Sådan opbygger du en kultur med datadrevet estimering (med Marcel Petitpas fra Parakeeto)

Read The Transcript:

We’re trying out transcribing our podcasts using a software program. Please forgive any typos as the bot isn’t correct 100% of the time.

Galen Low:

Well, it’s happened again. You’re putting together an estimate for a project that’s basically your company’s bread and butter. The scope is kind of like Julie’s e-commerce project from last August, but it has a bigger group of really vocal stakeholders kind of like DV is voice of the customer project from last April. It’s as risky as Fernando’s last project. That one went 150% over budget, but we’ve learned our lessons from that, probably. In other words, you’re dealing with another unique snowflake that can’t be estimated using auto-unique snowflakes.

Sound familiar? If you’re on a path towards data-driven estimation, but you’re struggling to get people to see project estimates the same way, keep listening. We’re going to be talking about the qualitative side of data-driven estimation, and we’re going to get really tactical about how you can build a culture of estimation to avoid these unique snowflake situations.

Thanks for tuning in, my name is Galen Low with The Digital Project Manager. We are a community of digital professionals on a mission to help each other get skilled, get confident, and get connected so that we can deliver projects better. If you want to hear more about that, head over to thedigitalprojectmanager.com.

Hey everyone, thanks for hanging out with us on the DPM podcast. My guest today is someone who is absolutely obsessed with project estimation after watching countless hours being spent creating estimates that are never accurate. He took up the entrepreneurial mantle to try to crack the nut on how to produce well-informed estimates with as little resource investment as possible. Today, he is the CEO of Parakeeto, a tool that makes sense of historical data to help produce estimates for agencies and in-house teams alike.

With what spare time he has, he still manages to coach CrossFit and if it’s gone on hiking, biking, and camping excursions with his fiance. Folks, please welcome Mr. Marcel Petitpas. Hello, Marcel.

Marcel Petitpas:

Hey, Galen. How are you doing, man?

Galen Low:

Not so bad. Thanks for being on the show. Great to have you here.

Marcel Petitpas:

Thank you for having me. What a, what a great introduction. Hopefully I can live up to the hype here.

Galen Low:

Hopefully it was all true.

Very cool. Very cool. Uh, Marcel, you I’ve really enjoyed our chats. I really enjoyed to get to know you. Um, and my takeaway has been that you are someone who is entirely on a mission to figure out estimation, which I think is a really noble cause considering how many people struggle with it. Your business Parakeeto, it offers consulting services.

Uh, it has its own data-driven estimation platform, it’s growing rapidly. Meanwhile, you still seem to have time to connect with like-minded folks in your network like myself. You still run the conference circuit and you still offer individual coaching. So, was just wondering, where do you get your energy from? What, what drives you?

Marcel Petitpas:

I get my energy mostly from sleep. I do a lot of it. Uh, I make sure I get my eight, sometimes nine hours a night. Uh, thankfully my fiance’s on board with that game plan. So, uh, that’s where I get it from that in CrossFit, you know, and just staying active. Um, and yeah, it’s interesting.

Like the motivation for this is really it’s, we’re focused on estimation right now, but the bigger problem is like agencies, especially small agencies that don’t have access to the same resources that big multinationals do really struggle to make a profit and to understand like what’s happening in their business.

And, you know, the reason we started focusing so much on estimation was that’s at the foundation of that problem. It’s the, it’s the root of all of the symptoms that a lot of people are feeling, even though they don’t often realize that. And that’s what we’ve discovered through the years of consulting that we’ve done around this problem space.

Um, and it’s just, it’s a big, hairy, difficult problem to solve. And, um, I think I’ve realized that like I need, I need things to work on. Otherwise I’m a person that gets very bored very quickly. So thankfully I’ve found this really nice problem to sink my teeth into and spend several years on it. And we’re still not even close to cracking, you know, that whole problem space wide open.

So I’m sure there’s many, many years to come of me continuing to invest in this.

Galen Low:

Yeah. And I’m serious about the whole taking up the mantle thing, because I think that’s what a lot of people find. They’re like, Oh my gosh, I wish I could just dedicate all my time to thinking about estimation. And even then it would still take me a decade to figure it out.

And you’re like, hold my beer. I got this.

Marcel Petitpas:

Exactly.

Galen Low:

I love that. Uh, inside and outside of work, is there anything specific that you’re trying to get better at these days? What’s on your sort of self-improvement bucket list?

Marcel Petitpas:

Hmm. Yeah, that’s a good question. Um, I think just being present is always something that I’m struggling with and there’s been an ebb and flow to that, I think, uh, throughout my life.

But it’s funny though, the older you get, the wiser you get and the more you realize how fleeting moments are, and um, you know, as an entrepreneur, Uh, and as somebody who, you know, works a lot, you tend to kind of always be thinking about the next thing. You kind of tend to always be planning. Um, especially as somebody who has been unorganized for the majority of like my life growing up and then had to learn how to be a very organized person, like you then start to always be in planning mode all the time.

You’re always assessing what’s coming next. Um, but then there’s moments like when you’re sitting at the table Christmas dinner with your family in the middle of a pandemic, thinking like this might be the last time I see them for a year and you’re having this existential crisis of like I’m in this moment trying to appreciate it.

And yet I’m viscerally aware of how fleeting it is and I’m actually like experiencing it, go by and already be like, every moment is becoming a moment to the past. So I’m just trying to reconcile that with myself and really learning to value the little things and try to turn my brain off and be present.

That’s probably the biggest struggle for me. And the thing that I’m always trying to improve.

Galen Low:

I think that’s a good one as well, for sure. For sure.

Marcel Petitpas:

And for all the PM’s listening to this can, uh, can associate to that feeling, like you’re planning for the future. Always trying to figure out what’s happening next.

Galen Low:

Why do I’m thinking about this milestone while I’m at Christmas dinner? Just be present guys. Just be present.

All right. Let’s get into it. Let’s, let’s talk about data-driven estimation and your take on how organizations can stop throwing so much of their most senior people’s time into the seemingly moot exercise of creating a project estimate. Um, but I thought maybe let’s talk about sort of the origin story a little bit.

Um, tell me, how did you become so interested in estimation to begin with?

Marcel Petitpas:

Yeah. So, I mean, it started with a fascination with why it was so hard for small agencies and service businesses to answer simple questions about their business without wasting a ton of time in spreadsheets, you know. Things that as an agency owner or an executive or an operations manager, even a project manager, you’re asking yourself.

