In a recent The Tech Trek podcast, Wellsky Chief Technology Officer Joel Dolisy talked about innovation and how different aspects impact your engineering team's organizational design. He covers private equity, venture capital, growth, digital transformation and more.
Full episode can be found here: https://eleva.no/joeld
Below is the transcript:
Amir Bormand (Host): On this episode of the podcast, I have with me Joel Dolisy who is the CTO at WellSky. We're going to be talking about organizational design, and Joel has worked for different type of companies, different size of companies and we're going to cover different aspects of organizational design when it comes to innovation versus transformation, PE versus VC, and just different sizes. Joel, thanks for being on the show.
Joel Dolisy: Hi. Great to be here. Thanks for inviting me. I'm Joel Dolisy. I'm the CTO for WellSky. What is WellSky? WellSky is one of America's largest and most innovative healthcare technology companies. We build a range of very proven software solutions, including EMRs but also a lot of other software solutions, analytics capabilities on top of those services around the software.
We really help healthcare organizations to optimize care for their patients, but also their performance. And we do that across the acute space, the post-acute, and the community care continuum, which basically means that we're serving a really large swath of the healthcare world, basically pretty much everywhere, whether it's in a hospital or when you get discharged from a hospital or from community services. We have some software solutions, some analytics solutions, and some services that we deliver to the market.
Amir Bormand: Absolutely. I know you're a CTO, how broad are your responsibilities? Every CTO's a little bit different.
Joel Dolisy: I think CTOs tend to have various kinds of responsibilities. I'm primarily responsible for delivering our software, which means all the engineering aspects, all the hosting, and cloud aspects of the infrastructure. And the corporate IT part. So keeping the lights on and allowing our teammates around the country here in the US and connectivity with our contractors and what not all around the world. And then as part of that, there's information security, which is a big thing as you can imagine also in healthcare.
And then the last part also is UX, which is sometime a little bit unusual, but we see as an important part of our R&D capabilities, it does report in my organization.
Amir Bormand: I know this episode we're going to be talking about organizational design, and I know you, your background has been experienced at different size companies, different stage companies, VC , private equity, and I know we were talking about how every company's needs will shift based on all these different parameters and no two are probably the same, but I guess let's start off at the top when we're talking about organizational design and just looking at. Just size of company as we're seeing one go through the transition of startup to mid-size, to enterprise, you're looking at all three.
Are there some basic commonalities across organizational design at any size company, or are they all individually driven based on those stages they're at?
Joel Dolisy: Yeah, I think at a high level, you're always trying to maximize collaboration between people. So that's the goal at the end.
You're trying to optimize what a set of people can do together to create bigger and greater things than the sum of the individual parts there. As I look into this and the continuum of companies from startup or pre-product market fit all the way up to super large companies, they're all trying to do that.
And the problems get more and more complex. The more and more people are working inside of the company to try to get this multiplier effect. And if you think about it, then every stage and every kind of growth stage of company will require I think a different set of capabilities in order to maximize that collaboration and that multiplier effect that we're still looking at.
It's very easy when you have four people collaborating and if they can be in the same room, that's even better. But if you think about it and the number of communication channels that need to happen between four people is relatively straight forward. When you move to 10 people that starts to get more complicated.
Then you start splitting that into two teams then you are starting to increase the level of overhead that is needed to actually get there. And that's why to some extent you are trying to find that right balance between what is the bare minimum model of overhead that I need to add to actually continue to get in the collaboration and the flow of communication to work between the different individuals that are part of the team.
Amir Bormand: I guess a question that comes to mind when we're looking at obviously the communication aspect I could see being very key across any size org. When you're looking at an org that is growing and is transitioning from one stage to the other, how much of that identification is proactive, meaning we're anticipating it, we're going to start putting changes in place or is it, hey, when we hit the bottleneck, we address it because we just might not know.
Joel Dolisy: Yeah, I think it's one of those things where you would always like to be able to predict and work and anticipate. I think that's what you're trying to do as much as possible as a manager and as a leader, you're always trying to anticipate what the next move is.
