The Vertical Sage
Mark Leonard and Constellation
Today we take a step back and look at one of the legends of vertical software. Capital allocation is incredibly hard and there are few better than Mark Leonard and Constellation.
Mark Leonard is perhaps the single best CEO that nobody has ever heard of. He has a cult-like status amongst a certain group of investors, but by and large nobody else thinks too much about Mark. I know I sure hadn’t before I went down the vertical rabbit hole.
This is especially surprising when you look at Constellation stock over the years. At a current market cap of around $29B, Constellation is one of the largest public companies on the Canadian stock exchanges and an incredible business. Guys who create that kind of enterprise value tend to get press. But Mark has a tendency to eschew notoriety.
It’s intentional. Leonard is a sort of rare breed. He’s incredibly private. There are no interviews.1 Only 3 photos of the man exist (including the 1 above). And the only writings we really have are his shareholder letters over the years.
As a result, it’s quite difficult to assemble a portrait of the man, the company, and Constellation. What’s really needed is an authorized biography of the man and his thinking. But until we get that, we must rely upon an amalgamation of sources.
As we enter a heightened wave of vertical SaaS adoption, it feels incredibly important to get to know the man and the company who have scouted more vertical businesses than anyone else out there. After all, they are probably your competitors, whether you are an investor or founder. But it also feels important to give Constellation credit. It is truly an incredible company with unique quirks and highly intelligent practices.
So what you will find in this piece will be a mix of all things around Constellation. While some of it is around the business model, it will not primarily focus on financial prospects. There is plenty of great financial analysis that I will link in the footnotes. The intention of this piece is not to pick stocks, but paint a picture of how Constellation works and how it will function in the coming vertical age.
Leonard and Software Co.
I was 39 when I started CSI. I had 4 kids, all under the age of 10, a modest house, and a mortgage. I drove a Toyota Camry and my wife had an ancient Chrysler minivan. I knew how to work ridiculously hard, and I had a decade of experience in the venture industry. That decade taught me how to collaborate with small, motivated teams, and the difference between a good and a bad business. I badly wanted to start a business where I would be fulfilled.2
Much of Mark’s life prior to founding Constellation doesn’t look particularly remarkable. Mark doesn’t have a storied education nor was he well-connected. He worked plenty of jobs in his early career spanning masonry and more. After business school, Mark really wanted to get into VC. But since the VC community in Canada was incredibly small, he couldn’t find a job. Instead, he ended up in banking while looking for a way in.
He found it by outworking everyone. Even though he was broke, he would attend every conference VCs were at. Soon they started to recognize him, and one day when a role opened up, Mark got his shot. While at the firm, he loved venture capital and the breadth of ideas he was exposed to. But he also began to avidly read some of the greatest investors. Buffet was one of them, but there were others. They convinced him of the value of great businesses with certain characteristics. Soon Mark began to look for these “great businesses.” He soon found them in vertical market software companies (Constellation’s preferred nomenclature).
Vertical software businesses have all the classic traits that you would want in a great company. They have high margins, are capable of generating great cash flows, and most importantly, tend to have high switching costs.
The only problem for Mark in his current role was that these companies were not typically venture scale. Most were destined to only ever do millions of free cash flow, not hundreds of millions. It quickly became clear to Mark that his desire to work on great VMS businesses was coming into conflict with his role as a VC. So Leonard began to figure out a different path.
Luckily, the venture fund Leonard was at was pretty receptive to new ideas and Mark began pitching the initial idea for Constellation which at the time he called Software Co.
Investors bit and Mark had $25m CAD in ammo to go and buy VMS businesses. He quickly got Bernard “Bernie” Anzarouth on board to lead the investment and finance strategy and they were off to the races.
As far as I can tell, there simply isn’t a lot out there about the early years of Constellation. From about 1995 until when Constellation goes public in 2006, you won’t find much. Again, the man has only three pictures on the internet.
But from 2006 and onwards, we really get to know Leonard and Constellation through his annual letter to shareholders. So from here, we go a bit more into how Constellation actually ticks.
