How Agencies Should Think About and Manage Risk

Sei-Wook Kim (00:06.104) On today's episode, we're breaking down six levers that drive profitability across all client engagements, whether you run fixed fee projects, retainers, or long-term service programs. We'll also fold in core principles around gross margins, utilization, forecasting, and delivery discipline to show how they all reinforce each other. Peter Kang (00:27.182) All right, so this topic is very top of mind for us these days. In our Barrel Holdings portfolio, we've had a pretty good year in the aggregate. But one of our portfolio companies, an agency, is going through some challenges with profitability. We've been thinking a lot about how they can correct course. So this is an opportune time to dive into some of the fundamentals and see how we're thinking about it, and also share some of the things that we're already doing or plan to do to make sure the agency gets back on track. So maybe we could kick it off by talking about a very important concept: gross margin. Maybe we can dive into that. Sei-Wook Kim (01:22.423) Yeah, definitely. So when we talk about margin today, it's about gross margin. Gross margin is the revenue that you receive as a company minus the cost of delivery — so the labor to service that revenue. And really that metric is an indicator of how efficiently you're doing the work and getting the work done. Peter Kang (01:45.579) Yeah, so in many respects, people talk about profit and this is one form of profit, right? It's a very important form of profit. Cool. So when we think about where gross margin gets compressed or where it leaks, what do you think the three sources we typically come across are? Sei-Wook Kim (02:10.624) Yeah, it usually starts at the pricing and scoping stage. So if there's misalignment there right at the beginning of signing a new engagement, that's where a lot of — it's the saying like you're setting up the team to fail from the get-go. The next part is poor utilization or overstaffing. So if you have way too many team members for the amount of work that you have. And then the third is inefficient delivery. So you may have scoped things, you may have the right amount of people, but if you're not delivering consistently — let's say your projects are going way over timeline from what you expected — it's all about the delivery there. Peter Kang (02:57.099) Yeah, like how you do the projects or the engagements, right? Cool. So we'll dive into each one of those. We've split those into actually six separate levers. And I think we should note that this is profitability not just on fixed fee projects, but it applies to retainers, any hybrid models. You can apply these principles across the board because at the end of the day, you're talking about the work that you need to do for your clients, the people involved, the promises you're making, and then how you're going to deliver it. So we think that these are pretty transferable. Let's dive in. Let's start with lever number one. You want to kick it off? Sei-Wook Kim (03:42.964) Yeah, so lever number one is pricing and scoping. Like I just mentioned, it's really aligning what you are delivering — or promising that you'll deliver — to the price that you're charging for it. In general, the more you can charge for a project, it gives you more buffer. More buffer for expenses, if things go sideways, if there are delays. And if the team makes mistakes along the way, you just have a little bit more buffer. So the more you can charge for a project helps with the margins from the get-go. Peter Kang (04:21.247) Yeah, but that's a very loaded topic, right? Because what's stopping anyone from saying, if you want more profitability, just jack up your prices, charge your client more — why is that not always the solution to this? Sei-Wook Kim (04:34.888) Yeah. It's funny because one of our earlier interns said, hey, let's solve this problem — why don't you just increase your prices? And there's definitely truth to that. But what you're missing there is, if you increase your prices, clients may just go somewhere else where the prices are lower. They'll say, I don't want to pay what you're asking for. You may end up in a situation where you lose a lot of projects because you're trying to ask for more and more money. Peter Kang (05:11.141) Yeah, and this is a topic we'll definitely cover in a later episode, but there's so much that goes into pricing. Part of it is how you're positioned, part of it is the psychology of how you're talking about the value you're bringing to the client, and then just what the market conditions are like to support the pricing that you put forward. But all that aside, higher pricing is better — it gives you a buffer. And then how does that relate to the scoping side? Sei-Wook Kim (05:41.043) Yeah, so basically the pricing is the pricing, but really you need to match the scope to the fee that you're charging and be clear about what it is that you're promising to do. We'll touch more on this in some of the other levers. But if you don't have alignment from the get-go, that's a quick way to lose margin when it's not clear what you're delivering. Peter Kang (06:07.21) Yeah, and one thing — maybe a good example of how we think about it across our agencies — is if there's a price that you've had the conversation with the client about, let's say they're going to buy a web project or a six-month retainer for $100k. A good way to start is actually to think about your desired gross margin on it first. You might think, okay, let's target 70%. So we want the cost of delivery to be 30%, or $30k, and we get $70k of gross margin. The way to think about scoping is, hey, can we get this work done for $30k of cost? I think that's a helpful approach to really structure the