That's what I do. That's my personality overall is to always try to identify what is behind the next corner and what will we need to actually be able to handle and mitigate any risk that's coming with that. Depending, overall, it does work. But they are always surprised they are coming because you don't control the external factors that are hitting you.
Whether it's competitors that are coming in or regulation frameworks that are changing that you didn't anticipate, for instance, six months ago very few people talked about ChatGPT. A few people knew about the ChatGPT library in kind of the models, but nobody was at the level that we are right now.
I know everybody looks at that and what is the impact on businesses. So I think that you always try to look ahead. I think that's what leaders do. Overall, you celebrate the moment, but you can never be happy that something else is coming and there's a little bit of that paranoia about looking forward.
And try to anticipate those changes and so you have to really balance those two things in those two aspects of the job and sometime most of the time, hopefully you'll get it right and sometime, you'll still have to adapt. But, to some extent, I really like this analogy to some extent of what you're trying to do in order to do all of those things is you're trying to implement an operating system for your company.
And the need for an over green system for a small company is way different than the need of an operating system for a larger company. And so if you think about it that way, then you're trying to optimize different things at different stages of companies.
Amir Bormand: What happens, I guess as I understand you, you can only be so forward thinking. I mean it's kind of hard. You brought up the AI component of that everyone's evaluating now and obviously who would've seen that last year this time? For most, that is when you're starting to look at. Companies that are dealing with innovation, right? They're building software versus companies that are trying to leverage software and move their company along.
So they're not software development type companies. They both approach team design, organizational design very differently. Are there any similarities, any differences that, that you've seen in the past?
Joel Dolisy: Yeah, I think Marc Andreessen who said, a long time, 10 years or 15 years ago, hey, all companies are software companies and software is eating the world.
I think since then pretty much everybody has said “Okay.” I think we indeed, whether you are an internet native company or you are to what used to be called ISV company in the early 2000 that is producing software as its output basically.
Or you are just a “brick and mortar” type of company and need software to actually run your business nowadays and it's an integral part of your success. I think all of those have gone and are going to transformation and innovation. I think there is still a difference.
I think in terms of true innovation and transformation, to some extent I don't like transformation too much because I think it is something that you end up doing all the time anyway. If you want to stay current and to some extent also protect your core business, you have to continue to reinvent yourself and its transformation to some extent.
So it's a never-ending process. There's this sweet spot that you find where companies are. Really great companies have a way to look at the different horizons in front of them and allowing funding for those type of horizons, which allows them to continue to innovate along the line and not just resting on their existing success.
But innovation also takes a lot of different forms. I think people talk about, okay it's a great product that you come up with innovation, but there's also a lot of innovation that companies have to do in terms of getting more efficient and more effective. Because it is expected as you grow to some extent that you're going to get more economies of scale, because you're going to try to do the same work that you're doing and maintaining the products that you're creating by being more efficient and requiring less people to do that so that you can devote people to the new things that will generate new revenue stream for companies.
And so that's why there's a lot of innovation that needs to happen at that level also because even if a product has been built 10 years ago, you're still trying to how can you leverage the new best practices, the new technologies like AI for instance in order to give you disadvantage that allows you to spend less time maintaining those software and put more of your people on the new things that are coming in.
Amir Bormand: That's actually a hundred percent. When I hear digital transformation, and obviously comparing contrast with software companies, sometimes you sit back and go, software companies are going through the same thing as you mentioned to keep up. They can't build what they build and then figure the new features are just going to be antiquated or they're going to not stay current with the latest technologies. I guess the one thing you mentioned there about maybe software being viewed as, where they sit within the cost versus potentially becoming a revenue center. Is that something that you see how, you know, based on that alone, do people start viewing how to design their technology a little bit differently?
Joel Dolisy: Yeah. I think so. I think that's a key observation that you mentioned there it is. I think companies that just see software as a cost center, I don't think they will ever be great innovators. They will ever be taking advantage of what the software world can offer.
