Rob Harr, Vice President, Sparkbox
Rob Harr can see into the future. He knows what’s coming in, what’s going out, who’s working on what and for how many hours. What kind of sorcery is this, you ask?
Over the past nine years at Sparkbox, this scrum master-turned-Nostradamus has engineered a finely tuned Ops system built on Friday forecasts, micro-cash sheets, rolling six-month billable averages and two years of feeling really uncomfortable.
Tune in to hear the method behind his magic and learn how to uncover truth in your own data sets, repetitive grading and a fresh mindset.
Carl Smith: Hey everybody, and welcome back to the Bureau Briefing. It's Carl. Before I get started, I want to say something. We've suddenly had more listeners than we used to, and I make jokes all the time about the seven people listening. There's a lot of you out there now, so thank you, and I'll stop making that joke. Just, you keep listening.
With an episode like we have today, I think you're going to wanna because today we're welcoming back to the show, the co-founder and vice president of Sparkbox, Mr. Rob Harr. How you doing Rob?
Rob Harr: I'm good Carl. How are you?
Carl Smith: I'm good. Did you not hear Rob? More people are listening. I'm fantastic.
Rob Harr: I'm, we're up to like eight, right?
Carl Smith: It could be double digits. I don't want to over commit. You know I want to over commit so bad. So, Rob, we're friends. Been to a lot of Bureau events together, been to non-Bureau events together. Hung out, all this kind of stuff, and the main reason I wanted to get you back on the show was to educate myself and hopefully make me be a better facilitator at the Bureau because I am totally confuse on digital agency operations. Everybody I talk to seems to have a different definition. Do you have that same experience?
Rob Harr: Yeah. There seems to be a lot of different ideas on what it is, how important it is, and who should care about it.
Carl Smith: Right. So, tell me how you got into operations?
Rob Harr: I kind of got into operations by default. We were running our studio here for a number of years and there really wasn't anyone who had dug in to the business side of things, the, how all the things worked together to actually make money. Well, Carl, we weren't making money.
Carl Smith: And that's changed.
Rob Harr: Yeah. I kind of decided that if somebody needed to do it, and that was going to be me, and I kind of threw myself into it and said I'm going to go figure all this out. I'm going to figure out how the pricing works, the business model works and some kind of way to make sure that we're making money.
Carl Smith: So you started to kind of build a little definition there.
Rob Harr: Yeah.
Carl Smith: Of operations. I think that's one of the things that I've noticed, is when I read about operations, I'm reading about traditional companies. Even service oriented companies, but they don't necessarily have some of the challenges that digital agencies have. But when I started looking at a definition for myself, it boiled down to everything that happens to keep your company running and making money.
Rob Harr: Yeah, I think the most simplistic definition of operations at a digital agency is turning sales into profit. Everything that happens after the sale is done, and shipping the project. That's really everything it encompasses and that's a lot.
Carl Smith: It's a bit of a junk drawer, right? I mean everything that isn't part of business development ... I mean some of it is part of business development, but let's say that isn't part of new business or sales, and even then some of it does slip over, right?
Rob Harr: Yeah.
Carl Smith: Or necessarily the pure financial aspect of the numbers. You're going to look at the numbers. You need the numbers. They help you, but you're not necessarily doing the books. Right?
Rob Harr: No. No, no.
Carl Smith: Everything in between those two becomes operations.
Rob Harr: It does. I think that even how the books get done and what reports get generated out of that are really important to the Ops people. It's the how of the business. I know there's all this talk at the different Bureau events over why we want to inspire people. The why we're doing all of this, and that's all really important, and I'm really lucky that I have a business partner who is amazing at the why, but I kind of own the how. How do we get this stuff done and make money at it?
Carl Smith: When we look at that, when we look at the how, what would you say are the biggest challenges to determining the how?
Rob Harr: I think that one of the biggest challenges, and we have in operations, is being able to predict the future accurately.
Carl Smith: Right. It's forecasting. It's resource planning, right?
Rob Harr: Yeah.
Carl Smith: Knowing what's coming down the pike and being ready for it.
Rob Harr: Yeah. It's forecasting. I hate the word resources because I think it comes from an old mentality of like machines and things that you use up. I like staffing and people's, humans way better.
Carl Smith: We'll go with humans and human planning. So how do you overcome that? First of all, before that, what are the special challenges a digital agency faces that other companies might not when it comes to trying to predict the future?