Um, and like, they’re just, you can’t answer the question without going and pulling data from a whole bunch of different places. Like, did we make money on this project? Simple question. One you should have the answer to most people listening probably can’t easily answer that question in a couple of minutes, same thing with like, do we have the capacity to take on this new project?

Do we need to hire people? If so, when, um, you know, what, what does our utilization look like? What, what was our profitability? What services are more profitable than others? What clients are more profitable than others? Like all these questions require, um, a ton of data crunching in spreadsheets if you’re a small agency, because the tooling that’s out there is just so it’s sparse and you have this data spread out across several different places.

Often it’s not well-structured, often that the connections between them is not very good. And, um, that just creates a whole host of problems. So that was the original kind of problem space that I became interested in. And we started, you know, doing consulting with agencies to get closer to that problem space with the end goal, always having been to build a software product.

And the more we spent time, um, you know, helping agencies start to figure this out. The more we realized that the basis of all of this is estimation, because that is where you are creating the assumptions that most of these operations functions are built on top of. And you’re also creating the structure in which that data is being organized, which determines what kind of questions you can and cannot answer about projects.

Um, and that’s why we really started to kind of narrow in on that because we had spent a long time trying to build technology to solve the problem without actually getting down to the root of like, Why, why are we struggling to solve this problem? Why is, why are people’s data sets so disorganized and incomplete and poorly organized and inaccurate?

And, you know, it just led us to this estimation thing, which today I realize is like, if you can’t do that well, you can’t actually scale your agency well, because you can’t predict and plan the things that you need to have forward visibility on in order to hire people at the right time, get them ramped up, make sure that they’re not working crazy hours.

Make sure you’re protecting your margins. Like the best thing you can do for your agency and the people that work at it is get good at understanding the scope of work so that you can plan that going forward.

Galen Low:

I really liked that. I like that sort of, uh, for those of you who are watching the audio recording of this in Marcel’s background is a copy of Start with Why by Simon Sinek.

And I like that notion you kind of alluded to it that, you know, you kind of started looking at the estimation problem, uh, and then, you know, not a, how can we sort of estimate better or what is a good estimate, but why do estimates suck in the first place?

Marcel Petitpas:

Yeah. Yeah. And of course we uncovered a whole bunch of problems around that as well. You know, most agencies when we asked them, like, how do you figure out what the scope of a project is? The answers range from like bad to terrible, you know, very few agencies who we run into were like, Oh, we feel really good about how we do this. Like, we feel like it’s data-driven like, we were confident in the numbers we’re throwing out.

And for most of them, it was like, we get all the most expensive, most senior people in the organization together in a room, and we spend thousands of dollars in potential billable time putting together an estimate. And then we’re constantly stuck in this dichotomy of like, do we spend a bunch of time on this to make it accurate, knowing that there’s like a 60% chance we’re not going to close the deal anyway?

Or do we just kind of like, you know, pull, pull an estimate from the last project, make a couple of tweaks and cross our fingers, hoping this doesn’t turn into a dumpster fire. Um, and none of those are really great answers for really fixing this problem.

Galen Low:

No, for sure. And you mentioned about like all the spreadsheets and all the time spent and the sort of. The spectrum of bad to terrible in terms of estimation techniques and tactics, uh, like were you finding that a lot of folks are just kind of saying, okay, well, let’s just treat this like a net new project and think about how many hours it will take to do X or where some people are were a lot of people pulling from their historicals and just trying to make it work. Um, but maybe sometimes failing.

Marcel Petitpas:

Yeah, it kind of runs the gamut. I think that there are a lot of agencies that overestimate how unique all of their projects are. And so they do take this approach of like building every project as though it’s like really a bespoke. And I think what they’re missing out on there is the opportunity to create efficiencies in what is actually a pretty expensive process.

And it can be the reason that you lose a deal. If you, if you need four days or six days to put together a scope of work for a client, like that’s a, that that could be the reason that you lose that deal. And it’s, it’s expensive, like objectively that, that requires a lot of time from people in your organization that should be working on earning your revenue.

Um, the flip side of that is those that are trying to pull back, uh, estimates from past projects. The mistake a lot of people make there is they don’t actually close the loop on estimates versus actual. So they might grab a proposal from a previous client, but they’re not actually going to reconcile, like, was this accurate in the first place?

And so, um, if they’re not doing that, of course they’re using what could be an inaccurate basis for their assumptions going forward. Um, and if they are trying to look at actuals and base their next estimate on that, most of them struggled to do it with a volume of data. That’s actually going to give them like a benefit in terms of making their estimate more accurate because it’s just inefficient to create that feedback loop.

It’s clunky, it’s happening in a spreadsheet. It requires manual pulling of data. And a lot of times the reason that that doesn’t happen, and this is one of the biggest thing is I spend my time on with agencies is because there is a disconnection between the way their estimates are structured from a data schema perspective and the way their time tracking is structured. So when you go look at your time tracking records and you’re like how many design hours did we spend on this project? It’s like okay well I have a task here for a wire-framing UI content outline. And I have this thing over here for you know client this that and the other thing but it’s not tagged back to a schema that maps to the estimate. And so, the reason you don’t do it as often as you should or the reason a lot of agencies struggle to do it at the volume that they should is because there’s this huge overhead of having to reconcile their data in order to actually just get that question answered in the first place.

And so there’s too much friction like the way they’ve structured their data has introduced unnecessary friction into answering that question that they are tracking time for the purpose of answering in the first place.

Galen Low:

I had actually the question here why doesn’t everyone do this sort of data-driven estimation process? And I think you just answered it there because A and B don’t really connect. They don’t want to invest the time to actually you know, I put the puzzle pieces together that actually ends up taking even more time and it seems like even more fruitless. Um but really again coming back to the why isn’t it working in the first place it’s usually because of the way people create an estimate is different from the way people are actually tracking project metrics. That’s really interesting.

Marcel Petitpas: Yeah Yeah And it’s interesting. What I found is that there’s kind of two you know what we’re really trying to do is create, I think in the estimation process this formula that we use for how we scope work. Right? So we have inputs, those inputs are generally things that we’re asking the client for trying and get a sense of the complexity of the project. So it might be like you know how many webpages do you need on this um on this website and how much of the content have you already written?