thinking around scope. And that's what we mean by matching the scope to the fee — depending on the agency and your situation, the gross margins might be less or more, so you have to apply it to your situation. But generally that's what we're talking about. Sei-Wook Kim (07:24.456) Yeah, and an important thing there is typically you're pricing based on the outcome of the work, not necessarily like you're going to buy X amount of hours that translates to $100k. So with that lens, you can really be creative on how we design the project to achieve that outcome. Do we need to use the exact same process we did on another project or retainer where the budget was double? Do we need to use the same type of resources? Maybe you can change the combination of senior and junior, or potentially some offshore resources, to help achieve the same outcome with the budget that you have. Peter Kang (08:06.675) Yeah, speaking of which, that is lever number two: staffing and utilization. Let's dig into that and how it impacts profitability. Sei-Wook Kim (08:16.084) This can also go more macro — just thinking about your whole team, all the roles, and what is their billable expectation. A senior person who oversees a team member may have a lower billable expectation versus a more junior team member that's fully dedicated to client work. So at the get-go, just understand that each role is slightly different, and then track it consistently. Doing utilization reports and actually taking a look at the numbers to see where people's time is going — because when you look at overall gross margin at a company, idle hours where someone has zero availability really crushes your margin really quickly. Peter Kang (08:56.362) Yeah, across our agencies we prefer to track time. But for those that might push back — hey, we as an agency absolutely don't track time and have no plans to in the future — what would you say to those folks in terms of tracking utilization? I think it's still possible. How do we go about that? Sei-Wook Kim (09:22.407) Yeah, the other way to look at it is the planned or expected utilization of a person. That's the exercise of going through and forecasting what everybody should be working on and updating that on an ongoing basis — so if a project gets delayed, you increase the expected amount of time or percentage of allocation. However you prefer to do it, saying, all right, maybe this person is working 25% on this project, 50% on that project, and having some semblance of how the time is being spent. The person doesn't need to track time, but you know through another measure where their time and allocation is going. Peter Kang (10:08.104) Yeah, basically some sense of how people are spending time is better than nothing. Great. And then just going back to what you said about the type of people on a project — seniority levels especially — want to expand on that? Sei-Wook Kim (10:28.674) Yeah, a couple things there. You may want to have senior people working on a client account, but obviously they're the most expensive resources on your team. And if you have only junior people working on an account, you may realize there's going to be a bunch of rework — there are things a senior person would take 10 hours on that a junior person would take 40 hours on. Suddenly the margin is lost. So it's really a balance of how do you support junior resources with senior resources and make sure it's the right alignment so that a project goes efficiently and you're not reworking a lot of the engagement. The other part of the mix is thinking about full-time versus contract or part-time resources, especially when work comes in peaks and valleys. You might have a really busy season or a really busy period, or one of your clients needs help for a short period of time. Using contractors to scale up versus full-time resources means that at minimum you can utilize your full-time resources 100% and then use the contractors to flex up and down as needed. Peter Kang (11:38.12) Yeah, we've experienced this ourselves and we hear it so many times — as an agency, you might be super profitable on a retainer or project basis. But then as a whole, let's say you did a project in perfect time, maybe even a bit early. You did everything you needed to do, huge margins if you measure it on that project basis. But then the week or two weeks after it wrapped up, you had a bunch of people who weren't working on anything and you're still paying them a full-time salary. When you look back a month later, you're like, wow, that wasn't that profitable at all. That's where the utilization concept really comes into focus — you have to keep that utilization high. The freelancers allow you to flex down immediately after a project is done so that you're not carrying those costs when people aren't busy. Sei-Wook Kim (12:45.69) Sure. Cool. Peter Kang (12:46.951) Okay, lever three. This one is engagement design and approach — how projects are done or engagements are handled. One of the profitability killers is when you don't have a good plan, don't have a good process structure for how you're going to get it done. There's a lot of unnecessary meetings, a lot of back and forth, and that eats time up. This is not too uncommon for agencies doing something for the first time. Like, we've never done this type of project before, but it looks exciting — dive in, figure out how to do it, build the plane while you're flying it. There's a lot of waste that happens in those situations, which sometimes is okay, it's an investment, and maybe you learn from it for a similar project later. This is where engagement design and approach is super important. Maybe we can speak to some ways to be smart about engagement design and approach. Sei-Wook Kim (13:57.112) Yeah, where possible, don't reinvent the