Rob Harr: I think one of the things that is really hard is that estimating our work is really difficult. I think that estimates are really hard because our clients are, they have a tough time explaining what they need instead of what they want. We're optimistic and we give rosy colored estimates. Most tell us the time, we're not being flexible and letting the business needs that change over time flow into our process. So I think one of the hardest things is that we're just bad estimators of our work.
Carl Smith: Right. So that estimate's obviously not just financial. It's about the effort that's going to be required.
Rob Harr: It's exactly the effort and that makes its way into our financial stuff because almost every studio you talk to that gets started, starts out with a fixed price kind of model. It becomes part of their way of life. Try to figure out what the client wants, try to estimate that poorly, be wrong, work too many hours, but still charge the same price.
Carl Smith: Right.
Rob Harr: I don't really ...
Carl Smith: That's the beauty of fixed.
Rob Harr: Yeah. Yeah. We don't want to like ... This isn't a pricing talk, but you can't get away from it. I think that the biggest challenge really of getting really good at forecasting is not being able to accurately say what happened yesterday. It's actually a reporting problem to start with.
Carl Smith: Okay. So how have you addressed it at Sparkbox?
Rob Harr: I think step one to fix forecasting is, to have really good data on what happened yesterday. You have zero chance of being able to predict what's going to happen tomorrow, next week, next month, unless you can accurately say this is what happened yesterday. Yes we all agree this is what happened yesterday and I have metrics to say this is what's happening, because then you can use that to start laying out in the future and grading yourself and seeing how you're actually doing.
I think that's the most important part. Now, I think one of the things that, if your business model has a lot to do with it and everything else, and what you expect and having clear expectations on the humans involved, but I think it all has to start with how good is your reporting?
Carl Smith: Right, because if you don't have good information, then you're already done.
Rob Harr: Yeah.
Carl Smith: This is part of what gets me, is that there's so much focus, and rightfully so. It's important. On optimizing a process, on figuring out the metrics for how you're going to measure if something was successful, right? There's so much going on out there, but if you don't know at some level what to be prepared for, all that stuff starts to fall apart, right?
Rob Harr: Yeah. I mean it's ... You have to have the ability to actually make adjustments.
Carl Smith: Right.
Rob Harr: If you can't tell what happened, then you can't see if your forecasts were correct, which means you can't make adjustments and get better at it.
Carl Smith: So talk about those adjustments. So say that ... Like, you right now, and tell me if you don't but I get the sense, which is why I wanted to talk with you about it, that you've got a pretty good handle on this.
Rob Harr: Yeah.
Carl Smith: Do you think that's fair, yeah?
Rob Harr: Getting better. We're always working on it.
Carl Smith: Exactly. That's why you're where you are, is because you've been working on it. How do you, and this is another one. People want to find the new tool, but if you don't know what it is you're going to use that tool for or how it's going to help you, but how are you looking at what happened yesterday, last week, last month, however you do it?
Rob Harr: Yeah. One quick comment and then I'll answer your question. The tool conversation drives me crazy. I think that we're all obsessed with tools and really what we need is some good fundamentals and figure out what data we need, and then find a tool to match, rather than to find a tool because we have a broken process. The tools are not a fix for what ails most of us.
But, so we've actually gone through several iterations of this now, where we're in a place now where all of the project managers are forecasting for everything they can see into the future, as far as they can see, based on the projects they're managing, and actually the forecast tool by Harvest. The tool is not important, but they're basically just putting people and assigning them based on their project plan, out as far as we can see. We've got projects right now that run through the end of the year. We've got a couple projects that are only a couple months out, but everything is updated every single Friday before they leave so that on Friday, we have the best information available to us on Monday morning.
Carl Smith: Okay. How far out can they normally see? Like what would you say is visibility in a standard time?
Rob Harr: It depends on the project size. We have some projects that having visibility six to nine months out isn't uncommon at all.
Carl Smith: Wow.
Rob Harr: We have other projects that are, you know ... It's a little bit ... They're much shorter. They're maybe a month long and this is the end of the project. It's only signed work that goes into forecasts for us.
Carl Smith: Right. Because that's the only stuff you know for sure. And even that can change, but you've at least got an agreement, so if it does change, there's some sort of repercussion.
Rob Harr: That's exactly right.
Carl Smith: The digital PMs are the front lines, right, who've got the best view. They put all this information in there by the end of the day Friday, before they leave on Friday, and then it's there for review on Monday. Who's looking at it?
Rob Harr: Yeah. There's several people. One, I think one of the most important things we did when we started using and forecasting all of this was to open it up to the entire team.
Carl Smith: Okay.
