Back

• Business Models

Decentralized structures: a competitive advantage | EP31

March 13, 2019 Print

---

---

Your Host

Rob CampbellCFA

Institutional Portfolio Manager

Rob joined Mawer in 2016 and brought with him a near-obsessive passion for nicely formatted spreadsheets. He is fervent about the importance of taking the long-term view in investing; living a life of “non sibi,” Latin for “not oneself”; and putting clients’ interests first.

When not at the office, Rob is binge listening to all kinds of podcasts—investing and otherwise—and sharing in the interests of his three young children, which include trucks, trains, and reading the same stories over and over. Good thing he is curious.

Transcript

Download a PDF copy

Rob Campbell:	00:23	I'm Rob Campbell filling in for Cam this week, and on today's episode, equity analyst John Wilson joins me to talk about the competitive advantages of a decentralized structure. We get into companies where we are seeing this work, and the keys to making it a success.
Rob Campbell:	00:53	Well, John, welcome back to the podcast.
John Wilson:	00:55	Thanks.
Rob Campbell:	00:55	Thanks for being here! Maybe just before we dive into the topic today, a little bit more on your background. You joined Mawer in 2012?
John Wilson:	01:03	That's right—2012.
Rob Campbell:	01:04	So, maybe just a reminder for folks who are listening, how does it work when you join Mawer as an analyst? And more specifically, how did you wind up on the global small cap team?
John Wilson:	01:12	Sure, so the way it works at Mawer is you have this rotation program. When you join, you'll actually rotate through all the different portfolios. We have a common investment philosophy, obviously: wealth-creating companies, run by capable and honest management teams, and purchased at a discount intrinsic value. But you find that, depending on the geography that you're in, the way you actually go about executing on that is very different. So rotating through is very helpful as it allows you to see how the philosophy is applied in different manners.
Rob Campbell:	01:39	I would think it's not just the different geographies or asset classes, but it's also probably the people. So there's this common philosophy, but every individual on the team might have some nuances in terms of their approach. Is that fair?
John Wilson:	01:51	Absolutely. And I think the other thing that you find—different asset classes have different opportunity sets. So if I look at global small cap, our investable universe is over 20 thousand companies. And I was actually talking about this with Karan the other day (the other analyst on the fund), and his comment was just—we were just talking about how we have these opportunities to find these companies—it's almost like a treasure hunt where you're walking along the beach and you find some gold. And it's actually really exciting. Our bodies have a similar (almost) reaction to finding treasure. My heart starts beating quicker, my hands are sweaty. It's an amazing feeling. So that's one of the aspects on why I wanted to go to global small cap—just given the size of the universe, you're going to find new companies that you haven't seen before. Whether they're recently listed, or whether they're just in areas that you haven't looked before. It's always new.
Rob Campbell:	02:42	I would imagine, as well, (tell me if this is true or not), but I would think that it's also a function of where the team needs some help, too.
John Wilson:	02:48	Oh, absolutely. I talked about from my preference standpoint, but there's also a matching up on the other side. Where are the gaps on the team? Where do we need to add resources too? So absolutely the second point as well.
Rob Campbell:	03:03	Okay, so maybe let's dive into what we're here to talk about today, which is this concept of “decentralized structures.” And before we get into what makes it work, where you've seen it applied, and how you think about it, can you tell us what that actually means just from a definition standpoint? What is a decentralized structure?
John Wilson:	03:16	So the way I define a decentralized structure is, pushing decision-making down to lower levels in the organization. And you can really see that manifest itself in two different types of decision-making. The ones around operations: so, how can you run the plant better? That's probably the most common one I see. There's also [some] around how to create a better product for customers. So if you're closer to the customers, you understand it better. And even it can go as far as customer service. With some of these luxury hotel chains [for example], you find that the people that actually deal with the customers have a lot of autonomy in serving them. Let's say you go to a luxury hotel, and you get there, and the room's not clean or it's not ready. If the customer service rep had to go and get approval from headquarters every time there was an issue like that, they'd have really poor customer service. So that's another example of where decision-making is pushed down from an operational standpoint.
