How Data Teams Scale Project Management Without Slowing Down

0:00On this episode of the show, I have with me, Cam Crow. He's director, data, and analytics at vacation. And we're going to be talking about formalizing project management frameworks. And we're going to talk about that in the context if this data background. We're going to talk about being in a startup. Obviously, I have to optimize for a few things there that might be different if you were to enterprise. And how do you layer in project management and choosing the right framework and got to bring the stakeholders along as well and have a couple of things in there to chat with Cam? Thanks for taking the time, Cam. Thank you. Great to be here. Absolutely. OK. Vacacia, what do you guys do there? Vacacia is a time share resort property management company. So we do a whole range of things, kind of ever evolving. But kind of the bread and butter for us is managing time share resorts. Everything they need to do from an operational standpoint, helping owners come and stay for their week, exchange their week, welcome exchange guests,

1:03and then we also run rental management for these resorts. So all about helping these resorts be healthy and strong and generate revenues so that they can continue to develop their resort and improve their owner and guest experiences. Absolutely. Very cool. All right, this episode, we'll be talking about project management frameworks. I guess I mean, whether you're working at a startup, you're working at a big company. Obviously, project management is always a factor. When you think about startup environments and project management, I mean, obviously we know companies want to be fast and nimble. But talk to us a little bit about if you are at a startup and you don't see project management, project management framework rather, where do you start? What do you start doing to look at that aspect of the business? Yeah. So my tenure at vacation has kind of really

2:03spanned the startup, scale up, enterprise, bridge, golf transition, whatever you want to call it. But like, so I've kind of been in all key phases where from a project management standpoint, I think yeah, like you said, it's common to not really have any kind of formal project management working at startup. And actually, like, I think that makes the most sense. So I'm like, my background is startups. And I've been very reluctant to do a kind of like adopt like sprint planning approaches in that startup background. My belief is you should wait until like you should add process very conservatively only as you need it, not based on some kind of like philosophical like rationale. Because when you're going to start up, you need to be nimble

3:06and you need to be fast. And like, we optimized it for that. And we wanted to kind of like keep pace with the business. So what they needed, like we wanted to like be a rapid response, delivering value to them as quickly as we possibly could and like keeping them moving forward as quickly as they could. So there are definitely challenges with that. But I think when you're small enough and the number of stakeholders you have is small enough, like that can work just fine. And that was actually kind of like how I've often preferred it. So that's how we did it in startup phase. And I don't think like going back and doing it all again, I would really change that probably actually. But we could go into like some of these different phases, but when we had a period of massive growth where there was an acquisition, migrations, merger,

4:09and all of a sudden there's just like exponential growth in the number of like stakeholders we needed to service and the number of like dependencies and use cases that were dependent on us. And over time, like we didn't change our process for a while. And we were just stretching, stretching, stretching, stretching, and it got, it became unbearable. And so we really had to change. And that's I think I wish we like started that process a little bit sooner. But probably I don't think like I don't necessarily wish that we like added process like at the early, earlier startup stages because it didn't, it wasn't really fully needed then. Absolutely. Yeah, I guess as you're talking through that in the obviously you've seen various phases of growth. And you've been adjusting to those phases. I always find that a little interesting when we're going from, hey, we're at this startup and we're really nimble to now, we're growing.

5:12And we need to actually give visibility. I can't be involved in everything. It's almost the whole team has to go through that and accept that change because they're used to working in a certain way. I don't know if you remember this, but when you think back that transition period of going, hey, we actually need to get a handle on things because we can't just remember everything. We just can't be on top of everything as you said, requests are increasing. That initial implementation of a framework, do you still think about right sizing that or are you now going, you know what, the company's growing. We need to actually maybe think a step past where we're at and might slow us down, but might help us. How do you deal with that transition? Yeah, exactly. So yeah, we've got like period of constant stretching. We're all kind of multitasking. That's like in a startup. I mean, pretty much the norm everywhere.