Rob Harr: So that everyone can see what their expectation is. We're actually working on a tool right now that will go through forecasts and put everyone's expectations for them into slack. So it will say, hey, welcome John. Here's your expectations for the week, billing this many hours on these different projects.
Carl Smith: Right.
Rob Harr: That's, the team member has access to all of that, so they know their expectations. If you don't clearly communicate what someone's expectations are, you can't hold them accountable to hitting it.
Carl Smith: That's definitely true. Then, just for people listening, so Sparkbox is a Time and Materials shop.
Rob Harr: Yes.
Carl Smith: You can say, is that what it say? It says this many hours on these projects?
Rob Harr: Yeah.
Carl Smith: Or is there tasks associated with it as well?
Rob Harr: Nope. There's no tasks. That's way too granular. I look at ...
Carl Smith: Okay.
Rob Harr: I look at the forecasting piece as an operations tool. It's not a project management tool.
Carl Smith: Okay.
Rob Harr: It has some of the same type of data, but task lists and cards and how we run sprints and all of that stuff is somewhere else.
Carl Smith: Okay.
Rob Harr: This is merely to get the staffing right and to get the forecasting right and get good at predicting what's going to happen.
Carl Smith: Again, not to make this a pricing discussion, how does somebody who's not tracking time or is not utilizing the time they track in their reports, so value-based or fixed bid? I guess retainers, you're still looking at that time, but how would a value-based or a fixed bid company do that?
Rob Harr: I don't think that, I think everybody should be tracking their time.
Carl Smith: Right.
Rob Harr: It's the universal constant. We only have so much of it. I think that, I think it's crazy to say that people who do value-based pricing don't track time. I know that, our good buddy Dan Moll, tracks all of his time. There is not a bigger proponent of value-based pricing out there.
Carl Smith: No.
Rob Harr: Because he wants ...
Carl Smith: Well that's how he ... Yeah, that's how he knows what he did.
Rob Harr: Oh, right because he has accurate ... He can accurately predict what happened yesterday, or report on what happened yesterday to predict better next time.
Carl Smith: Right.
Rob Harr: It's the same principle.
Carl Smith: When did you open up the reporting ... So forecasts, the tool from Harvest has been around maybe three years, four years? Has it been longer than that?
Rob Harr: Three or four years. I think we adopted it very, very early.
Carl Smith: Okay. I know when it first came out it had some issues, but my understanding is it's gotten tremendously better.
Rob Harr: Yeah.
Carl Smith: So then, when did you make that decision to open it up, where everybody was going to be seeing it?
Rob Harr: We ran that way for probably about six months before we opened it up to everybody. We wanted to see how good we were at it. We wanted to make sure that we had time to make improvements, figure out what we were going to track, what we were going to forecast, how we were going to do all this, and make improvements from a PM operations team before we decided to subject everyone to that data.
Carl Smith: Okay. Then you do it. What kind of questions did you get from the team when they started seeing that data?
Rob Harr: Well, a couple things happened. One of the questions was, is like, oh I'm blocked. There's no way I'm going to make this number. It was like, oh really? That's really good to know because if the project manager is the one that's saying you should work this many hours this week on this project, they're also the one that should be responsible for unblocking you.
Carl Smith: Gotcha.
Rob Harr: One of the other questions was, well I have other things that I'm responsible for too. So we track that kind of stuff in Forecast now. Anything that we're asking an employee to commit to and do gets forecasted.
Carl Smith: Gotcha. Is that, what about with your time with operations?
Rob Harr: Yeah. We mostly worry about focusing on forecasting ...
Carl Smith: Things that are going to be building the tool or the project.
Rob Harr: Yes.
Carl Smith: Yeah, yeah.
Rob Harr: Billable stuff or things that impact people who are going to be billable.
Carl Smith: Right. Okay. Which makes perfect sense. Because any data in there that's not that data is just irrelevant.
Rob Harr: Yeah, and the other thing that it's made really, really visible to us from a forecasting point of view is when people are going to be out or at conferences or on vacation. The PM can look at it and be like, oh crap, I'm not going to get the hours that I thought I was going to get.
Carl Smith: Right, and also I mean, somebody goes home sick or has a family emergency or something else comes up. Because you're dealing with it ... Well, I mean you're dealing with it a week at a time, but an unexpected absence creates a bit of a gap in there that you can't just plug another human in.
Rob Harr: It can. One of the things that we run from an expectation point of view is all of our billable employees have a billable amount they're suppose to average over a rolling six month period. If you're unexpectedly out, that's fine. Please, go take care of your family, or whatever you need to go do, but you're responsible for that average. It's going to ebb and flow a little bit. If you can't make the hours up today and it's not PTO, there's next week or the week after.