The model they followed has continued to this day: buy VMS businesses, incentivize leaders to achieve 40% free cash flow, and then reinvest cash flows into buying more and more vertical businesses. The model is the easy part, operating effectively as you continue to scale is the stuff of legend. Conglomerates tend to only outperform for limited amounts of time. The reason Constellation is legendary amongst investors is because they have refused to revert to the mean. There’s really no indication that they are slowing down either.
Live Free or Die
To really understand Constellation, you need to understand one crucial reason for its success. Mark Leonard doesn’t wear crowns.
What I actually mean is that Mark Leonard doesn’t give a damn how big Constellation gets, he’s will never reign with an iron fist. Constellation is all about autonomy and freedom. Mark never liked being ordered around and refuses to do the same to his employees.
Constellation works by decentralizing as much power and decision making as possible to the smallest unit within Constellation: a business unit. When decisions need to be made at a higher level than an individual business unit, they are made by one of the six operating groups that comprise Constellation: Harris, Volaris, Jonas, Vela, Perseus, and Topicus.
Each operating group is again autonomous and responsible for their own legal, finance, M&A departments, and more.
Constellation’s main office then acts primarily as a huge data compiler. Data flows up from the individual business units through to the operating group and into the Constellation main office. Constellation is then able to feed this data into their other operating groups for them to utilize in making operational decisions. But again, operating groups are never ordered to act upon the data or to adopt a best practice from another operating group. Instead, they must be convinced.
This has allowed Constellation to run semi-independent experiments in each of the operating groups and to move far faster since operating groups are trusted to act as they see fit.
This is not par for the course at most conglomerates.
Only one other HPC has followed a strategy of buying hundreds of small businesses and managing them autonomously. They eventually caved in to increased centralisation. My hunch is that it takes an unusually trusting culture and a long investment horizon to support a multitude of small businesses and their entrepreneurial leaders. If trust falters the BU’s can be choked by bureaucracy. If short term results are paramount, the siren song of consolidation synergies is powerful. We continue to believe that autonomy and responsibility attract and motivate the best managers and employees.3
Everything else springs up from this central value around autonomy. It’s what renders Constellation so idiosyncratic and so successful.
From the above quote, you can also start to get a hint around what also matters to Constellation here: hodl’ing. Constellation doesn’t sell their businesses.
“We Don’t Do Deals”
Operating companies isn’t the only thing that’s done by operating groups and business units. They are also given primary responsibility for acquiring new businesses. And you can get a sense of how seriously they take this job from this quote from Jeff Bender, the Harris Operating Group President.
So we don't do deals. We make investments. This is a pet peeve of mine.4
Constellation doesn’t do deals. When they buy, they buy for the long haul and thus everyone is incentivized upon one big metric: ROIC.
Our compensation system at the senior level is driven around return on capital because we don't sell our businesses. If you've done something and it hasn't worked out, it is in your capital base forever… leaders are compensated off of ROIC. What this means is that nobody can escape bad acquisitions and reduce their invested capital base through selling and they need to deploy capital effectively to generate personal return.
ROIC as a defining metric allows Constellation to safely decentralize M&A. Since leaders are judged off of ROIC, their incentives are entirely aligned.
The result has been an incredible return for shareholders and an organizational prowess across the entire company to scout acquisition targets, make acquisitions only at the right price to generate sufficient return, and close at a high clip without decision-making being overly dependent on a centralized M&A function.
It’s remarkable what this sort of discipline, personal accountability, and upside has created in the culture of Constellation. Everyone is highly conscious and strives for excellent capital deployment.
The Long Haul
Since capital allocation decisions are mostly decentralized, it’s also critical that employees don’t choose short term results by juicing ROIC temporarily. Constellation solves this by requiring employees above a certain rank to invest a certain percentage of their bonus into company stock (purchased on the open market). This has several effects.
First, it focuses everyone upon the long term health of Constellation and long term enterprise value rather than juicing short term gains. Second it has created several hundred employee millionaires. Third, it has created a culture of Constellation die-hards.
Constellation’s best employees don’t tend to leave and are typically in it for the long haul.