John Wilson:	04:13	The other angle that you see it in, is what we call “capital allocation.” So, [to recap], in our philosophy, we invest in wealth-creating companies. That often means that there's excess cashflow that those businesses throw off, and the decision on where to put that cashflow—whether [it’s] to reinvest in the business, or to do M&A—those decisions are also pushed down to the lower levels in the organization.
		And maybe the one caveat I would say about this, is that it's a spectrum, it's not black or white. You can't say, “this company is decentralized, and this company is centralized.” Rather, they exist along a spectrum. And even within the same company, there are different activities that a company does.
		Think about a brewer…it's not just brewing, they actually have marketing, they actually have distribution. So the level of decentralization might actually vary depending on the activity. They might have very decentralized marketing. Marketing might be done at the national level or at the local level, but production is done centrally. So even within the business, it can differ on how much it's decentralized or centralized.
Rob Campbell:	05:15	Okay, so I'm trying to think of analogies for this. The first one I think of is an army. I have this old-school notion of these armies in the 17/1800s where there would be a general who's really making all decisions for his troops, and if something happens, that's got to work its way all the way back up to the general to make a decision, and then flow all the way back down to the troops.
		My understanding—at least with the current setup within the U.S. Army for example—is they are trying to decentralize [and] give a little more autonomy to those pods out there on the field; to be able to be flexible and make decisions. But I think what you're saying is, there are some decisions that ultimately do probably have to go all the way back up. It's a spectrum. There are not just these two end points, there are lots of levels of grey in between. Is that a fair way to think of it?
John Wilson:	05:55	Yeah, that's a great way. Great analogy.
Rob Campbell:	05:56	I was also thinking about football or high-school football. Pretty much every play is dictated by the coach. What happens on the field—obviously there's some discretion there—but play-by-play, it's the coach. It's a very centralized structure as to what's going to actually happen or how the play is going to take place.
		I played a lot of rugby where, sure, there was a coach, but decision-making was really made by people on the field. And I think it was the nature of the sport that dictated why a certain structure worked better than others. Is that also true of the companies that you're looking at? Are there specific types of companies that maybe lend themselves better or worse to having a decentralized structure?
John Wilson:	06:30	A particular business model…I would say that it, again, depends on the activity. A business model is obviously a collection of activities, it's not just one thing, and a decentralized model tends to really work well from a decision-making standpoint. Where it works less well is when there is some benefit to economies of scale. So if you think about operations that have manufacturing, distribution, procurement…oftentimes, those things are better to centralize because, yeah, sure there might be an advantage from a decision-making standpoint, but the advantage that you get on the economies of scale is so great that it doesn't make sense to do that.
		So generally speaking, it would probably be fair to say: decentralized decision-making, centralized things that have economies of scale.
Rob Campbell:	07:13	Okay. So, where there's a need for—I guess related to the economies of scale—where there's a need for this massive amount of coordination between different groups for the process to work. I'm thinking about, like, a car manufacturing plant. Clearly, there has to be this overarching structure to enable everything to work, so a need for a more centralized [approach]. Am I understanding that right?
John Wilson:	07:32	It can be around coordination, absolutely. … I think about one of the companies we own, Bravida, which provides electrical and HVAC installation and maintenance services. They run a portion of their business in a very decentralized manner. At the branch level, the manager is compensated on local branch level profitability, but they centralize the procurement. So all the equipment they purchase is purchased at headquarters because they find they get a cost advantage in actually centralizing that component. It goes into that comment around the advantage that you can get around economies of scale in some of these activities.