wheel for things that are the same every single time — like how you onboard your project team, how you onboard your client, how you think about QA and making sure you're allocating time for it. One of the big things in the design of an engagement is creating the project schedule all the way through and really thinking about how you're going to get that engagement done, including what are the meetings. Do you need a meeting, or could it just be a post, an email, some kind of async communication? Do you need client feedback? Have you budgeted for that time? And if that's downtime for your team, could your team be working on something else? Just thinking about all of the nitty-gritty aspects of a project from start to finish and accounting for as much as possible — because then there's going to be some variation, but there is some predictability to how the project is going to get done and how you want it to get done. Peter Kang (15:05.181) Yeah, 100%. And I can't not talk about this because it's such a pet peeve of mine. The project schedule or project plan — a lot of times, and I don't know what it is, the thing I would hate to see is when project managers, account managers, whoever is creating it, they're so focused on it being perfect and super detailed that it just never gets done. And in some ways it's an incomplete picture. To that I say, look, what's more important is just first getting a basic sense of the shape of a project from start to finish. You don't have to have it perfect. Don't let perfect be the enemy of good. Fill it in so that everyone has a sense of, hey, over the next four or five months, these are the things that are going to happen, this is the launch date — and then work backwards. What happens a lot of times is people are trying to meticulously plan every single week, getting more and more inputs. Because of that, it's an open-ended launch date, and by the time they finish adding everything, you realize, wow, this is two months longer than what we'd talked to the client about. So the big thing is start with the end in mind, don't sweat the details until you've gotten the structure and the container done, and then from there you've got some constraints to work with to solve the problem. So that's my rant on projects. Sei-Wook Kim (16:41.635) Yeah. It's almost like: start with the end, work backwards, what are the big milestones, and then how do you get to those big milestones? The analogy of building the plane as you're flying it — there are some things you'll need to figure out and the next week might be the most clear. But as long as you know, hey, we've got a big deadline in two weeks, how do we get there and make sure we're hitting those major milestones — that's what's important. Peter Kang (17:14.627) Yeah, definitely. Anything else on the design and approach side of engagements for improving profitability? Sei-Wook Kim (17:23.431) Yeah, the other thing is just thinking about what does our team need to do versus what does the client's team need to do in an engagement. Sometimes if that's not clear, you may just end up taking on work you didn't need to — maybe the client could have gone in and populated a bunch of content on the site, or gone through and done some kind of UAT testing that wasn't clear upfront. So just think through that it's not one-sided when you're working through an engagement. Peter Kang (17:55.938) Yeah, definitely. Okay, let's move on to lever four: scope management and change orders. Let's talk about this because it has to do with scope creep, which in many ways is one of the biggest threats to profitability. Sei-Wook Kim (18:11.311) Yeah. So we talked about scoping — that was the first lever. Scope management honestly starts there at the business stage when you create an SOW and agree on a scope. The more you can be specific about deliverables — rounds of revisions, who does what in the user acceptance phase — if all of that isn't outlined, it gives your team very little to push back on. It's very open-ended. But as long as those aspects are clearly documented, you know exactly what's being done. And if something is clearly out of scope, having a formal process to document how to handle those situations is really important. Peter Kang (18:57.476) Yeah, vague scopes are one of the biggest liabilities when it comes to profitability because you can get taken for a ride as an agency if you're super vague about what you're actually delivering. 100% there. But let's say you've written a tight scope and the client is definitely asking for more. What are some ways to handle that? Sei-Wook Kim (19:23.982) Yeah, it's sometimes an art to figure out how to navigate that. Let's say a request comes in. You can say, hey, let's do this in the second phase — we'll defer this to the future. That might be an acceptable solution for now. Or there can be a trade-off: this thing that you wanted, maybe that's not as important, and this new request that's come in should take priority. So why don't we trade those? We'll do this new thing and won't do the other thing — no change in budget, but a change in scope. And then the third option is obviously just scope it out, ask for more money, and get more budget for the change that's about to happen. Peter Kang (20:01.988) Yeah, and like you said, there is an art because you have to put yourself in the position of being a buyer of anything, right? We expect to pay a certain price and maybe some things weren't clear, so we're just asking for this thing we thought was part of it. For us to immediately get hit with a change order won't feel good. Sei-Wook Kim (20:30.549) Yeah. Peter Kang (20:31.351) But if there's a conversation around it and maybe some understanding, you're going to feel better. So it is definitely an art. You don't want to get