My motivation is to help create a company where worthy people succeed… I want to support and encourage employees who work hard, treat others well, continuously learn, and share best practices. I try to make sure that sycophants, spin-doctors, and mercenaries don’t survive in Constellation’s senior ranks. Harder, but not impossible, is helping identify and remove hidebound managers who rely upon habit and folklore to run their businesses rather than rational enquiry and experimentation. Constellation is as close to a meritocracy as I have experienced.5
What makes Constellation work is exactly this blend of astute capital allocation, long term thinking, and meritocratic culture. Mark has created a high-trust and high ability employee base which is given vast autonomy and incentivized to think about the long term health of the business. Even better, they can realistically see themselves rising through the ranks.
One of the things I respect most about Mark is his honesty. Mark tends to be incredibly thoughtful and honest about the tradeoffs involved in the Constellation model. This is probably what makes him so great.
Great capital allocation requires a holistic assessment of the tradeoffs involved in an approach and then the guts to go through with it. You can grasp this from Leonard's thinking on the other high performing conglomerates (HPCs) that he often measures Constellation up against.6
One comes up quite frequently as a point of comparison: Jack Henry & Associates. Jack Henry is a HPC focused entirely on financial institutions. Leonard is an admirer of JHA’s prowess and you can fully grok it from his 2015 Letter.
JHA has been able to do something quite remarkable from an investor's perspective. Despite the total number of financial institutions declining year after year, JHA has continued to generate greater profits. The secret is that JHA has become highly effective at acquiring existing solutions within the industry and then cross-selling them into their customer base. The net effect is an incredible moat around its customer base and the ability to print consistent revenue.7
But the reason Leonard writes and thinks about JHA is not only because of their savviness but because Mark thinks deeply about the tradeoffs involved in this approach.
In the pursuit of a deep moat and continuous expansion through cross-selling, JHA began to do acquisitions without regard for price. So long as it was strategically important, they would buy it. In other words, JHA did not have the same discipline around ROIC that characterizes Constellation.
The result is certainly an incredible business with high revenue per share, retention, and free cash flow.
But in acquiring the moat, Jack Henry made a choice to not maximize return on invested capital. And this is the tradeoff that Leonard and Constellation have historically been unwilling to accept.8 Leonard isn't that interested in developing moats if they result in inferior returns. It's a conscious tradeoff and one he still thinks about often.
This self-reflective tendency is the reason Leonard is so highly trusted as a capital allocator. He’s willing to acknowledge real tradeoffs and choose a path forward while acknowledging the real limitations of his approach.
The reason I think this tradeoff is particularly important is because of what it means for Constellation’s future. They haven’t tried to create large moats through cross-selling. Has this left them vulnerable? What happens when more and more vSaaS companies start to attack their businesses without firm moats?
There’s plenty that could be said about Constellation's financials, structural challenges, and more.9 But this newsletter is fundamentally about vertical SaaS and the new paradigm enabling venture returns out of vertical markets.
So what awaits Constellation in the new vertical age?
The Next Decade
First, Constellation’s success with smaller VMS companies has convinced them to go after larger acquisition targets. In 2021’s Letter, Leonard wrote:
One of our directors has been calling me irresponsible for years. His thesis goes like this: CSI can invest capital more effectively than the vast majority of CSI's shareholders, hence we should stop paying dividends and invest all of the cash that we produce, even if it means lowering our hurdle rates…
I have stopped arguing. I have converted, and with the fervour of the newly converted, I am busy demonstrating my new-found faith.10
All cash is now going into acquisitions. And with this huge pile of cash, Constellation is expanding into larger deals.
To do so, they have centralized a new large acquisition group inside of Constellation proper. It represents a change in strategy in several ways. First, it’s obviously not autonomously governed by an operating group. Second, within this size, ROIC is less important (to an extent) than finding great businesses that Constellation wants to hold forever.
I’m very interested to see the types of companies Constellation acquires in this new group. They won’t pay the highest multiples, but theoretically there are other reasons to go with Constellation. Of course you get the autonomy. But say you also have aspirations to go public long term. Constellation is happy to let you do so (they’ve already spun out Topicus into an independent company).