If they see software as an enabler of their growth and an enabler of basically what they can achieve, that takes a very different view at that point because it's going to impact a lot of the way they think about funding those efforts and what not. And not that you don't need to be profitable and what not.
All of those things matters when you're trying to build and run the business but there's a key differentiation about looking purely at this as an extension of it, which is a cost center versus us, okay, this is where our growth is going to come from, and it's an enabler for our growth and it's core enabler for our growth.
Amir Bormand: You mentioned something like ChatGPT, that obviously I know there's alternatives out there too, but that was first to markets. Everyone's talking about ChatGPT, obviously you mentioned we talk about different stages. You can't talk about external forces.
As a leader, obviously you have to evolve your team, your org, but as you see these developments coming, how do you evaluate which ones you have to? I don't want to say take seriously, but which ones do you go, I'm going to move on, or I'm going to direct energy and resources too because obviously that'll shift priorities and it might not have been on the roadmap six months ago even.
Joel Dolisy: Yeah, it's, I think it's more of an art than a science, honestly. I don't think that there's any silver bullet that you can apply. I think you look at the potential of change and the potential of an impact that site of technologies or technology or innovation can actually have.
And what's the scope of that impact? And when you start looking at legislators, a new web UI framework comes out. At this point, it's a relatively mature market at that level. Do you need to jump on the latest greatest Web UI framework? Probably not.
You can probably still go in with react and some of the things that you're using, but you're looking at this and those large language models, whether they are ChatGPT or Bard or whatever, I think we are just starting to see, and I don't think that we can even see what the true potential of those disruptor technologies are.
And so I look at this and I feel I've seen people saying it looked like we are back when the internet came out and a lot of people knew or forecasted that there was a ton of potential in the mid-nineties but nobody could see that everything that changed around that.
And I think we are like that with those kind of large language models and what they are able to do and the foundational models and the transformers that you have on top. I think that's when I see that type of stuff just a little bit like mobile, a little bit like cloud computing and what not.
All of those are just massive waves and you just start to see that, and you really have to pay attention to that. It doesn't necessarily mean that you're going to be able to make a use of that directly into a product of yours, but you look at some of the capabilities that GitHub is coming up with, or even Google is showing.
I think even Amazon has their own kind of developer productivity tools and some of those things are really just transformational. Like the fact that you can start getting something to write boilerplate code for you intelligently. It's just unbelievably powerful and that's what has been out for a year now with GitHub co-pilot.
They've got a new set of capabilities that are coming out with some of their extensions. And you look at that and you think that is something that if we don't pay attention to this, we are going to be outran. Because somebody else will take advantage of the good parts and come up on the market with solutions that basically come without all the baggage that you necessarily have as an established company and they will create disruption.
So that's what we're always trying to weigh, the pros and cons, the return that you may get on that and the risk of doing nothing. What I look at is what is the level of destruction that those kinds of waves will have.
When it's at the magnitude that it is, right? No, you can't ignore that. And the thing also that is very different compared to even, you look at the way that started with the internet or the mobile stuff, the time for adoption for those type of technologies is compressing dramatically. While in the mid-nineties you may have had 10 years to figure out what to do with the internet as a commercial solution. Nowadays, you don't have 10 years to figure that out because the rate of innovation and change on top of those technologies keeps going and accelerating.
You still have to balance that and really make sure that you are not doing a knee-jerk reaction and just jumping on the bandwagon for hype. On the other side, you can't just say, oh, I'm going to wait for, I'm going to wait for the adoption curve to go and hide, because you may just be left behind and it might be too late at that point.
Amir Bormand: I think we were going to touch on timing impacts. And think this talking about ChatGPT, Google, Co-Pilot, all these different tools and I see the current bubbling and I think back to two years ago and people were trying to implement Bitcoin transaction in every application.
Tip me with Bitcoin. It's okay, hey, that's going to be a thing. Hey, it's popular. And you start seeing that and you're like, you see a lot of companies integrating ChatGPT, to immediately rewrite your words. And sometimes I guess when we're talking about organizational design and timing, being early is very good because everyone's looking for a differentiator of something for first mover.