And you know like you’re trying to figure out like how much complexity is there here? And then based on those inputs, you’re trying to create some kind of relationship line between like if there’s X number of webpages it’s going to add this much additional effort and then categorizing that by you know the way that you resource plan which is generally on some kind of a rules basis. So like how much more design, how much more development, how much more copywriting strategy, et cetera is this gonna require. Um and there’s two ways to build that line in a way that’s actually accurate and reliable.

The first is collecting enough data to get a sense of like, What does this actually look like? And what is the correlation between these inputs that we’re collecting and effort. And then the second, and this is one that a lot of people overlook and this is where actually talking to people becomes really important. And I, you know I don’t want to be uh put in the box of being the person that thinks that like, tracking time and looking at data is how you’re going to solve this whole problem because it’s not. Is looking at the process because the reality is there’s things that you’re trying to predict that are probably just not that predictable because the process for how you do them is not very well-defined. And so naturally you’re going to have this really huge distribution of data points on that graph. The the the spectrum is going to be too wide for you to get a reliable line. So what you want to try and do is collect a lot of data points and then squeeze them towards the middle by placing rails on, like what is the possible variance of the amount of time and effort it’s going to take to get this thing done by defining your processes.

But it’s a bit of a chicken and egg game, isn’t it? Right? There’s unlimited processes for you to invest in in your agencies, so how do you figure out which ones are the biggest opportunities are going to be worth the squeeze so to speak. And the answer to that in my opinion is you look at the data to figure out where you consistently blowing budgets, where you consistently not predicting the effort of getting something done.

Galen Low:

Makes sense, I like that. Um you had at one point walk me through a bit of a framework for estimation. Like a bit of a cycle. Uh it has various different stages I wondered if maybe you could walk our listeners through that a little bit.

So those are the exact specimens less the less mate Sorry

Marcel Petitpas:

There’s four steps, um and this is a framework that we’ve kind of called the agency profitability flywheel. So this is really at the core of the process that we use for consulting engagements and how we think about you know our product and how we think about improving profitability in the agency in the first place. And of course the first step is getting good at estimation. That’s the foundation as I said before.

So the first step in the flywheel is defining your process for how you estimate, Um work inside of your agency and then also defining the structure of what those estimates look like, because if that structure is changing all the time and if your process is changing all the time and you have a moving target that the rest of these pieces that I’m going to talk about are trying to align to and you’re just creating unnecessary complexity there. So the first step is just like define what that process is and then standardize the format that an estimate comes out in and what’s more important it’s not like, I don’t mean what font you use for the document but like the hierarchy of the data.

There’s a client and then under clients there’s projects and maybe inside of projects, there’s deliverables or phases. And then within those you have like the standard kind of line items that you roll time estimates up to. Maybe it’s roles, maybe it’s people, you know however you want to do that for your agency it doesn’t matter but just structure it in a way that’s consistent because the next step is then to measure actuals. So you have your estimate, now you want to go and structure all the tooling that you use for tracking time and costs. You want to set those up so that they map it back to the estimate. And this is a thing that seems so obvious and it’s like why am I even talking about this, but like go audit your tooling and hold up an estimate next to your time tracking reports and tell me if they match up. If they do, congratulations. But most of the time when I talked to agencies it’s the thing that seems so obvious but it’s not happening. So that’s the next step is to align your estimates to your time and cost tracking tooling. And what that does is it decreases the amount of work that it takes to actually check if your estimates are accurate or not and you create a data feedback loop there and now it doesn’t take you an hour to answer the question of was this accurate or not.

It takes you two minutes and you’re more likely to do it. And now you can install a cadence that weekly, biweekly, monthly, whatever the rate of change you know in your agency. Whatever is appropriate for capturing you know, relevant rate of change. Set a cadence, go on a review. What did we estimate, what’s actually happening, and you can start to see patterns of where the gaps are in the way that you estimate where the consistencies or inconsistencies exist. And that’s going to drive the next part of this framework. So this is kind of quantitative, it’s very data. It’s very objective, right? That should as a leader in the organization, as a PM, as an operations person give you data points to understand like where should I be investing my attention. Where are the opportunities?

The second part of the framework is all about qualitative. So what are the people that I need to go talk to, to figure out why this stuff is happening. Because the data is not going to be able to tell you that. And I think the mistake that a lot of people make and it’s the reason that their data set is a mess is they try to answer every question with their data. And so they’re trying to segment it down to like by task and then by sub-task within the task. And that’s how you end up with a time sheet that’s just like unreadable and unreconcilable. So like don’t try to do that because you’re not going to get good insight, instead understand that it’s like okay, We’ve consistently been under estimating project management by 20% on all of our you know e-commerce website design projects.

So then you go and you sit down with the team and you go, Hey guys we’ve noticed this pattern. What do we think is happening here? Are we not estimating it properly? Is there some clunkiness in the hand-off that’s happening here, and you let the team kind of surface insight on why they think those things are happening. They’ll surface insight on better ways to estimate, they’ll surface insight on better ways to do this process, and streamline it and make it more consistent. And the beautiful part about that is their ideas are coming to the surface, and then you can say that’s a great idea. I think that would be really valuable. Do you want to help us get that implemented? And you can actually start to recruit the team to get involved in creating the processes that are going to make the agency better and hopefully make their lives better and make their work more consistent and keep them you know getting home in time for dinner every night. Um and you’re not now having to just like dictate these processes which I think is something that a lot of people struggle with inside of agencies. So the next two steps are meet with the team regularly to review what you’ve discovered from your data sets. So that’s like meetings and reports. And then the fourth part of this is create a backlog of process improvements, updates, you know maintenance that you can prioritize and implement. And that will in turn make your estimates more accurate. So it’s a, it’s a circle. It’s like a circle.

Galen Low:

You wanna might say it’s a bit of a flywheel.

Marcel Petitpas:

It is a flywheel. Exactly, cause like what you’re doing is you’re doing exactly what I described earlier. When you go through this process you’re collecting the data points that allow you to create this relationship line of you know, how do projects increase in terms of effort. When you know the inputs that we’re asking for in the discovery call look like this, and then on the backend you’re also looking at where can we improve our processes so that those data points start to get pushed towards the middle. If you do this for just a couple of months I promise you you will identify a handful of processes that will dramatically smooth out you know, the distribution of like how much effort and the risk of how much effort or time it might take to get things done. And you’ll start to see things become way more consistent, moving more quickly.