The upside for these larger companies with dreams of IPO’ing is that so long as Constellation is a shareholder, you won’t be subject to a hostile takeover. They aren’t selling. And I can see certain companies biting at this.
More recently, Constellation has also spun out a venture capital arm, VMS Ventures, with a 200 million dollar fund. In typical Constellation fashion, it carries some quirks. First, it’s highly idiosyncratic and focused on its employees as the entrepreneurs it will fund. Second, individual business units are responsible for sponsoring the idea and team for funding.11 In Mark's own words, this is "outsourcing diligence to the business units."12
There is also a tremendous emphasis on “earned secrets” - unique insights into a vertical.
If you have a unique insight into a particular customer problem through deep knowledge and experience in a vertical, and this insight is not obvious to agile and deep pocketed competitors, then you have an earned secret. That allows you the opportunity to grab a sticky greenfield market by subsidising customer adoption. We believe that our employees are the fortunate holders of thousands of unexploited earned secrets. So, bring them on.13
At the heart of the VMS Fund is a realization that while Constellation is fantastic at ROIC, it is not great at organic growth. The solution then is to use the VMS Fund as a way to pursue organic growth by leveraging human capital already at the company and who are deeply familiar with vertical software. In Constellation’s mind, this will have rippling effects on organic growth across the firm.
It’s a very different way of doing venture and I am extremely keen to see what happens and what sorts of companies come out of it.
Lastly, Constellation has been taking a look at non-VMS businesses with the same attributes that their core business has had. They haven’t exactly pulled the trigger:
I mean the uncertainty of non-VMS is that there are hundreds of thousands of Harvard and Stanford and MIT and God knows what all else, MBAs, hustling around, trying to find places where you can generate alpha. And so the likelihood of us being able to find another place to generate high rates of return on significant amounts of capital in a predictable way for decades is bloody low. And so I would tell you, don't count on it, but we're working on it.14
The closest Mark seems to have gotten is around a 1 billion dollar tax-advantaged thermal oil investment where the “oil markets ran away” on him. Yet it’s still pretty unclear if Constellation is really serious about this type of expansion.
Over the next decade, Constellation will keep doing Constellation things. They will acquire great VMS businesses, optimize them for cash flow, and keep building an incredible culture. Sure, there are open questions about returns when you are sitting on this amount of cash, on large acquisition strategy, and more. But, leaving aside the financial performance questions, the really big question is how the vertical SaaS wave both impacts their own business units and how Leonard is thinking about the rise of venture dollars in the vSaaS market.
Newspapers, Airplanes, and Automobiles
There’s one aspect of Mark Leonard and Constellation that every vertical entrepreneur and VC firm focused on vertical SaaS has to think about.
…[We] are very worried about limits to the number of good VMS businesses that we will be able to buy at reasonable prices.
A large part of Leonard’s success is his skepticism. He doesn’t buy into hype or vision. Instead, he buys great businesses with predictable attributes so long as they are at valuations that give him a high ROIC. But what happens when money starts to realize the steady returns inherent to vertical SaaS and is willing to buy them at larger multiples?
That’s the future that Mark is thinking a lot about.
His basic concern is that in growth markets you end up with a lot more shipwrecks than safe passages. Whenever new opportunities come to light, money starts to chase the grand prize: winning a vast majority of the vertical. The problem is that due to the winner take all dynamics, it’s somewhat of a race to the bottom, with markets subsidizing the businesses to gain wallet share. It’s what happened in the newspaper industry as newspapers consolidated, and similar dynamics also happened in the airline and automobile industries.15
Thus, Leonard is quite concerned about how these new VC-backed businesses will affect the overall VMS industry and more specifically, Constellation’s ability to acquire these businesses at good prices.
Frankly, my hunch is that this should be a concern for Constellation specifically but less of a concern for those around the new vertical SaaS markets. I’ll get to this next week, but VCs aren’t making irrational bets in these markets. The legacy software simply isn’t cutting it and entrepreneurs and investors are seeking returns by providing verticals with far better software!
Further, the potential return profile has changed substantially and these businesses can be quite a bit larger than the VMS businesses that Leonard has tended to acquire.