6:14So it's like maybe you've got like three balls in there, four balls in there. And then all of a sudden, you have 10 balls in there. And then you can't catch them all. And this is like a, I read about a metaphor, and I can't remember exactly who said it. But the trick is figuring out which balls are rubber and can bounce and which of them are glass and will shatter. You got to catch all the glass ones. And like you can let some of the rubber ones bounce, like follow because they bounce. But I think when you get in that point where you've got so much you're juggling, the effect is that sooner or later your speed and your quality is going to decline. Because you just can't keep that many things. You can't keep track of that many things. And even if you could, even if you're not going to drop a single ball, but you're spreading your attention over 10 balls at once,

7:16none of them are going to move forward very fast. And none of them are going to get the highest level of quality that you can deliver. Because you are just like going as fast as you possibly can. And at that point, nobody's happy. So the stakeholders start to get frustrated that they can't really get what they need. The team starts to get frustrated, because they know they're a lot of the team is happy to the extent that their stakeholders are happy. And they know they're doing a good job and bring impact to the business. And at that point, it starts to get dangerous. And if you stay in that mode for very long, I think you don't like there's no equilibrium. You're getting better, you're getting worse. So as you're getting worse, you start to kind of kind of start to see the brink. You see the edge of the cliff, and you

8:18need to fix it before you get there. I'm like, edge back away from it into something more sustainable. And like you said, you need to get the team on board. If you wait till the point where, I think if you try to do this too soon, your team might be like, do we really need to do this? Do I need to add all this overhead and this process? And I was just getting in the way. But at some point, they're just like, I can't breathe. Like we need to do something. And it's very easy to get everybody on board. So you don't want to wait too long, but if you do wait, it becomes easier from getting the team on board. I think what the more challenging thing is getting the stakeholders and your leadership on board, because sometimes I think they can be worried thinking about going from not having sprints or development cycles to implementing that.

9:20And it might be the first time where your supply has been limited, where it's not just like you ask for more. And we'll just try to give you more. Where there are some limits on supply, because we can only do what we can do in this period of time. Therefore, we now have to talk about what's the highest priority and where do we allocate our limited resources. And I think the key of getting the stakeholders on board is convincing them that we are going to improve both velocity and quality on the highest priority work. By trying to do all of this, we're not able to satisfy what you need from us. And what we can guarantee is that we're going to move the needle on what's most important. But we can't. We have far more requests than we can do in a period of time. So we need to focus. So that, getting them on board, I think, is the hard part.

10:23And you really need to have an executive sponsor. You need to have somebody in leadership that is fully behind you to chart that transition. If you're going to like, if you haven't had it all, to like, you're putting something in place, I think without that, the deck would be stacked against you. It's actually some fantastic advice. And you mentioned the stakeholders. And you mentioned obviously sometimes as we see companies scale and process and size. Proximity is interesting because they're used to knocking on Cam's virtual door and they have, I need this. And then all of a sudden, there's a team. All of a sudden, oh, we just can't get that in email. I need to have something more, right? That proximity sometimes is actually to me one of the harder change management aspects. I think we all have, we all feel this in different ways.

11:25I don't think it's just his work. Sometimes you're your favorite restaurant. The chef's there, the owner's there because it's small and then gets bigger. And you're like, oh, I never see that guy anymore. And there's 1,000 things like that. But proximity, when you look at stakeholders and that proximity, I guess, talk to us a little bit how you handle that. Because that, to me, is a big one of setting change management policy in place. Yes. Yeah, in the startup kind of mode, there's often like maybe one or two people that do the function. So it is very common for people just like send a direct message and slack. Like, hey, I've got this thing, let's talk. And be like, cool. You know, like I'm free right now. And then you like talk and they can play in. And you go from there. When you suddenly have like 20 people trying

12:25to like have that kind of interaction with you, it just doesn't work. Or if you try to do it, then you'll be wall-to-wall meetings. And you can't actually do anything. So there's, I think, a couple of things we did about to help with this is, one, we vary, I would say very aggressively, but like in a friendly way as much as possible, try to like outlaw direct messages. So we basically like need to get people working in channels. So that, because if you like just direct message me, you know, like I might not see it, I might not see it for half a day. And then like, you know, at that point, I've got so much to get like piled up to respond to. Like it's not a good experience for a stakeholder. And like there's no reason for that bottlenecking

13:25to be going on. So like we started to organize our stakeholders into what we can call domains. Each domain has a public slacking also. And we call these data talk in the domain name. So it's oriented around like a public kind of communication lobby, if you will, for like the data team that's supporting that domain and the stakeholders bring that domain. And we try to like really cultivate a culture of like talking out in the open. And the value of that is that like people, you might like, they'll be instant responses if it's like in a public channel. It doesn't need to be me. That's like the first person to respond. It could be anybody on my team. It could be other stakeholders. If you if you talk that way, it might actually be like another stakeholder in the same domain has the answer for another stakeholder