Rob Campbell:	08:10	I want to dig into that a little bit more, but just before we get there—and maybe we've talked about this a little bit, but I want to make sure we cover it in full—as you think about a company’s lifecycle, are there other things beyond operational or capital allocation decisions that a decentralized structure might help? And I guess what I'm thinking about is, the company is small. There's a lot of autonomy within that small group. As they grow, do you necessarily lose some of that autonomy? Or do you have to? And what does a decentralized structure try to solve there?
John Wilson:	08:34	I think that that's one of the things that the decentralized organization can help with. Daniel Pink has this great book called Drive, and the whole book is about what creates intrinsic motivation in employees. And one of the things that he found is that it's autonomy, mastery, and purpose. So, very simple acronym: AMP. (What gets you “amped” up.) … if you think about a decentralized structure, I think it helps with all those.
		If you think about autonomy, that's basically being left alone to do what you think is best for the organization—not being micromanaged. That can actually work very well not only from a motivation standpoint, from the employee's standpoint, but also from the business standpoint. One of the businesses that we own is Addtech, a company that produces industrial components. They're very small cost, but they're very important to the end customer, so the customers tend to not be that price sensitive.
Rob Campbell:	09:23	What type of components?
John Wilson:	09:25	Valves would be a big thing for them. It's a conglomerate, so it's across a number of different industries. But one of the advantages that they have in their model is just, they tell potential acquisitions, “hey, we're not going to come in and change how you operate, you know your customer better. We're going to introduce some oversight, we're going to introduce some frameworks, and provide education if you want to take your business to the next level, but we're not going to come in and wholesale change things.”
		Oftentimes they're the acquirer of choice for businesses that are looking to maybe take some cash off the table, but still be involved in operations. So, from an autonomy standpoint, decentralized organization helps.
John Wilson:	10:00	From a mastery standpoint, which is the ability to get very, very, good at something, a decentralized organization is really great for that. Because you've taken the actual decision-making process and you've empowered your employees to make those decisions, so they actually get trained up in how to operate the business.
		One of the businesses that we own is Constellation Software, and what they have been talking about for years is the main bottleneck for their growth—the main thing that's holding them back—is talented individuals that can allocate capital and that can run the business. When you think about that, their strategy of running a decentralized organization actually makes a lot of sense because they're training tomorrow's leaders.
Rob Campbell:	10:38	I would think that there's a flip side to that too, though, right? I mean, if you're giving lots of autonomy, and you're decentralizing that structure…I would imagine that also poses a problem in some degree to succession because there isn't this clear plan, or somebody that you can go back up to. … If a “business unit” (to use a Constellation Software term) has little connection to the rest of the organization…what happens when the key individual there moves on? Does that not pose a risk?
John Wilson:	11:03	I think that actually removes key person risk because, again, you're pushing down the decision-making so it doesn't all run through that one person in the organization at headquarters. Instead, you have a collection of trained-up decision-makers. I think it can actually alleviate those problems.
Rob Campbell:	11:21	Okay, so we talked about autonomy. Just to go back to Daniel Pink—the autonomy on decision rights. I can say as an employee, I like not being micromanaged. It's just empowering. There's the mastery element, so you're a little bit more in control of your destiny to build skillsets that you otherwise wouldn't. I don't know that we covered the purpose piece in as much detail as I would have liked. Can you say a little bit more just about that?
John Wilson:	11:40	Sure. So, purpose is really just the feeling that the work means something to you. And the goal—if you think about a decentralized business—you want people to feel as though they're in charge of…almost an entrepreneurial enterprise. They have a business within a larger business that they take ownership in. And if you ever talk to an entrepreneur, they often refer to their companies as “their baby.” They have a lot invested in it. They want to see it succeed.
		And that's what you're trying to create with a decentralized structure. It's something that we use in global small cap as well that we call “Entrepreneurs Within the Enterprise.” And because all great ideas need to have an acronym, we call it EWE. The idea is that—
Rob Campbell:	12:21	EWE?
John Wilson:	12:22	EWE, yeah.