Sure, more entrants into markets might not be great if you have been highly successful by finding these markets and buying the existing businesses, but it doesn’t mean that those entering the markets are racing to the bottom. In fact, quite the opposite.
The reason is pretty simple: Constellation has prioritized free cash flow mostly at the expense of R&D. And while it’s tough to know the exact breakdown of SaaS vs. on-premise businesses, they have large amounts of on-prem business units.
Based simply on these two data points, their business units largely aren’t set up to compete in winner take all markets. But since (and to be explored next week), the total revenue that can be generated from a vSaaS business is far greater than it has been historically, they also can’t rely entirely on their historic strategy of playing in markets where they aren’t winner take all because a certain group of entrepreneurs and VCS are willing to play in every market.16
I think you can sense some of the ways in which Constellation is trying to grapple with this even from their new strategic initiatives detailed above. First, develop a culture of entrepreneurship and fund them. Second, acquire larger VMS players who more clearly look like winners. Lastly, look for diversification opportunities in non-VMS industries as a failsafe. All of these are responses to a real existential threat that Constellation feels.
In many ways, I think this new concern with the future of VMS investment mirrors some of Mark’s writings on Jack Henry. Recall that Jack Henry traded off ROIC for better moats and products to cross-sell. This new paradigm is a similar matter of tradeoffs.
By seeking good businesses but not world-class businesses within their verticals, Constellation has been able to build an incredible capital allocation machine with high ROIC. What they have forsaken is building deep moats within verticals. Yes, they have sticky products, but their moat is entirely centered on high switching costs. In most of his writings around VMS, Leonard tends to act as if this is the only real moat available to VMS businesses. It’s most definitely a highly important one but it’s not the only one.
It will be the task of vFounders to figure out defensible moats even as they lessen switching costs by moving their customers to the cloud. This isn’t an impossible problem, it’s just one that Leonard really hasn’t had to grapple with. There’s already really good ideas on how to create moats in new vSaaS companies.
It’s not all a loss for Constellation. Even if some of their businesses get taken out by new entrants, not all of these entrants are destined for an IPO. And surely Constellation will be there at the right price to give these companies an everlasting home.
In the meantime, business is still war. And the best way to respect the previous conquerors is to take shots at their holdings. There are plenty of Constellation business units that are worth competing with. Do Mark the honor of giving him a new business to prioritize acquiring.
Next week, I want to dive a bit more into what separates new vSaaS solutions from old ones and why they actually can be far larger companies. We will take a look at one specific vertical which I think will be a lot of fun.
Liked what you read? Consider subscribing and joining the Verticalized community.
There’s one and it was pretty much instantly taken down when it was posted publicly. I’ll leave it to you to find the bootleg audio recording.
From the 2015 Annual Letter.
Jeff Bender, the President of Harris, in an interview.
From the 2018 Letter.
Leonard’s HPCs of choice are: Ametek, Berkshire Hathaway, Danaher, Dover, Illinois Tool Works, Roper, Jack Henry & Associates, Transdigm, and United Technologies.
JHA has a 99% retention rate.
There’s an adjacent reason for this: a tradeoff around autonomy at the business unit level. It’s really hard to think in terms of long term moats as a business unit leader when you are judged in part upon ROIC.
JKHY appears to be willing to pay high prices for some third party add-on product businesses that might sell well into their installed base. We have tended to be more sceptical of such cross-selling synergies, perhaps because the investment decision-making has not historically been at the BU manager level. A lesson from JKHY, is that we may have been overly cautious regarding cross selling synergies.
From the 2018 Letter.
You can get a grasp of the idiosyncracies in this case study There’s also a great line about rejecting “ramen pay.”
Taken from the 2022 Annual Meeting
I have never seen the document above nor this one discussed. You can read through Leonard’s Q&A on the Fund from where this was taken here.
The 2022 Annual Meeting
Mark apparently gave a talk on this at a business school this year. If anyone ever finds, the audio, give me a ping.
I also don’t know if it’s really fair to categorize venture investment in verticals as a race to the bottom. Even in highly competitive verticals, we aren’t seeing hundreds of the same solution. We are seeing vertical ecosystems. This could change, but I don’t think it’s likely.