Carl Smith: Okay. How does that work? The reason I was asking was how does that work for the project? Does the project just kind of slow down until they're back?
Rob Harr: Yeah. It does. It can. That's okay for the most part because unless there's a tight deadline on the end, since everything is time and materials, then we just invoice less that week.
Carl Smith: Right.
Rob Harr: The client isn't charged the same amount. Invoicing goes up and down. It fluctuates a little bit, as long as it all comes out in a wash, and it almost always does.
Carl Smith: When you're going through and talking with clients, do you explain any of this with them, about the way that the resourcing, or sorry the human process works?
Rob Harr: The staffing.
Carl Smith: Staffing.
Rob Harr: We do. We definitely talk about it. We talk about we invoice actuals and we invoice every single Monday for the work done the previous week.
Carl Smith: Right.
Rob Harr: There's not a whole lot of reason to explain to them how many, who's going to be off when and why. We work on cards. We have a velocity based on story points. They can basically see their invoice, so it's here's how many hours went in to, to close this many cards, this many points worth of tasks. That stays pretty consistent because even if I take, somebody takes a Friday off, there was less hours for less story points. It kind of corrects for itself.
Carl Smith: You're using story points, you're talking through some different methodologies. Did you kind of just put together what was going to work for Sparkbox in kind of a pure play mode with certain philosophies?
Rob Harr: I think we've kind of iterated. We've been doing this for nine years, and we've kind of figured out the things that work. We tried to estimated all of our tasks and hours and it got really, really messy. We've kind of moved to more of a story point kind of thing. We definitely work in an iterative way, but there's no specific development methodology we use unless a client requires it.
Carl Smith: Right.
Rob Harr: We've gotten [inaudible 00:19:55] we worked on things like that, and we're happy to fall into those processes, but we try to stay away from things that add too much ceremony if we get a pick.
Carl Smith: Did you have places you went to, resources that you used to learn about these different opportunities to iterate?
Rob Harr: Yeah. That's a good question. I was a certified scrum master in a former life. I kind of had that baggage with me I guess. Our technical director was a scrum master trainer, actually. We've send some of our PMs to scrum training. I think the biggest influence for me personally, from an operations point of view is there was a CFO type that really early we brought in because nobody knew anything about any of the financial pieces here.
He used to come in and spend a couple of hours with me every single week. On Thursday mornings he would show up at 9:00 and he would be here until about noon. We did that for I think a couple years, and he would just beat me up every single week. What's this? Show me your utilization on this? I would be up on Wednesday night, getting ready for my Thursday morning meeting every single week, and he was, this guy was great because he would, he was so good at just giving me the next little piece I could handle, and never dumping on me the whole entire thing that I needed to be doing.
Carl Smith: Right.
Rob Harr: What that meant is for the two, two and a half years I did this, I was never comfortable. I never had the answers he wanted, but I was never so overwhelmed that I couldn't catch up by the next week, for what he asked for.
Carl Smith: What was his background? Did he have digital agency experience, or he was just a pure CFO?
Rob Harr: Pure CFO. He actually came from a manufacturing background.
Carl Smith: Okay.
Rob Harr: Was a CFO. Was retire consultant. He learned things along the way too. I would, we would kind of figure things out together and look at things and look at the numbers that were driving the business. For a business like ours, utilization is king. Good billable utilization, everything else takes care of itself.
Carl Smith: Everything else works.
Rob Harr: Yeah.
Carl Smith: For people out there who are struggling with forecasting, which is just about everybody, and you obviously, you're still iterating and still working on it, what advice would you give them? Obviously they've heard a lot already about the importance of, regardless of how you're billing, to understand where your time is going. How would you recommend that they make a transition to something that gives them better data?
Rob Harr: I think step one is to make sure your reporting is correct, and step two, use that data to project out what next week is going to look like, or the next period, whatever that is. Then grade yourself. You've made a prediction. You could accurately say what happened, and then give yourself a grade, and do that over and over again, until you close the gap.
Carl Smith: While they're doing that, I'm just curious, do you have any type of buffer that you put in? We're normally 15 or 20% off on this, so we're just going to go ahead and add it up?
Rob Harr: No. I don't want to ...
Carl Smith: I knew that question would kill you.
Rob Harr: Yeah. I want to be right. I want to get it right. I want to be accurate.
Carl Smith: You don't want it to be magic. You want it to be science.