Rob Campbell:	12:22	Okay [laughs]. All right.
John Wilson:	12:23	So, EWE is: you want to set up the structure, which I think we'll be talking about in a bit, so that you can run as though you're an entrepreneur and have a lot of autonomy. It allows you to improve your skills and hopefully create purpose for your[self] as well.
Rob Campbell:	12:37	And inherent in all of this, I would imagine, is you have to have the right people. The right people on the bus.
John Wilson:	12:41	Absolutely.
Rob Campbell:	12:41	So, you've got to have people that are naturally inclined, or just wired to want to operate autonomously—who are just driven to constantly make improvements and to help out.
		Am I right that this decentralized structure wouldn't work if you've got a bunch of people who are just…sort of coasting and looking for direction?
John Wilson:	12:57	Absolutely. If you have an organization where employees aren't interested in being proactive, then you're not going to get the behaviour that you want.
Rob Campbell:	13:06	And maybe just to finish up on, what's the problem [a decentralized structure is] trying to solve, can you talk a little bit about how this helps a business scale? Maybe Mark Leonard is a good example there. [If you] think about the Constellation Software business model: acquisitions at really attractive rates of return…you become a bigger company. How can one person—or how can Mark Leonard—do all the acquisitions? Surely there reaches a point where his time is just not divisible that way.
John Wilson:	13:29	Yeah, I think that's absolutely it. Oftentimes, as you scale up—if you think about information flowing through the organization—the larger the organization grows, the more difficult it is for one decision-maker to handle all those decisions and to allocate capital across the firm, or to make operational decisions. So, it's very helpful to have this decentralized structure.
Rob Campbell:	13:52	So, I'm running a business, [and] I'm trying to scale my business. I see the advantages of having a decentralized structure that I'd like to implement. What are the keys to making this work? In other words, what do I need? Or where have you seen in the global small cap universe this done successfully? And can you draw some key lessons out of that in terms of, what really makes this type of structure work?
John Wilson:	14:12	Sure! There are a couple mental models or thoughts around this. The first one would be what I would call, “clarity and alignment.” It's one of the things that we talk about internally. And one of the things we look for when we're talking to management that have a decentralized structure is “field marking.”
Rob Campbell:	14:30	Sorry—field marking?
John Wilson:	14:31	Field marking.
Rob Campbell:	14:32	Okay.
John Wilson:	14:35	There are three categories within that. It's: “where are we going?” So, what's the goal, and how do we measure progress towards that goal? “What's your role in getting closer to that goal?” And, “what does that involve?” What activity does that involve and what activities does that not involve?
Rob Campbell:	14:50	Okay, just to make sure I got this…I'm setting up a decentralized structure, and again, we're speaking in black or white terms—I’m sure this happens more organically, but companies would decide to do this. What you're saying is, in those conversations as they seek to push out the decision-making below them, these are conversations that would have to happen with those managers who are going off to run the underlying units.
John Wilson:	15:14	Yes, I think that's a really good point. When I'm talking about the field marking, these are the decisions that should be made centrally.
Rob Campbell:	15:22	Okay.
John Wilson:	15:25	The big decisions, the big direction. And the other component along with direction and the actual role, is, what they classify as “values.” Areas that we've all agreed upon, [and] this is how we're going to act. So for example, one of the things at Mawer we believe and one of the things we talk a lot about is candidness. That's not just telling the truth, but it’s if you disagree, you need to speak up because we think that helps make better decisions. So from a value standpoint, that's something that we've all agreed upon.
Rob Campbell:	15:52	And values and culture are pretty much synonymous in a lot of ways in that sense, right?
John Wilson:	15:57	Yeah.
Rob Campbell:	15:58	Your corporate values are going to dictate the culture that you have in this organization.
John Wilson:	16:00	Exactly. Values are kind of the rules of engagement that you've decided upon. And the reason why we call it “field marking” is … you kind of say, “hey, this is what the goal is. Here's your role on the team, and here are the boundaries.” That's kind of the value piece. “Now, within those boundaries, go ahead, be an entrepreneur.” That's why we call it field marking.
Rob Campbell:	16:19	Okay, so for an organization to do this successfully, they have to really clear parameters around what the autonomous decision-making can operate within.