But being early also means committing resources at a higher risk. And I guess as you're looking at just the org overall and you have to make those decisions, not whether you implement, let's say, these types of tools, but just in general, being early to take advantage versus taking the risk is again, art and science seems more art than science to me.
Joel Dolisy: Yeah. Obviously, that's where you need to go back to some type of framework that you have a core business, and if you go with the McKinsey’s “Three Horizons” kind of management framework, your core is still horizon one. That's what's paying the bills today. The whole question after that is how you need to spend any type of incremental dollar or move dollars from the core business to spend on your horizon two, which is what you may have in the pipeline and it's starting to grow, may not have hit escape velocity just yet, but you're really close to.
And then you've got horizon three, which is more speculative. You just need a framework of understanding, okay, here's how the company is actually going to move funds between those different kind of horizons, so that the teams that are working on horizon one and the core business don't feel like, they don't get to play with the shiny object, but you still count on them to some extent to continue to adopt some of those technologies just as a set of tools and toolbox for them.
I think those frameworks are really helpful in terms of, first of all, aligning internally about, okay, here's how we're going to spend our dollars, and what is the responsibility and what are we expecting from the teams? And how do you structure the teams for each of those efforts? And it does help you to use a reference framework there.
Amir Bormand: Hopefully most people, when they see these technologies, are taking that kind of pragmatic view of what to implement. The excitement sometimes does bubble and swell, and you know you have to deliver, right?
I think end of the day, business, whether you're a software company, you produce widgets, whatever it is, and I know with your background you've seen different drivers of whether it's been private equity or VC, and the impact that has on how people go about building the tech org. I mean maybe you could just highlight a little of those differences for everyone.
Joel Dolisy: Yeah, sure. Early in my career I worked for more VC backed companies where venture capital typically is more about trying to get to that hyper growth stage, scale up model or even pre-revenue and really putting a lot of gasoline on the fire for growth.
Depending on the period that you are in, sometimes profitability was not such a high importance. Basically for them it’s grow up any way possible, get the highest growth rate possible. We'll focus on profitability later. The latest part of, probably, the past 10 years of my career has been more centered around private equity and growth equity. At that level that has a more balanced approach to growth where profitability is always super important for those companies, and for those investment firms and their limited partners. And at that level, that means that you need to apply different techniques and you have different priorities when you are funding different efforts around engineering and product development and what not.
One of the key things that is super important to understand in particular with private equity is where are you at the holding period? When private equity firms make an investment into a company to some extent, the clocks start ticking. And on average private company tend to hold the investment for five to seven years. For companies in their portfolio, there are exceptions to that. There are some that are holding longer, but on average that's about the holding period there. That’s going to be very different as you progress to the holding period, there's going to be a very different type of profile of investment that you're going to have to react to, because the further you go inside of the holding period of a venture capital of a PE firm holding period, the more growth on the profitability becomes important. We still want revenue growth, but you want to see an acceleration of the growth of profitability, basically, and EBITDA margins in general. So at that level you have to adapt new investment strategies and the planning that you do, to some extent, when the holding period starts all the way through the five to seven years, you have to think and have a plan. During the first maybe two years, you can have a lot more liberties and freedom in terms of doing things that are accelerating growth, and putting in place a lot of the changes that are putting dividends and producing dividends in terms of efficiencies and what not down the line. You have to really keep that in the back of your mind because if you don't and you're trying to apply the same tactics and strategies in year two or four, the holding period as you were trying into year one, you're going to have a lot of friction and a lot of challenges in front of you to actually get up, get that done, and be successful.
Amir Bormand: When you're hiring leaders — I guess this maybe staying with private equity versus VC backed companies —when you're hiring engineering leaders to come in to execute the plan, and we you just outlined there are different drivers. Does it impact the type of leader that you bring in?
