line and we, we hit it 85% of the time, go figure. Right. So it's, that's helped us a lot when it comes to knowing what resources are we going to need in advance, are we going to need more resource resources to apply to this? When can we take on additional work or if we need to take on additional work, do we need to hire more people? Right. And if we can't take on additional work, how do we need to think about pricing this if they want it done sooner, that way we can think about okay, if we're going to have to work extra for this, if we're going to put in overtime, something like that, how does that affect price? And we tell that to the customer and they can then decide is that reasonable or not and make that decision. So they go into it knowing and it's definitely, definitely helped us.

Yeah. So interesting to learn more about the Monte Carlo simulation. How, how does this impact. Well actually one of the questions I was asked, are you still time recording in your agency and should, if people are implementing this approach, do they need to still be recording time or is it now just based on, you know, that, that timeline and that outcome.

So I think it depends on how you operate your business. If you bill based on hours, then you still need to track time. We don't bill based on ours. We bill based on effectively the output, the deliverable. And so, yeah, so, so we don't. I track my time and some other people do, but I do it as a way to understand what's like, how long am I spending on something like what, what, what's like the cost. For me, I really just do it to see like, where's my time being spent. But for the most part, instead of tracking, say like at the granular hour level, we're looking at it. What we're looking at is when did the item start and when did the item go into a queue. And so when we're moving stuff around the column in the back end of our tools, like you can do this with asana, wrike, trello, ClickUp, whatever. When you move them, it'll kind of date stamp it as it moves between the columns or the sections or whatever you're working with. And we use that to calculate, okay, well, how long was this? And that gives a sense. It is somewhat dependent upon people moving it to the actual column whenever they're done with it. But so is depending on somebody to hit a stop start button on a time tracker.

I can't help but feel like you're more likely to get people to do that than record their time. Just for the record, I am a massive fan of not selling time to clients because clients are never ever buying your time. They're buying an outcome, as you said, Skylar. So I'm a big fan of that. However, up until now, I've always said to people, you need to track time internally to check on efficiency and profitability. But actually if you're focusing on these throughputs, then. And you'll get saying, okay, there's an 85% chance that we're going to get this done by this day. And these are the various components. And if everyone's doing that, then I guess you know, you're going to be profitable and you're going to be kind of hitting your targets and people are being efficient with their time because they're hitting this 85% timeline.

Yeah, I mean it's, it's when it comes to say, utilization, because again, we're still looking at it, we're just not looking at it from this perspective of timeless tract because we tried this and if someone has some sort of technique or ability, some sort of way that can just get people to do this correctly without any sort of, you know, resistance, please tell me. And not only that, if you can figure out how to tell me how to do it properly for myself so that I can actually lead from the frontier, please tell me that too, because it's very easy for me to tell everyone else to do it. And then I get to working on, because I still develop a lot of products. So I get deep in the weeds writing code and next thing I know, like, I have no idea how long I've spent on this, you know. And so I don't really set like the best example there. But when we said, okay, well the time tracking is never really 100% accurate. So let me stop worrying about that so much and instead let me just look at rather than hours, I'm looking at it in terms of days. It's how many days did it take for me to complete this? Was this a one day project or a half day project or was this a two, three day project? It's the same thing as tracking your time, except it's a lot easier when the system just says, you've moved it to this column. It does the math. It says, well, okay, there's the time, right? If you want to divide it by three to get a more realistic estimation of hours, sure. But when you try to get hyper obsessive or accurate or precise or perfectionist with how you're tracking time, I think you end up wasting more time than what you would do if you just said, let me just focus on getting the work done.

Yeah, no, it's so true. And I just resonate with so much what you just said there. We've got to lead by example if we want to put these tools in place and demonstrate that we can use them as well. When I ran my agency back in the 1990s, in the years and years ago, we did track our time. We sold time because we didn't know any better. We sold day rates to clients back then and we did. So we had to track our time. But in those days we did it in a spreadsheet because there were no stream times or toggles or harvests or any of those tools available to make your life a lot easier. Just switching this up because I'm conscious of time. Have you got any practical tips that you can give to agency owners if they want to improve their throughput and their flow efficiency?