John Wilson:	16:27	Absolutely.
Rob Campbell:	16:27	Got it. So, field marking is one of them. What else is key to making this work?
John Wilson:	16:32	The other key thing is aligned incentives. You want to make sure that the incentives are aligned within the organization. So, ideally when you think of aligned incentives—you want something that they can control, and something that gets you closer to the goal that's aligned.
		We've seen good incentive structures around return on capital; net promoter score, which is a measure of customer satisfaction; on time deliveries…things like that.
Rob Campbell:	16:55	I want to go back to the Bravida example. Their branches operate autonomously, those branch managers have leeway in terms of making decisions as to what they should do. But, ultimately, the alignment is in the way that they're compensated, which is tied to a more centralized agreement—return on capital, profitability, whatever that is.
John Wilson:	17:14	And in that case, what they tie it to is factors the manager can control. … They have control of their costs, they're compensated on a margin, and working capital is the other component. So they're compensated on lowering their working capital.
Rob Campbell:	17:26	Okay, so, alignment. We talked about a lot of things when it comes to alignment, but that's really the first key to making it work. What's the next one?
John Wilson:	17:34	The other important thing that you have in place is, a system for having oversight over the decentralized organization.
Rob Campbell:	17:41	Okay, and is that different than alignment?
John Wilson:	17:42	Yes, I would say it's different. Alignment is: we all agree that this is the right direction to head, and, “hey, you're going to be compensated for heading in this direction.” This is a control, this is, “hey, are you actually heading in that direction?” Or, “hey, are there issues that crop up?” And ... there are these really interesting cases where you use social pressure. I've seen it before. I've talked to a manager—the way he runs his business … whenever there's a meeting of all the subsidiary managers, on their name tag, they'll have their name tag and I think it's return on capital, or customer satisfaction. It just creates this little incentive to focus on, again, go back to that goal that you're—
Rob Campbell:	18:29	You're walking around with a report card on.
John Wilson:	18:30	Exactly! And we do that here at Mawer as well. You'll notice, on the wall over there there's a list of—we call it “the bathroom list.” Every year, we need to go and review all the companies in our portfolio, and there are stats on that.
Rob Campbell:	18:48	So, these are minimum standards, or things that you need to do at a minimum annually to be able to keep holding a stock in the portfolio.
John Wilson:	18:55	Absolutely. It's on a 12-month lap. And you'll look—and most of the stats are above 90%. We've talked to management in the last 12 months, we've updated the model. Probably higher than that. But still, because you see that, let's say, Canadian large cap is at 98% and global small cap is at 95%...that gives you that little extra nudge to—
Rob Campbell:	19:16	So nobody wants to be the one who is sort of last, or not important.
John Wilson:	19:21	That's it. Exactly.
Rob Campbell:	19:22	Okay, so let's get back to the oversight piece. It's ultimately the concept of a dashboard.
John Wilson:	19:27	Yes. Often, these companies…I think about Freightways, which is a company based out of New Zealand that does courier services. They are able to look at branch profitability all the way down to the local level and they can compare that to how that branch performed the year before, the year before [that]. So, they can flag if there are anomalies; things that are outside of the historic range; to try to understand what's happening and whether they need to step in and coach and provide advice to a local manager.
Rob Campbell:	19:56	So, we've got alignment of incentives, which is sort of setting the playing field. You've got effective oversight systems, which, I guess is the monitoring aspect, but I would also think is sort of the accountability piece too—we're ensuring accountability. You mentioned three things. What's the final one?
John Wilson:	20:12	The final one is “aggregating up” information. One of the potential disadvantages of a decentralized organization is that people might learn the same lesson again, and again, and again. You might make an error at one plant, and that might be replicated at another plant—
Rob Campbell:	20:29	Or don't realize that you're making a mistake.
John Wilson:	20:31	Or don't realize it.
Rob Campbell:	20:32	Or learning happens at some other plant, and it just doesn't filter through the rest of the organization.
John Wilson:	20:34	From a best practice standpoint, absolutely. So, the idea is you should aggregate up learnings, whether it's a mistake, or whether it's best practices, so that people can learn from each other.