14:26before like anybody on my team even like noticed. And like that's great. It's just like decreasing the time and friction to get to any answer or a solution. So public communication is key. That also really helps with accountability by the way. When like more eyes are on something then people feel a lot more motivated to respond quickly and to you know, hold up their end. And then a little bit more about the domain. So one thing that we did with our kind of approach or our design for project management or development cycles, we used the tool linear and they call them development cycles. We put a lot of thought obviously into kind of our routines and ceremonies within the team. Like how are we gonna do our work? But equally important is like, what are the routines we want our stakeholders to do

15:29to like kind of meet us halfway? So and this I think was a big part of the strategy of getting those stakeholders on board with us attempting this is that they needed to really believe that they were gonna have access and input and we would like truly address their highest priorities. So we actually put a lot of emphasis on stakeholder engagement as part of this process. So we often we chose a two week a pretty standard like two week development cycle length because one week's too short and it's like too much like overhead to plan the cycle that's like only one week and three weeks felt like too long. That's like too long for like stakeholders to kind of like wait to like refresh the plan. So two weeks felt like the sweet spot and really like the max we could kind of get stakeholders to buy into and then but instead of going two weeks

16:32done start the next two weeks, we planned a one week in between which linear calls a cool down week but we have rebranded internally it's anything but a cool down. It's the planning week and which means it's the most chaotic week out of like a three week cycle. So we've got two weeks on cycle, one week off cycle but that one week off cycle is when we convene all the stakeholders. Every domain gets convened in that planning week and it's so there are typically 45 minute meetings we invite every stakeholder from that domain and we give everybody an opportunity to say like what is most important right now and we write it all down and then that and there's I invite everybody on my team that works in that domain and this I think kind of unique when I've talked of peer kind of data managers that most leaders don't like bring their team

17:33into all those stakeholder communications they kind of try to shield their team from that like discussion, deliberation but to me like it's critically important to bring them into the indirect stakeholder communication because they're gonna be way better at connecting the dots on the work they're doing if they hear it straight from the stakeholders mouth and they can ask a follow-up question. Do you mean this when you say that? And that also builds relationships and helps to build kind of trust and like credibility kind of across stakeholders and team members. So that rapport is important. So we take that week and it's intense. Like the I like encourage the team to really like when you're in the cycle like you can kind of be like live under a rock like I'm like like focus on your projects I'll be the triage manager when new curve balls come in

18:36and like we got to figure out what we got to do about something that wasn't part of the plan. Let me figure that out and I will let you know if we need you to like pivot or like bring something into your cycle but like you don't need to be scanning all the slack channels all the time like checking your email every hour like you don't need to like be responding to these things coming in. Let me do that. You focus I'll let you know if you need to if you need to shift at all. So so like they get a focus but then in this planning week they get engaged and it's the whole like the point is like come out from under your rock like you know like update your radar like see what's going on connect with people you know like really the focus is on communications and getting up to speed with everything. And then based on that like stakeholder check-in meeting that's the basis of when each team like plans there makes their plan for what they're gonna work on

19:37the next two weeks addressing the highest priorities from the stakeholder check-ins. I love that. I guess when you look at projects within the data space obviously there is a notion of needing yeah needing what you need from the stakeholders and obviously from a data perspective you might be able to include something up you know a short-term when obviously long-term might not be so portable we're starting to see this shift into you know is code something we should hold on to with all the coding agents out there that are helping teams and seeing it even now touch some of the data side and even some analytic side. When you start looking at the landscape and when you start thinking about your team if you were to put on your I guess your future goggles

20:37of this notion that AI is starting to touch all different aspects of technology are you starting to even look at that within because obviously you've put in a PM framework you handle the requests you're understanding that everyone wants to go faster these tools potentially are an option is that something that's showing up on your radar is it's still one of those things where listen we're actually executing and don't necessarily need to make that a part of the core offering it. Yeah, I am thinking about it and I think I'm thinking about it in like two ways, AI in about two ways. So there's the kind of value delivery process so like your workflow are you using AI agents to do things faster, better, you know than kind of bespoke, you know, coding by hand

21:40you know, like custom code. That's one aspect of it. The other aspect of it is enabling AI, agentic AI experiences from data and that is, so that's not really about like the data workflow, you know, delivering faster but better. It's really about building the tool that the business can basically get insights from on demand. So that is that kind of area is some people call it the semantic layer. It's the concept that you are, we're already engineering the data models, the data sets that like we build reporting off of, we build system integrations off of, we're crafting those data sets,

22:42but now we don't stop there, we need to craft the context and the definitions on top of that data set. So you build a great data set and then you build the instructions that make it clear like what's what and how it interacts with other data assets. And then an AI agent knows what it's looking at because that's the fear is that you're going to get totally bogus answers if you just like point an AI agent and we've done some proof of concepts that have kind of born this out. Like if you just point it at your data set as it exists in your data warehouse, it's going to make some assumptions, it's going to do the best it can, but it's going to kind of like take some liberties on like what are some of these business definitions because the business definitions don't like inherently live on that data set if you haven't built the semantic layer.