So this the simplest one, I guess, because again, we've Been learning this as we go. Well, the one big lesson I learned very early on was don't try to reduce your whip too fast. So like, don't try to just overhaul your whole system. Instead, just focus on, take what you've got in progress right now and just tell yourself, I'm going to be a bit more mindful about what we're starting and let's just focus on finishing what we've got in progress. And so one thing we did early on is we said, let's, let's see how much work we have in progress currently. Let's track that, not try to reduce it. Let's just see where we're at. And so all you got to do there is just say, okay, well, how many items are you working on right now? And just calculate them up, put them in spreadsheet, whatever you want to do with them. And then we would look to say, okay, let me finish what I've got and at what rate am I finishing things? So we would just say, okay, did I finish one thing today? Two things a day, Three things a day. And when we, when we're talking about things, what we're talking about here, and this is important, is that a work item? Some people might think of this as a task or deliverable. It's. It's something that provides tangible value to the customer. So this was a big, big shift for us was instead of thinking about, how do we. It's not saying you can't get granular or break things down, it's just they're all just subcomponents. But the actual thing that you're working on, you're completing, is the thing that provides tangible value to the customer. So there's a big shift towards, instead of focusing internally on utilization, which is a very internal focus metric, we focused on throughput, which was getting things done faster for a customer in isolation, a single customer, not all customers. Because when we try to do everything well for everyone, we end up doing nothing well for anyone. Right. And so we would rather say, let me finish this one thing for this one customer that delivers value to them as soon as we possibly can, that makes them happy, that has impacts on their satisfaction. If you're doing any sort of NPSC set scores, etc. And then we'd work on the next thing. And it's very, very difficult. You're going to feel like, I can't not work on these eight things at once. How can I just focus on this one thing? But we tracked it and I said, let's Just see what happens if we do this, if we just focus on one thing. And when we looked at the math, we found out that we were actually getting more things done by working on fewer things at a time. And once we had realized that, try doing it with an internal project or something that may help if it feels like it lowers the stakes a bit. Once you see the numbers, then it's a lot easier to give yourself permission to say, wow, the math actually works here. And there's actually like an entire like mathematical like law that backs this up. Little's law is the law if you ever want to look it up. But so that was a big thing. And then beyond that it was focusing on again identifying how long does it take for us to do a thing and then trying to balance the number of things we're allowing to come in to work on at the same rate that they're leaving. But yeah, simple, simple way of don't try to do it all at once. Just look to see how much you got going on at once and just try to say, I'm just going to do this one thing this day and just take it like that.

Great. What I like about this approach is, is that it aligns how you operate internally much more with what's important to the client. And too many agencies, I talk about four things we deliver. Inputs, outputs, outcomes and impact. And inputs, the time and materials to do something. Outputs are the output of like creating a wireframe or initial graphic or something. The outcome is what the client cares about and then the impact is the longer term results of the outcome. So this starts to align what you do internally a bit more with those outputs and outcomes and a bit less with the inputs. And too many times I see people kind of reporting on tasks that they've done. The client doesn't really care about the tasks that you've done, they care about the outcome. So if we start reporting more on the outcome, we're going to be much more aligned with what the client's looking for. And that's fundamentally why clients aren't buying time. Because they don't care if it takes an hour or 10 hours.

Yep. I mean, it's crazy thought, right? But that just like we see this in the world of like traditional marketing, if you're marketing yourself or marketing your clients is, it's really crazy what tends to happen when you shift your focus for what from what's best for you or what you feel like is best for you to what's best for the, for the customer. And that's where, like throughput, that's them receiving value, right. And we shifted it towards that. And all of a sudden, clients are happier, we're happier. We're actually getting more work done. Right. We get more referrals. You know, they want more work. You know, things like that. So just like in marketing, it's like, you know, you. You can think you know, what customers want, or you can actually just go talk to them and focus on what do they want? How do I deliver that to them without my own, you know, presuppositions or, you know.

Yeah, yeah, absolutely. Yeah. That's a whole other conversation, isn't it, about not making assumptions, but actually talking to clients. Really interesting conversation. Skyler. I'm conscious of time, and I want to make sure I ask you the final question that I ask all my guests when they come on the podcast, which is, if you go back in time and give your younger self just starting out in business a piece of advice, what would it be?

Hire people that you can't afford sooner. I look back on it now, and I think about a lot of times where I felt like I needed to make sure that we had, you know, a safe cushion to be able to hire somebody if we needed someone really, really important. And kind of told myself, oh, we're not ready yet. We'll, you know, I'll pick up the slack myself. And in hindsight, I would have been a lot farther ahead than I am right now if I would have just hired people sooner who knew how to do, you know, do the job and, and not thought about it as a. What happens if we.

Yeah.

If we can't run money. They make you money. That's. That's what happens whenever, whenever you hire. Right.

So an investment, not a cost. A really interesting one. And the sub question to this is, do you think your younger self would actually listen to your advice and heed it and action it?

Oh, no. If we did, we would never. We would listen to our parents. Right. Half the time. So it's. I think it's.

I don't think anyone said yes to that. Anyone said yes to that question. Brilliant. Listen, really great to talk to you. If people wanted to find out more about you and your agency and what we've been talking about today, where is the best place for them to go?

If you want to learn more about myself or the agency, you can look me up on LinkedIn. Scott Reeves, you can go to our website, artangrowth.com if you want to learn more about what we've talked about. I haven't written a lot about that. This is really just me sharing the great thing that we've discovered and helped us a lot with other folks. You can just search for anything related to Lean Management or read the Goal by, I think it's last name's Gold Rat. He's got a book called the Goal. It's a, it's a really good sort of narrative driven story, but you can search for it on YouTube, just learn anything about Leaner flow. And yeah, that's probably the best place to find. It's where I found it, so.

Oops. Great. Okay, well, I'll include those links into the show notes. And just a big thank you to coming on the podcast today and sharing some sort of interesting insights on how we can sort of run our agency differently and also what an impact it's had on your agency as well.

Thanks for having me, Rob. Sam.