Joel Dolisy: I think it does. I wouldn't say necessarily the type of leaders. I think you're still looking for people who are, to some extent, great operators and people that are superb in executing and being a great fit to the company. Though I think those two things don't necessarily change between the two. But I think that what I've come to realize over the years is that I think a lot of the leaders that you may have in your organization are focusing on solving problems on a day-to-day basis. Here's why some of the decisions that are impacting them are being taken. Here's why some of the constraints that are impacting them. And I think there what is super important is the context and sharing that context and repeating it often and over because those things tend to just go back in the background, because, okay, we are dealing with the problem the usual of the week, focusing on delivering this set of capabilities to clients. We have escalations, we have production issues that you have to work with and what not. It's easy to lose track of the forest, because you're so focused on the trees that you have in front of you there.
And that's what I think my job, the job of my direct team, it’s really giving that context, ensuring that people understand that there's a good reason of why we're doing that, and there's a lot of benefit of us following those strategies that are allowing us to meet those goals.
And yet it's different that maybe we would've done it two years ago but giving them the context of why that is. This is something that I think I've certainly underestimated earlier and I've not doing and we're doing a lot more around that and see that's probably the most dramatic change that I've done over the past 10 years is really focusing on the context there.
Amir Bormand: Final question for you. You know whether a company is a true software company versus non, but everyone does software. Now we talked about private equity VC, we talked about different size companies, and I guess the last question was about leadership.
I want to ask a follow up regarding leadership. As the company transitions, it could be any state to any state. It seems to me that leadership is difficult to plan for because as your company is growing, and you're growing in size, growing complexity, you need different leaders at different stages. And it's not at the highest level, not talking about succession of a CTO but the actual like hey, the org grown from 20 people to 50, and being able to manage that like 20 to 30, or 50 to a hundred. That seems to be as when you're looking at the organizational design and the shift that happens and who's responsible for the smaller teams.
That seems to be in the vein of what we're talking about. How do you view that?
Joel Dolisy: Yeah, I think there's a great book that came out several years ago called Team Topologies, help summarize a lot of thoughts that people had around, what are some of the successful ways of thinking about team design, and how do you assign responsibilities to team.
Do you think about teams that are working on features and product features? Do you think about teams that are working more on platform and shared services? You think about enablement teams that are helping the teams adopt some key technologies for instance. To some extent what you really try to get at is a model that can scale.
You need to have clear ideas about your design principles for your org, and what you don't necessarily want to create is a skeleton ahead of time in filling it. I think you have to build the skeleton as you are filling it, but you need to have a set of design principles that you're trying to follow that allow you to know, when do I need to add a manager? When do I need to add overhead or oversight to a set of teams that are working on that. And that's why it's important to have. Our teams will be around eight people. We don't want teams lower than five people for instance because they're not teams. You don't get the acceleration that you need to and the scale that you need and codifying, and it's going to vary from one organization to another.
But going to that process of identifying what are your design principles for your organization, how do you want it to look two years from now? And what is going to allow you to get to that point? You need to have those starts and you need to write them down and share that with the rest of the organization, whether it's engineering but also finance and the rest of the executive team there so that they know that there's a process behind what may seem sometime a chaotic kind of way, and then you're going to have to adjust because sometime you have downturns and you're going to have to contract. And so how do you contract those, that organization that you may have thought about? Okay, I was planning to actually get to a hundred people, but I only get to 90. Do you still need all of these oversight layers and being able to react to that. But that's why I think it's super important to have those kind of design principles about what are you trying, what is going to guide your choice, basically, and guiding your decisions as you are building it.
Amir Bormand: Absolutely. I like that and I could continue with talking about contraction and the impact. We might have to save that for a different day. I gotta let you get back to your day job as well. But I was going to say thank you for coming on.
Thank you for sharing. Definitely have to have you on again and expand on some of these other side topics that I think we could talk about. If someone does want to reach out to you and touch base on anything you said on the show what is a good way of getting a hold of you?
Joel Dolisy: On LinkedIn or you can reach me on my work email, joel.dolisy@wellsky.com.
Amir Bormand: Yeah. Okay. We'll include your LinkedIn on the show notes. If somebody does want to reach out to you, they will. But Joel, thanks for coming on. Thanks for sharing. I really appreciate it.
Joel Dolisy: All right, Amir. Thank you very much.