23:43The semantic layer is like a configuration infrastructure that an AI can see, oh, when you say revenue or net revenue or gross revenue, or that means this field filtered by these columns and it joins to this table in this way. So you can actually, if you do that in AI agent can arrive at like correct answers, the like a type of answer that like a senior analyst would deliver to the business, but it might take them like two days to get around to it because they're juggling a bunch of other requests like that whereas like an AI analyst could give it to you in 30 seconds. So I think that's where data and analytics, it's going to fundamentally shift

24:44to speed to insight. If we build sufficient infrastructures that AI agents can look at the data the right way because they're not just going to know all the business logic just from looking at the data. So that's I think like where the big opportunity is can also almost be as if like every business stakeholder has their like analyst, you know, like sitting right next to them and answering their questions all the time and that. So it won't necessarily be like direct services provided by, you know, a human in that regard, but there is going to be a need for like the humans here because you need, it's an infrastructure solution that enables that. So there's always going to be corners of your business that you don't have that in place yet or there's a new product or a new feature and like you don't have that coverage on on,

25:46like you might start, you might get to the point where you've got 50% coverage on your data, data sets. And that might give you like 75% of the value you need from like an insight standpoint from like what stakeholders are used to asking. But it's an ever shifting landscape and you're going to need to like get more coverage and the coverage is going to change and the logic's going to change. So it's going to be more like infrastructure-oriented analysis and engineering to support those kind of capabilities. So we're thinking about both right now and to a large extent like the focus, so we started our development cycle strategy, we rolled it out in November. So we're now finishing up cycle six, going into cycle seven and we've evolved and iterated on each cycle

26:49to the point now where like there's a whole bunch of like new opportunities that weren't even visible like in cycle one, two, and three. The like like having this approach is kind of like a platform for evolution and experimentation. Like it gives us the ability to evolve in much like higher ways that we never could before. But my focus with the team has been get us into like sustainable working routines so that we have like, we're set up for success. And now I think we need to shift our focus a little bit more towards having a capacity budget for strategic work. Strategic work could be evolving how the team works together. Sometimes we're actually now bringing people outside of our team into development cycles and working in concert with our team hand in hand in these domains.

27:49That's new and exciting. I think that's like strategic work. Like take some time to like integrate somebody new and help them kind of find their way and like build like team building with that. There's also and then these other two AI things are like the top two things I'm thinking about from a strategic standpoint that we need to work into are like capacity allocations is. Both the workflow, like how do we develop faster and more completely and consistently and like avoid some of the basic errors and how do we build the infrastructure to enable AI. And maybe the workflow AI helps us get faster to the like, you know the semantic layer coverage that enables AI for the business stakeholders. So that's kind of how I'm thinking about that right now. I love it. You guys are, I mean, you've seen the company go through its growth cycles and looking at new opportunities with emerging technologies.

28:52You've seen it all love it. Cam, I want to thank you for taking the time to share insights with us. I'm sure somebody might have a question that I did not remember to ask here and they may want to reach out and ask you follow up. Is there a good way connecting with you? Yeah, LinkedIn is perfect. I'm in there as Cameron Crowe, parentheses cam. Just shoot me a message or a connection request. I'm watching that every day. So I should be able to get back to people pretty, pretty quick. Fantastic. I do appreciate the time. Thank you. Yeah, thank you. Absolutely. All right, that's it for this episode. Be back again, different guests, different topic. Until then, two things. One, Cam took us through, well, Project Management Framework, especially, he's seen this growth from a start of phase to definitely company that's at some acquisitions and some growth and how he's managed implementing a Project Management Framework that has support the company. And it's very stages of growth as well as some of his insights on AI

29:55and where it sits within data and analytics. So please share this episode with somebody else who might benefit from it. Also, like, subscribe, comment. Let me know how the show's going for you. Until next time, thank you. And goodbye.