Rob Harr: No. I want it to be science. I want it to be predictable. The things that I lose sleep over are when things, when I can't predict what's going to happen tomorrow. Now, it's ... That doesn't remove all of the unknowns, but every unknown that happens reveals an opportunity to figure out why it happened and then, turn an unknown to a known.
Carl Smith: Right.
Rob Harr: I want to make better mistakes. I want to make new mistakes. I don't want to repeat yesterday's mistakes next week. I think if you do that long enough, you've seen enough things where you can get in the 90, 95% accurate. Being able to accurately forecast things has so many great things that happen. The other thing that seems like all of these studio agencies struggle with is cash flow.
Carl Smith: Yeah, for sure.
Rob Harr: If you can accurately predict what things are going to happen, guess what? You can accurately predict cash flow. That's a superpower. Being able to say, we actually have a ... Part of what we do, every single week, is we are able to, we have basically ... We call it the micro cash sheet. It's a checkbook into the future. It starts with today's balance in our operating account, and puts in all of the expenses that we have. Usually it's a rolling 12 months into the future. It has all of the payrolls, all of those things, and the beauty of our business is our expenses are so predictable. So predictable.
Like the revenue is the part that we've got to figure out, right? That's the variable part, but we can predict all of our expenses into the future. We can put all of the things we've already invoiced and what due dates they're going to come in at. We can look at that and say, okay, we've got enough, with everything we've invoiced today, in our operating account, we're good til here, and whatever that date is. I can tell you exactly when that is.
Then we can layer in the forecast data, because we know we're good at forecasting, and be able to say it's like, oh, we've got enough money, right now to, with the weekly tax savings and other kinds of savings that we do, to say like, oh we've actually got enough signed work right now to be, to have money until this date. That's amazing, because then you can make purchasing decisions.
Carl Smith: Right.
Rob Harr: You can actually see slow downs coming. A couple years ago, when we first started, when we first kind of started really getting good at being able to really accurately forecast our cash, we actually had a slow quarter. It actually turned out to be, it was the most amazing and horrible experience of my career. It was amazing because we could see it coming and start planning months ahead of time. We could get tight. We could cut discretionary spending, and we actually were able to weather that storm, even though it was pretty bad for a quarter because we knew what was going to happen before it happened.
Carl Smith: Right.
Rob Harr: It was the most horrifying because it was kind of like watching a train wreck in slow motion that you know is going to happen.
Carl Smith: Right.
Rob Harr: And that you don't feel like you can do anything about. It was one of those that you ... But I don't think we would have made it through that, had we not had an idea of where our cash was going to be and like what it meant to go through that.
Carl Smith: Yeah. Well that's, I think that's, that's what a lot of shop owners and operators don't give themselves the time to build so they can see it. Right? They're so busy with the fires that are burning every day, that they don't have ... I don't want to see it's even that time. Maybe it's just not really knowing where to start, but maybe that start is finding that CFO who's going to beat you up on Thursdays, or finding a way that you're keeping your commitment internally and building that system out so that you're getting better over time, not thinking that you're going to throw a switch and everything's fine.
Rob Harr: Yeah. I think that's right, and I think it's a mindset switch that has to happen too.
Carl Smith: Yeah.
Rob Harr: I think it's really easy early on to figure out how to survive your studio and actually get addicted to that feeling. You can actually get stuck surviving and not have to make the mental switch and learn the new skills to be able to thrive. It's crazy to think that all the business fundamentals that are out there, that you don't have to learn, that you're immune to that. It's the overlap between where most of us are accidental owners and really good at a craft, and the business fundamentals, the overlap isn't there.
Carl Smith: Right.
Rob Harr: At some point, you have to forget everything you knew that you learned to survive to actually start to thrive in these things, and put these systems in place and to help hold people accountable but give them the data to hold them accountable. All the other pieces, and like, that's operations.
Carl Smith: Well, Rob, thank you. That was an amazing education in how you're doing things, and really, I mean, it makes such sense for how we can all make a step towards that and keep going. I know I'll never run a shop again, but man, so many things I would start off with differently now.
Rob Harr: It's crazy because really we talked about one little piece of what Ops is.
Carl Smith: Yeah.
Rob Harr: We just talked about some of the financial pieces and the reporting pieces, and how to, and the forecasting. Doesn't even get into the humans, the human side of all this.
Carl Smith: No. That will definitely be something we have to check back in on and talk about next time we get you on the show. I just, I want to thank you because you know what? I had this expectation that we kind of put operations in this nice little box, but I think by definition, that's never going to happen.
Rob Harr: Nope.
Carl Smith: Well, I appreciate you being on the show. Thanks everybody for listening and we will talk to you soon.