into a situation where the client says, stop nickeling and diming me — because then you're just jeopardizing the health of that relationship in the long run. Sei-Wook Kim (20:50.093) The other thing — an exercise along with change management — is, let's say you are doing that trade-off, doing a zero dollar change order. Documenting it, signing the scope, getting it in writing, and making sure the client understands what's happening. It's like signing a new contract, but it's just a zero dollar agreement. Peter Kang (21:11.658) Yeah, and that goes back to documentation because over the course of a project, you could have dozens, if not hundreds, of these kinds of scope conversations. If you don't have a paper trail, people's memories get very hazy over time and they might come back and be like, hey, I thought you said you would do this. If you can't point to something, those are tough conversations to have later on. Sei-Wook Kim (21:35.21) Yeah, definitely. Peter Kang (21:37.154) Cool. Anything on the retainer front, actually — I know that was very project-focused — but when it comes to scope and change orders on the retainer side, anything there? Sei-Wook Kim (21:47.21) Yeah, with retainers it somewhat depends on how you've structured it, but typically you want to go through a prioritization exercise to understand what we're going to do with the team that we've allocated. It's the same thing — when we think about a retainer, it's like a series of small projects that make up a retainer. If you continue to agree to things because maybe there isn't an explicit scope, that'll balloon the team that was supposed to work on it. Suddenly, if they were supposed to balance another account, maybe they're not able to do that. So it's the same exercise: really agreeing on what the priorities are and documenting that clearly. Peter Kang (22:35.654) Yeah, maybe the other element is carving out some time or resources for ad hoc requests, because sometimes retainers are a catchall for some of these things. If you know there's going to be a certain amount of ad hoc stuff that comes up — clients are going to ping you and say, hey, can you take care of this — and you want the retainer to service them in that fashion, that's something to pre-plan for. And sometimes certain requests are just too big to do comfortably within a retainer. It's even better to just say, hey, let's treat this one as a one-off project and preserve the retainer for the more ongoing things. Sei-Wook Kim (23:18.495) Yeah, definitely. Peter Kang (23:21.941) All right, lever five. We have reusable components and IP. Let's dive into that one. Sei-Wook Kim (23:28.108) Yeah, so this is similar to what we were talking about with how you execute a project. Oftentimes, the more you do similar engagements, you internally develop standardized templates on documentation. From a coding standpoint, there might be modules and scripts you've written that you could use again. Just thinking about what are those reusable components in a project that could reduce delivery cost. It could be a reporting framework — if you're doing the same kind of report every single time, having the same structure, the same PowerPoint or Keynote template that the team uses, where every subsequent team benefits from it. All of that can definitely help reduce the time and cost of delivery. Peter Kang (24:28.72) Yeah, absolutely. And once you have these reusable components and documentation on how to use them, you can then assign lower-cost, more junior team members, or rely on offshore or nearshore resources to handle more of these tasks, freeing up time for senior folks to do higher-value work. It creates a really nice sense of leverage. This is where specialization really plays out — over time, you have all of these components and IP, and you're able to come to the table in an engagement and say, hey, this is how we do it, these are all the tools and internal workflows we have in place, and we're able to do it super efficiently for you. Maybe you can charge a lot because the certainty and quality of it is great, and at the same time the cost to do it is low and the margins are super high. This is a huge part of why we beat the drum on specialization. Sei-Wook Kim (25:44.084) Yeah. One thing here is you do need to be intentional about setting aside time, because oftentimes it's internal or unbillable time being spent to organize all the components. All those little pockets of time — in between a project, or downtime within a project or engagement where the team has the ability to say, all right, here's everything we did, let's package it up cleanly so that other team members can benefit from it. Peter Kang (26:14.356) Yeah, absolutely. The more habitual you can make it — just build in some time right after the close of a project — the more likely these things are going to get built out. All right, the last lever. We saved the one that everyone's talking about for last: automations and AI. This is something that's top of mind for a lot of folks. In the context of engagement profitability, how does this lever factor in? Sei-Wook Kim (26:51.593) Yeah, this is all in service of how could you be more efficient at executing a project. Automations can help with a lot of the administrative tasks — lower-level tasks that happen in a project. As one example, think of an automation where a contract gets signed in DocuSign. Can you trigger something in HubSpot, or the other way — HubSpot gets updated as the contract is signed — and then it triggers on the billing side saying, now you can send the invoice, and at the same time activates your projects, so Google Drive gets created, Harvest gets set up. There are all these tools where typically someone would manually go through and create these things. If there's