And of course where that’s most important for the executive leadership or for the agency is people spend less time working over time because the deadline is rarely elastic. And when you underbid something that’s costing people their evenings and weekends. So that’s most important part. And the second thing is your profitability will go up, because when when you plan to make money when you sold the work hopefully actually make money when you deliver the work.

Galen Low:

I like it and I think that like, I feel like this qualitative stage is where a lot of people are falling down. I think you mentioned it earlier, right? And some people just expect the data to answer all the questions. Maybe they actually get too granular and then the timesheet and the estimate is so granular that nothing is similar, you know in the past or in the future in terms of projects. Um but I really want to dive into this sort of qualitative stage because I think folks will be interested in it. Um you mentioned getting together regularly and talking about you know estimates and you know who’s over, who’s under and then talking about the reasons why, would you say that that’s like uh like everybody together. Do you get people of a certain like craft or discipline together? Is it just that line item that went over or under that you want to analyze and get those people together? What does that what does that meeting look like? Who’s there and and and how do we start talking about it?

Marcel Petitpas:

Yeah. So, unfortunately I don’t have a cut and dry answer for you because we’re dealing with human beings and a lot of this is going to depend on what your organization looks like. So if you have, you know like pods inside of your agency that worked with you know a specific group of projects and you might run this at the pod level. If you have more vertical disciplines then you might run it you know at that level where you have the designers sit down and talk about their process and it really just depends on how your organization is structured, what your delivery processes look like. Um but ultimately, it’s important to just protect this time and make sure that it’s happening consistently.

And the two ways that we typically see this done is number one, Project Retros which most people are familiar with those. At the end of a project you schedule some time to get everybody that worked on the project in a room and you talk about you know, here’s what happened. This is what went good. This is what didn’t go the way we thought it was going to. You know, what can we learn from this and you know you use that to inform the process backlog.

The other way to do this and I kind of prefer this, um only because it’s a little bit easier to protect the time is I call these performance um, Project Performance Meetings. And we do these usually bi-weekly, or monthly um just depends again on like, what is the rate of change in your agency. Do you have big projects that take a long time to like really kind of move, or do you have a lot of smaller projects that change quickly? That’s going to determine like how often you want to do these. Um and with those you’re you’re going to define like a group of people across functional kind of section of the agency. You’re gonna bring them together and you’re just going to have a list of all the projects, the ones that were completed, the ones that are kind of ongoing and you can use simple cost performance indexing to get a sense of like, are we on track for this or not.

And then you’re just kind of looking at the outliers, the ones that are way better than we expected. And like I think not enough people focus on that stuff but like there’s gold in there of this project was super efficient, like what did we do here that we can apply to other projects. And people are like, Oh well you know like, we had all the designers work on this one phase you know, at the same time. And we ended up being able to bang it out in a day as opposed to like it usually takes us a week to do it or we did the handoff differently here and like it was super efficient. So like look at the outliers in terms of what went super well, what was super-efficient. And then also look at the outliers in terms of like, what’s not going the way you expect.

Um and it’s just a conversation with the team about like why do we think these things are happening. Um and what’s important here is we want to focus on the process not the people. And we just want to like be inquisitive and not you know attack people because that’s one of the best ways to just basically get people to like write these meetings off, not want to show up and not engage in them. Is if we show up and we’re just like, Hey Galen you messed up man. And you spent twice as much PM time on this. Like that’s not a productive way to have that discussion. It’s like Hey it looks like you know our assumptions about how much PM time was going to take for this project Um didn’t line up with what happened. Why do we think that happened. You know what went differently than what we expected and that’s the way to facilitate that conversation.

Galen Low:

And I think that’s a really important lens. I mean you mentioned project retros and I see the value in that for sure. But I love the idea of having this sort of maybe process subcommittee across projects looking at everything that’s happening now, making tweaks in real time but also like it’s it’s that comparative side of things um that helps people understand what can be compared across their projects. Um and have that discussion of, Oh yeah you know like it’s taken longer because it’s really complex or you know there’s a lot of stakeholders involved so I spend more time communicating. Um but I think you know one thing that uh that that sort of jumped out at me you mentioned earlier okay, getting like some PMs together talking about why PM time always goes over. Um and then there’s this question that uh you know we have in the in the circles I travel in which is you know, does everyone see the value of what a project manager is doing?

Uh and then I think that actually starts to tip it because I seen I’ve talked to a lot of folks in agencies that are like okay Yeah Like project management is like tax. It’s like a percentage that you just throw on top uh regardless of what kind of project it is Uh and then everyone’s like Oh why why did it go over? Or you know people who are paying for it are like well why do I have to pay for this big chunk of project management time, we don’t see the value but I love that notion of getting together and understanding where PMs are spending their time. How they’re adding value and then maybe reflecting that back. So maybe that percentage should be higher because they’re delivering more value and that’s also how you can sell it which is also how you’re going to estimate it because you know that there’s additional value on top of what most folks you know typically thought a project manager will be doing on a project like that is actually maybe a misconception. So I think that’s really cool.

Marcel Petitpas:

And since we’re on the digital project manager-focused show, I think it is important to riff on this just a little bit because the outcomes that I’ve seen come out of that discussion a lot are one of two things. It’s to your point it’s like actually helping everyone understand the value of the project management team is bringing. Um and then the other one is getting an understanding of like where all the shortcomings are in the rest of the organization that are holding the project manager back from doing their job.

Um and I think uh Ben and I had a great discussion on our podcast about this where we asked the question like what is the central responsibility of a project manager? Cause like I think everybody listening to this knows like project manager is one of those job titles that it’s like You have no idea what that means at one agency versus another, right? It’s like you could be doing basically sales. You could be doing basically operations. You could essentially be a designer or a developer or like doing like the spectrum of where you’re going to be pulled is different in every organization based on how mature they are. And based on the other thing is like where the weakness and the inputs are. And what we agreed on was the central responsibility outside of all of those kinds of exhilarate things that you get pulled into as a PM is building and maintaining the project plan.