Rob Campbell:	20:44	Is there an example of a company that you've looked at that does that really effectively?
John Wilson:	20:47	Going back—Addtech is a business that does that quite well. The company has what they call the “Addtech Business School.” Throughout the year, they hold seminars, where people [and] the various subsidiaries meet up. They share best practices from what they learned in their business, and the management team itself educates them around concepts around managing your working capital, allocating capital, serving the customer, R&D—things like that. So the managers are able to benefit from these education seminars in order to take their business to the next level.
Rob Campbell:	21:30	What are some of the areas where there are pitfalls associated with this? Where can companies get this wrong?
John Wilson:	21:34	Well, I think one of the big things is when you don't have these structures in place and when you're not deliberate about setting up the right incentives. One of the things that Charlie Munger said, who is Buffett’s right hand man is, “you show me the incentives and I'll show you the outcome.” So, that is something that's important to set up.
Rob Campbell:	21:52	We've talked a lot about the companies that you're looking at, but what about us? Are we a decentralized organization?
John Wilson:	21:58	Yeah, I think here at Mawer we absolutely are. When I think about that framework we talked about, just around clarity—this weekend we're going off on our annual retreat, so that's where we clarify where we're headed, the end goal. We also talk about roles and recommit to the values of the firm. I think we also have a very clear investment philosophy, so, we're very clear that we're looking for companies that are wealth-creating, run by capable and honest management teams, and purchased at a discounted intrinsic value. That's something that we talk a lot about and I think there's a high degree of clarity around that.
John Wilson:	22:36	And then, if you think about the second component, which is around effective oversight, there are a number of different mechanisms that are in place around that. We talked a bit about the bathroom list—just trying to be up to date on that. The other thing we have are these 60-day check-ins where every employee within Research will check-in with a leadership member on things like, what's holding them back in their current role? Are there any bottlenecks that they're encountering? It can be a check-in on whether they're encountering any issues, or what's going well. And then that ties in a bit to the aggregating of information because there are also these conversations around best practices that happen then, too.
John Wilson:	23:17	Finally, if you think about the aggregation of knowledge, we have this great tool that we use internally called M42. This is a great database of all the notes on companies that we've spoken with; our thoughts on those companies. That's a great way to see what other people think about a business, what they think about the quality of the management team. And that stokes further conversations.
Rob Campbell:	23:37	And what about the global small cap team? You guys operate as an asset class team independently, but in terms of the four of you—would you say that you guys within that group operate in a decentralized way?
John Wilson:	23:49	Absolutely. I think that was a big push, a big thing that we've been working on for a number of years now—almost out of necessity. In the last year, we had one of our members, Karan, move to Singapore. Obviously that makes it very hard to run the portfolio or the team in a very centralized manner because the time zones are completely off. I think it's 12 hours from Toronto, 10 from here, so it becomes very difficult. So, when you think about setting the field marking, you have a really good understanding of where you should spend your time and what sort of activities you should be involved in. Those things become very important. And oversight becomes important. So that whole framework applies right through global small cap, and we've found that its increased our efficiency a lot.
Rob Campbell:	24:39	Okay, great! Bunch of amped-up individuals on the global small cap team [laughter]! Again, thanks so much for being on the podcast. Love to have you back.
John Wilson:	24:46	Thanks Rob.

---

Related Reads

• Mawer Global Small Cap Fund
• The competitive advantages of decentralized structures

---

How to subscribe

The podcast is available to listen and subscribe through any of the following platforms:

• 
• 
• 
• 
• 
• 
• 

Subscribe to Art of Boring to receive email notifications when a new episode is available, as well as other insights through our blog and quarterly updates.

---

Have feedback?

If you enjoyed this episode, feel free to leave a review on iTunes, which will help more people discover the Be Boring. Make Money.™ philosophy. 

If you have any questions, comments, or suggestions about the podcast, please email podcast@mawer.com.

---