an automation that handles all these steps that are exactly the same in every single engagement, that can save hours of time for an account manager, for example. Peter Kang (27:48.77) Yeah, and even enriching it with AI so that as these workflows are happening, there's a brief for the engagement or some kind of research that gets automatically pre-populated and is ready to onboard the team. It might draw from call transcripts from your video calls, or go out and hit all the client sites, look at competitors, and tie it all together neatly — that could save hours. And that's something that is very real. People are experimenting with it, and we're certainly using it in a big way. Sei-Wook Kim (28:31.048) Yeah, definitely. And that's more of the administrative side. But even on the execution of the work itself — in development, AI has helped speed up a lot of tasks when you can prompt to get scripts or things off the ground. Even on the QA testing side, automatically generating tickets for the team. On the design side, there could be time spent on initial copy that you have to write for a website, but if you prompt AI, the copy that comes out could be pretty good or close to production-ready. So there are a lot of aspects of the journey that can be supplemented by AI. Peter Kang (29:23.986) Yeah, supplemented or even taken up as a whole, right? Like, you can put in prompts — hey, I'm looking to get a website made that has these qualities — and sometimes it could one-shot it to a 70%-good-enough site that can then be polished up and tweaked and delivered to clients. There's a lot of stuff that on a weekly and monthly basis keeps eating up more of what was previously manual. At the same time, I don't think that necessarily makes it any easier to some degree, because you still have to manage it, you still have clients, you have to make sure they understand what you're bringing to the table and get their buy-in on certain things. There's a whole aspect to engagements that AI is not going to completely take over. As long as it's people working with people, there's going to be an element of management there. But I think as an agency leader, you'd be foolish not to embrace these tools, because at the end of the day a vast majority of folks are leveraging them in some capacity. Peter Kang (30:42.909) Cool. All right, let's bring it all together. We are the Agency Habits podcast, so maybe we can talk about some habits that can reinforce all these levers in a repeatable way. Let's quickly list off these habits. Sei-Wook Kim (31:02.309) Yeah, first habit: weekly budget tracking. Every project or engagement should have a budget that is looked at and monitored weekly. Oftentimes this is the project manager, account manager, whoever fills that role at the company. What we've found is it's almost helpful to have some shared accountability — do it as a team, look at your budget together, see how your resources are tracking against estimated actual hours versus estimated, and what opportunities do we have to course correct before we go too far. Peter Kang (31:38.333) Yep. Next one is forecasting and capacity planning. Making sure you're looking ahead and forecasting your team's hours, and then matching the pipeline with the available capacity. If you have an influx of work coming in, you're going to have to size the team up. If there's less work coming in, you might want to size the team down. And we talked about keeping a freelancer or contractor bench to flex up or down as needed without carrying ongoing costs. Sei-Wook Kim (32:13.275) Yeah, and for that one, it could involve team leads, business dev to inform what's coming down the pipeline, and all the project managers and account managers who are working on live projects. The third habit is tight change control. We talked a lot about change management, but having a formal process where all changes are documented and avoiding doing free work along the way really helps protect your profitability. Getting the team involved to flag it — it could be a designer who's in a call with a client who brings up a new request. Making sure there's a process for that person to talk to the right people and have a conversation around a change request. Peter Kang (33:02.778) Yeah, this is all about going back to scope management and making sure we're handling that. Cool. And then the last habit is debriefs and feedback loops. We talk about this a lot — you have to learn and continue to make improvements from the learnings. At the end of every engagement, you want to have a structured debrief. Get your team together, always have folks share what went well, what could have gone better, and learn from those opportunities. Double down on things that went well, and use the ones where you didn't do such a good job to create new processes and improve the way you do things next time. This is a great opportunity to dive into where the margin got lost and how you could have preserved margin better. And once you have that debrief, it's very important to share it with the rest of the team so that others can learn from your experience. Okay, great. That brings it all together — six levers as a framework to think about ways to clean things up on the profitability side, and these habits are ways to keep it in motion. Anything else before we wrap up? Sei-Wook Kim (34:19.694) No, I think it's a whole team effort, as you've probably seen with gross margins and lifting profitability across the company. It's not just about one project — it's about a holistic view at all these different factors. Everyone has to come together and work together to achieve this goal. Peter Kang (34:38.938) Awesome, well, thanks for tuning in and good luck making all your engagements profitable. Take care. Sei-Wook Kim (34:44.783) Thanks.