And that’s super valuable because it tells everybody the team, the executive leadership team, the client like where things are at it’s the source of truth for everyone to be informed on what they should be doing, what the focus is, what the progress is, like where the project is. It’s so essential and all of your operation systems are built on top of that forecasting, resource planning all that stuff. So like the project plan is the central responsibility. It’s super valuable to everyone that touches the product, project all the stakeholders, and generally as a PM you’re going to get pulled in whatever direction is the weakest in terms of the inputs that you need to put that project plan together So if sales is doing a terrible job of scoping you’re probably going to end up spending a lot of time with the sales team helping them scope projects If the delivery team is doing a terrible job of updating you on project status you’re probably going to spend most of your time chasing down time sheets and trying to get updates from the team If this client is doing a terrible job of giving you the information that you need to try and like actually Get the team going and get them on blocked and get revisions back You’re probably gonna spend most of your time on account management So um that could be a big insight is just realizing like okay we’re spending all this time over here because we actually don’t have a great process for collecting you know these important things from the clients And that informs process improvements that hopefully allow you as a project manager to kind of keep coming back to that central Thing which is creating and maintaining your project plan and making sure that you have all the inputs and the correct tooling and the correct process for getting the outputs um that everybody needs and that you need to do that job effectively

Galen Low:

I like That’s why process is so important so that you can actually process should kind of run itself so that you could be left to do your job as a project manager and add value in other areas Um I I totally support that I love that Um I think I mean I want to get into the qualitative stuff in just a bit but I think even just to take a half step back um the thing that I get asked a lot is uh well Maybe just told a lot is okay My my project is very different I do things my own way like and it’s never going to be the same as somebody else’s project and you know I’ve got my own style and everything’s always different This is it’s probably pointless to try and use historical data to do anything or to try and unify processes because the fact of the matter is that every project is unique How do you defend against that What is what is your what is your argument against that

Marcel Petitpas:

It’s just not true I mean like I think like you you have to think about um You have to separate the creative from the process And the reality is that every creative has a process And um I think honestly one of the things that separates the mature creatives from the ones that are still kind of figuring themselves out as creatives is that the mature ones acknowledged that they have a process As ridiculous as their process might be as weird as it might be Like your process might be to you know smoke a joint and go for a run in central park That’s still a process if you do it every time or if you do it most of the time And so you have a process and it might be very high level it might be very loose but there is one there and defining kind of at an even at a very high level what those are even if it’s just like the order of the things that you think about and start to wire frame and start to design Like those things matter and they make up 80% of the the patterns in the way that the data shakes out And we know this from looking at the data from hundreds of agencies even the ones that say like everything we do is totally to bespoke It’s like well maybe that’s true but like either your process for figuring out like how you’re going to do these things is still relatively consistent Um and then you generally know what signals you’re looking for in the discovery process to figure out like how much complexity is going to exist in each of those Steps in your process Um and you can generally figure those things out by talking to the clients and trying to extract you know what they do or don’t know or what they do or don’t expect from the deliverable So um I don’t know if that answers the question but it’s like it’s just it’s not true Even if every design is different the process for getting there probably isn’t unique every single time

Galen Low:

Yeah And I I fully agree with that Um and it’s something that we’ve kind of come across especially in digital like on the web everyone thought you know their website was so unique I mean back in the days of flash driven websites everything was like a movie Uh and everything was very different and actually how it got massaged down and like now Like browsing the web is like it’s a vernacular it’s a language it’s it’s pattern driven It is you know uh it is componentized in such a way that it actually supports process And actually that there is a benefit Although I mean I was I was at Maverick PM who was like Oh I just do things my way And that’s my style and I’m not going to do what everyone did And until I realized the value of actually doing things consistently and the same way and learning from one another And then to your point you’re not going home on time because I’m not like trying to make up my own method as I go or be special or be this unique unique snowflake that I can do what all my peers are doing So that we’re this sort of you know platoon of people doing the same thing creating the same level of quality And also being able to say Oh yeah this is how much time it’s going to take because I know that’s how much time it’s taken on average historically across my peers to do this job uh at this level of complexity And I can sort of rest my hat on that know that we’re getting better as an organization and yeah Have have better work-life balance So I think that’s the thing that sort of tipped it for me

Marcel Petitpas:

Yeah And I think I want to just dispel one final myth around this which is and I think you get a lot of this again from the creative community is that there’s a sense that process is going to somehow inhibit their creativity or it’s going to you know diminish the quality of the work And what we found is that that the opposite is actually true Like there certainly is a point where adding too much process will restrict You know the creativity and the quality of the work but having a little bit there actually dramatically improves it because it gives you a framework to operate in and it starts to to your point start to protect the consistency of the quality that you’re producing for your client And at the end of the day it’s not not what is most important Like they’re coming to you buying an outcome You’re generally selling them on like We are going to get you this outcome And part of that conversation is because we’ve done it before and you should be able to describe why you’re able to consistently deliver that outcome Um and so like just having even again if it’s very high level having a process that you go through as a creative to get into that space where you’re doing great work Um you know understanding that that’s part of it is important And I think it can actually elevate your creative work and make you a more reliable and consistent you know creative that can produce that work the way that you want to every time And doesn’t find you uh it doesn’t leave you in a place where you’re just like finding these flee fleeting moments of creativity and needing to rely on those in order to do work for clients because that’s not a business model Um and it’s risky And it’s to your point it’s going to put you in a position where you’re probably going to burn yourself out

Galen Low:

Yeah no absolutely Having an anchoring workflow so that you’re not just kind of hoping for that inspiration Every time you get paid to do a job

Marcel Petitpas:

That’s right

Galen Low:

I like that Uh I want to dive into some of the qualitative sort of I guess data points you mentioned some things you know like complexity you mentioned things like okay well you know they have really demanding stakeholders or you know maybe the sales team needed more education on how to add you know how to sell this product or you know there’s these Jennicet clause for all of these projects that a lot of folks will be like well you know it’s It’s not that easy Sometimes you have you know a client that’s you know a bit of a pain to manage Uh how do you sort of turn that Uh or how do you apply a data lens to some of that qualitative aspect of how projects go in real life

Marcel Petitpas:

Yeah So again I think it’s it’s a question of creating um a bit of separation between how you’re going to answer those questions and understanding like where it makes sense to tool that up in the data and where it makes sense to just continue to rely on you know more of a qualitative approach conversations with the team in order to get to the root root cause of what’s going on Um but certainly like what is feasible from a data perspective is kind of defining a couple of standard Um I guess drivers that influence the complexity of work um within your your phases or you know within a project So if you have this kind of standard list of like three or four things that you’re asking for or that you’re kind of uncovering in the discovery process that you know have a dramatic impact On a project then you can start to create whether it’s a naming convention in the project or if it’s you know some kind of other way that you define that against the project you can start to create sets of projects that are similar based on that criteria and kind of where they rank And then the final piece is just kind of using contingency for all the other little things that just doesn’t make sense to track And that probably aren’t impacting um the scope of work Like as much as you think it is In the grand scheme of things when you actually look at the data normalize that at high volume Um but that’s where you can just kind of be like well this client feels like they’re kind of a pain in the ass so let’s just tack on an extra 10% douchebag tax for this client And hopefully that’ll make up for the extra emails we’re going to have to send uh to keep them in line Um and then on the back end of that it’s like again being consistent about having conversations with your team about these things is super super important And uh the pushback I get on this a lot is like That always gets bulldozed by client work And my answer to that is it’s because your client work isn’t happening in the amount of time that you thought it was going to And so you have to start cannibalizing time in the calendar within work hours before you start cannibalizing time outside of work hours And it’s like this is just a it’s another reason you need to start doing it and to do everything you can to protect it and start with the data And hopefully that will improve things enough that you can start to find the time to actually have these conversations but like the reason it’s not happening is because it needs to be happening You know what I mean

Galen Low:

I like that like the cannibalizing that time is the symptom that tells you that you need to fix the process

Marcel Petitpas:

Yeah it was just like kind of a crappy answer Right Cause it’s like that’s not a solution but it’s the reality of kind of where you are as an agency If you schedule these meetings these retros and they’re always getting bulldozed it’s like well that’s because if you plan to have that time and then you don’t have any more it’s because the things that you plan for project-wise are not happening in the way that you expect them to

Galen Low:

I mean there’s this like especially in agency world there’s this urgency and it’s very difficult to get anyone to sort of take their foot off the gas for a bit Even if they’ll benefit from pausing Taking a step back going okay what’s actually happening here and let’s fix the root of the problem before we continue because there’s just this rapid pace high pressure sort of culture and environment to just keep going and keep smashing through things and catching up and then finding the next you know the next job Um and really a lot of the sort of just anxiety and fervor and just chaos of it all could probably be just you know gently pushed back a little bit by you know sitting down and looking at the process and how it can actually be improved So

Marcel Petitpas:

I want to riff on this just a little bit because it’s it’s one of the big reasons that I’ve gone deep into this problem Um and it’s something that has driven me crazy about this industry for a long time which is like I think for a long time working at an agency is you kind of know that it’s going to be Hell right You’re going to work long hours You’re gonna you know it’s going to be crazy You’re going to live on coffee and cigarettes and like it’s just going to be terrible Um but you’re going to cut your teeth You’re going to get to work with cool brands And then eventually you’ll go like work for a brand that allows you to go home at five and pays you twice as much as you got paid at your agency Um and like that just doesn’t actually have to be true I think that that’s that’s the story that a lot of people tell themselves because that’s how it’s been in the industry And I think there’s a lot of bad players that their model was I’m going to let my employees subsidize my poor management and poor scoping of projects And that’s what ends up happening.

You work 70 hours a week and that’s actually helping your agency owners like and the people running the agency make up for the fact that they are not actually designing and running their business in a way that’s profitable The reality is you should be able to run a highly profitable business 25 30% net profit margins at the end of the year on 65% utilization 65% That should be your annual net target And you should have no problem being profitable at that level of utilization That means that like 35% of your team’s time is not spent doing work for clients Imagine that for a moment And this is within a 40 hour workweek right Normal hours you should be able to be profitable So like if you’re listening to this and you’re more on the executive leadership side or European who’s like maybe having this discussion like those that’s the reality.

Those are the numbers you go model it You should be able to achieve that There is no reason why you have to get your team to work 70 80% utilization to have a profit That just means that you’re not doing things right And again all of that generally comes back to the starting point is you’re not scoping projects Well you’re not You’re not planning for that And so you’re just kind of letting your team eat that cost instead of you absorbing as an owner And that drives me crazy And it’s the bigger reason why I want to solve this estimation process Cause like I’m sick and tired of people believing that it’s okay to you know work their employees to death and cannibalize their whole personal life because that’s just the way this industry is because it’s not true

Galen Low:

I really love that actually And I think that’s a really useful perspective because I know from past experience uh sorry sort of being at a leadership level at an agency and the process trying to make the organization attractive to finance the areas and make it sort of right for acquisition One of the one of the metrics actually that we looked at was utilization And we were like more of the 80% and that was a good thing It was like Oh we’re consistently at 80% utilization In other words our workforce is being sort of mobilized towards client work and nobody’s quote unquote Wasting time you know sitting on their hands Whereas the opposite is is is actually probably more valuable which is that Hey listen we’re able to put 35% of our time into thinking about how can we work better and more efficiently and more consistently and deliver a higher quality And then the 65% of the time we’re doing that thing and we’re doing it really well And we’re We’re working at a margin that is still like very like very healthy in terms of profitability

Marcel Petitpas:

And this is the thing that I fun Again this is an old world metric that’s romanticized today but it’s completely like the wrong metric to be tracking Now if you are a time and materials Pure time and materials agency and you bill for every hour that your team works then yes utilization is your most important metric and you want it to be as high as possible you know without burning your team out But the reality is most shops are not time and materials anymore And so just focusing on utilization is the wrong call because If you’re not looking at average billable rate or some kind of earning efficiency metric whether it’s average billable rate or gross margin or contribution margin next to that then yeah Maybe your team’s working 80% of the time on client work but maybe that extra 15 or 20% of their time that they’re spending is just over servicing the client and it’s driving down your actual Efficiency And so um I think acquirers are becoming more savvy to that now that they’re seeing the business models changing most people are charging on value or flat rates And so utilization cannot be the only metric that you look at because it doesn’t tell the whole story Um in fact it can obfuscate and draw your attention away from the fact that you’re actually just throwing human beings at the problem instead of fixing the fact that you’re not efficient at getting your revenue earned in the agency

Galen Low:

For sure Now efficiency is huge There was actually in a time materials world There is no motivation to actually be more efficient because you want the billable hours You want to you know you want that You want to hit your utilization target You’re not going to spend less time on something You’re gonna spend more time on something

Marcel Petitpas:

Yeah a hundred percent

Galen Low:

That’s very interesting I love that Um Oh so many other questions bubbling in my head but I think uh coming back to sort of process right And efficiency and and talking about you know uh how things are going or have gone in a project and then talking across projects Uh and then you mentioned it’s like yes this the this group or there needs to be champions of process to actually prioritize what processes to look at and improve and then roll out And then In rolling it out I think the only way this really sort of works is if people actually do it right the way that the process is meant to be Um and I don’t know if you know the people you talk to and the people that you work with you talk about like like how you reinforce adherence how do you sort of make sure that people follow the process once it’s been overhauled

Marcel Petitpas:

Yeah I mean this kind of comes back to organizational change 101 and you know like you can go to the most like corporate just like staunchy information on this And it’s all going to tell you the same thing is like your team has to be bought in to the process if you expect them to follow it and and reinforce it and actually be invested in making sure that their peers are following it And they have a vested interest in making sure that it’s improving And the only way to do that is to get them involved in the process of ideating and creating these processes early on because that’s where they develop a visceral relationship with this is how this process or lack thereof Impacts me and my job and it impacts my ability to do my best work It impacts my ability to get things done efficiently It impacts my ability to get home on time at five o’clock you know like and when they are drawing the line and this is what should be happening in those kinds of project performance meetings is like Hey you know Galen really sorry you you worked a slammer you worked over the weekend to get this done because we we miss scope this thing like You’re like okay that’s how this is impacting me And then we have the conversation about how do we solve that And you say well I think we need to change the way we do our handoff between design and dev because it’s creating problems now Like you understand viscerally like This is how I get my weekends back And now you’re motivated to make sure that that process is implemented that you follow it and that you get other people to follow it And I don’t have to be the bad guy as the PM anymore And I think that’s another unfortunate thing about this role is often you’re just kind of put in that position where you have to be the person that’s Holding everyone accountable to the process Well in in this case you don’t really have to be that person anymore Or you can just be a facilitator of a conversation about like Hey how do you think this is impacting your peers When you don’t follow this process We had this discussion you know and you’re it’s not just like follow the process because I said so Uh which unfortunately like when you’re taking more of a top-down approach to process ends up being the case So um yeah this is just change management 101 get the team involved And the way that you do that is you get them connected to the facts which is in the data and then the emotional side of it which is how it’s impacting Them and their work and how much fun they’re having on projects and how much time it’s taking them And then having them actually champion the ideas that are going to make that better And honestly no one’s really better qualified to do that than the people doing the work Right So um that’s another great way to remove bottlenecks at the executive level when it comes to maintaining process and creating accountability around it as well

Galen Low:

I like that Hold up I’m gonna shift gears a little bit Uh and I just want to talk about tools versus techniques Um so PMs we love tools but I think I think for me the question is uh And maybe this is an obvious one but does an organization like need a software tool to gather this data to produce estimates and also like do time tracking like is there a sign of kind of a smoking gun that everyone should be looking for to solve all their problems Maybe I’m asking this question myself Uh but like you know we even talked about spreadsheets and time tracking and just having you know the estimation process matched time-tracking like can an organization do this without having a new tool Is it more of a mindset and a framework or does the data get a bit overwhelming

Marcel Petitpas:

Yeah no I mean absolutely you can do it um without a tool and the most important way to make that possible is being really really deliberate about the way you structure your data so that when you pull a time-tracking report and you pull an estimate and you put them up next to each other you don’t have to do a bunch of work to make that That comparison actually makes sense like that will help speed this up and probably get you to a point where you can do a lot of this stuff manually but it doesn’t change the fact that as you know you know creating that feedback loop is cumbersome It’s probably gonna require you to build spreadsheets export things import them you know tidy things up um create visualizations and um that any amount of resistance that you place on that when you’re already in a process that feels kind of arduous and already feels like you know a big expense in the sales process which is creating a quote and where there’s time pressure means you’re likely not going to do it or we’re not doing it often enough And so you know this is part of the reason we created Parakeeto Um and to my knowledge it’s the only tool that really closes that loop And I think that’s what we do really differently Everybody else between what you estimated and what actually happened and allows you to put all of that data where you’re creating the estimate so that as you change your estimate for how many project management hours you think it’s going to take on this project on that same graph that same visualization you can see all the other projects that were similar and how much time it took for project management on those as well as that line I described earlier of what is the relationship between budget and effort and how does that change over time for your agency So you know we created this to make that As frictionless as possible so that it’s easy for you and your team to actually do it Um as opposed to having to rely on like your your process of doing it manually which even when you have your data structured really really well it’s still going to be cumbersome

Galen Low:

Yeah no definitely And you and I were talking about sort of the the operations layer and you know maybe PM’s aren’t always the best people to sort of decide how to structure a time card or decide on some of these processes uh for estimation but then also there’s this process of like Pulling all the data together Uh and if you’re doing it manually then that’s a lot of effort I’m like I’m just picturing someone going through every time card of every project and trying to bringing it into this Google sheet and hoping for the best It’s just at a certain scale I think it gets um it gets a little bit too complicated to actually reap the benefits of having You know and I’m just picturing that visual of like okay well yeah you’re estimating this for that Well here’s you know Bill’s project and here’s you know Jeremy’s project and here’s Sally’s project Um and they did the same sort of thing And you’re like Oh but you know like this is pretty complex It’s more like this project And even that even if it’s not you know telling you the answer it’s giving you like really solid guidance based on what has actually happened in the past to you know actually make good decisions

Marcel Petitpas:

Yeah a hundred percent And the thing that I’ll say is like our tool just like every other tool just like your time tracking tool your project management tool it’s only going to be as helpful as your data structure So you got to start there start with defining the data structures that are going to be able to answer the questions that you need And then go look at tools that map to that Um not the other way around I think that’s the mistake that You know everybody makes in everything that they do as they look to the tool to solve the problem But like the problem has to start with the process and the structure and get that um focused in first And this is like the thing that we spend most of our time uh consulting On at the beginning of an engagement is just like really going deep into data structures and mapping that out Um because without that we can’t really build anything else We can’t build a great process We can’t build great reports and we can’t build greats uh feedback loops with the team because the foundation just isn’t there

Galen Low:

I like that I think you’ve probably answered my next question which is you know if if someone wanted to start doing this uh and start building a culture of data-driven estimation, uh you know what’s the first thing that they ought to do

Marcel Petitpas:

Well they should go and uh opt in for our agency profitability toolkit where I have spreadsheets and frameworks and resources and videos walking you through each step of the flywheel I’ve worksheets you know to help you kind of go through the same audit process that we do with our consulting clients and kind of go through this process on your own So that’s what I would honestly you know it’s a little self-serving but that’s what I would encourage you to do Um you know there’s Five videos that you can kind of go through a watch to get real clarity on how all these pieces fit together And then there’s you we kind of go deep into each section and provide some resources for you to actually implement those things So um but the first step in that is again it’s that estimation Defining that how you do it what the structure looks like and then going and auditing your tools and making sure that that the way you’re structuring your data and their maps um you start there that’s 80% of the upside Um and you’ll start you’ll you’ll find that you’ll be able to answer questions way more easily You’ll have more insight and you’ll it’ll become very clear very quickly where you need to go improve a process or change something to really close the gaps that are probably creating most of the discomfort Most of the unpredictability in the agency Probably comes down to just a handful of things that are going to give you the high return

Galen Low:

That’s very cool And for our members listening I will include a link to that when I post this on our forum Uh and also I’ll put a link as well in the podcast description for folks Um because I know we talked about the flywheel Listen, these are all really valuable insights. I think we had a really good conversation today.

And I think the thing that really resonated with me because it’s been my past, which is that, you know, agency life doesn’t have to be hell. And I think like that really resonated with me that like, yeah, we cut our teeth in this high-pressure environment so that we can go like retire onto client side.

But actually, the goal shouldn’t be that. The goal should be to be more efficient, to have that work-life balance. And when we’re talking about process improvements, it’s not about top-down, you must do this thing, you must follow this checklist. You are a robot now. It’s actually bottom up. Like, actually I want to have time to like actually eat lunch.

Like I have a lunch break, uh, and, and, and, and catch the regular bus and not have to Uber home at night, uh, to see my family, you know, before they go to sleep. And I mean, the adults too. Right. Um, yeah. So I think it’s like, uh, yeah, I think that’s a really good motivator for anybody working in any environment where you have to estimate project work, uh, to just kind of returned back to that sanity and not accept that it’s going to be something that’s going to be crazy and that estimates are never accurate and that there’s always going to be over time. Um, it’s actually not necessarily even that. We can get there. And actually, I’m just reminded of one Catherine question that I have.

So we’ve been talking about sort of, you know, data different estimation. I’m really glad we dug in on the qualitative side, but just for, you know, everybody who’s producing an estimate and agency owners and leadership teams like. How, how accurate can we get using some of these processes? Like, are we talking about, you know, like just getting a bit closer to being accurate or how accurate have you been able to get some of your clients to be at estimating?

Marcel Petitpas:

Yeah. Yeah. There’s a lot of variability around this because of course, it comes down to what type of things are you doing? Like if you’re building custom software products, Yeah. I mean, you, you should just go time and materials if I’m being honest, like those things are just inherently hard to estimate.

Um, that’s the, like one of the very few examples where I would agree that it’s like every project really is different. Um, that’s only because I know enough about software to know that it’s just, it’s ridiculously hard to figure out, like, do I predict, um, but you know, on the other end of that, if you have like highly productized services that are like exactly the same, every single time, we’ve been able to get people into single-digit margins of error.

Um, and most of those come down to like, it’s newer people on the team that aren’t trained or they didn’t follow the SOP. And that’s why there is like a variance here, um, but like, I would say that if you run more of like a digital agency that with a foundation in web design development, your objective should be to get within fairly consistently within a 10% margin of error.

So what you actually track versus what was estimated is within 10%. If you can kind of get there, then that’s plenty good. That’s plenty of reliable enough for you to build all your operations systems on, um, make sure that your people get home consistently on time. Make sure you’re consistently profitable.

If you can start to close in on that 10%, that really should kind of be the objective.

Galen Low:

That’s awesome. That’s amazing. That’s really cool. I know it was so many people who would be like, Oh, I’d be so thrilled if my estimate came anywhere, anywhere close to that. So there you go.

Marcel Petitpas:

And I’m sure you have some, I’m sure you have some, but the problem is it’s not most of them, right?

Some of them are 50% over, under some of them are 30% over, under, and then every once in a while you get one that’s accurate and it’s maybe just a fluke. Um, but yeah, it’s, it’s making that consistency is, uh, is really the objective.

Galen Low:

For sure. Yeah. I love that. Very cool. Awesome, Marcel. Thanks for joining us today, I think it was a really good conversation.

I hope our listeners have learned a lot. Uh, I’m really excited about, uh, just the flywheel in general, the process. I think that’s really interesting for anybody, uh, but I really love the notion of Parakeeto. It’s something that really appeals to me as someone who’s done many, many, many estimates, both on the project management side and on the sales side.

Uh, so I really love what you’re doing. Uh, and thanks again for being here with us on the show.

Marcel Petitpas:

It’s my pleasure. And thanks everyone for tuning in. And I’m looking forward to meeting, um, hopefully lots of you, uh, inside The Digital Project Manager network.

Galen Low:

Awesome for sure. Yeah. And maybe what I’ll do is I’ll, I’ll post this, uh, in the forum.

I’ll get you in there and Marcel, and you can answer questions for folks. We’ll have a little dialogue about it. What do you think?

Marcel Petitpas:

Absolutely. Yeah. I mean, I’m even happy to do an AMA, um, you know, answer questions in the chat and if you’re listening to this, uh, you know, just reach out. Um, I love nerding out about agency stuff in case you haven’t noticed.

It’s one of my favorite things to do. So if you have nerdy questions, if you want to dig into it, uh, yeah, definitely leave a comment, uh, inside the DPM group and I’ll, I’ll chime in.

Galen Low:

Awesome. Very cool. Very cool.

All right, folks. So what do you think? What hacks, tips and tricks do you have for making project data usable for future estimates?

What works for you? What hasn’t worked for you? Tell us a story. Have you ever had data-driven estimate that was just way off or maybe a ballpark estimate that just was spot on? Tell us in the comments below. And if you want to learn more and get ahead in your work, come and join our tribe with DPM membership, head over to thedigitalprojectmanager.com/membership to get access to our experts forum, mastermind mentorship groups, our library of mini-courses, our live mentorship sessions, our ask me anything sessions with a variety of experts. Hopefully Marcel in the future, eBooks templates, and more. And if you liked what you heard today, please subscribe and stay in touch on thedigitalprojectmanager.com. Until next time. Thanks for listening.