The Constellation Model

We are suspicious of ‘vision’. Long-term studies suggest that visions are nearly always impractically vague or outright wrong.

The anti-conglomerate

Mark Leonard started Constellation Software in 1995 with C$25 million. When he stepped back as President in 2025, the company was worth over $70 billion. The transformation rested on a single insight: most serial acquirers fail because they eventually run out of good deals and start doing bad ones.

Leonard’s solution was counterintuitive. Instead of building a team at headquarters to find and integrate acquisitions, he pushed capital allocation down — to operating groups, then to business units, then to individual managers. The more people looking for deals, the more small deals you can do. The more small deals you do, the less pressure to do large ones.

Constellation now runs over 1,000 business units with a 14-person head office. The ratio is deliberate.

Hurdle rates are magnetic

Leonard’s sharpest insight concerns the psychology of capital allocation. When you lower your required return threshold, you don’t just accept marginal deals at the lower rate. You drag down returns across your entire opportunity set.

“Capital is magnetically attracted to mediocrity when hurdle rates are lowered.”

Constellation has changed its hurdle rates only three times in thirty years — once up, twice down — and tracks the effect meticulously. The discipline required is uncomfortable. It means walking away from deals that look good but don’t clear the bar. It means watching competitors win auctions you could have won. It means explaining to your board why you’re not deploying capital faster.

Most acquirers eventually break. Leonard didn’t.

Perpetual ownership

“Constellation’s objective is to be a perpetual owner of inherently attractive software businesses.”

They have sold exactly one business in thirty years — early on, for a very high price. This is a fundamental difference from private equity. Constellation doesn’t restructure and flip. The time horizon is forever.

The implications ripple through everything. When you’re never selling, you optimise for different things. You don’t load businesses with debt. You don’t strip costs to inflate exit multiples. You don’t churn management. You build for decades.

This positioning creates deal flow advantages. Founders trust Constellation with their life’s work in ways they don’t trust PE. Leonard understood that reputation for permanence is a competitive moat that compounds over time.

The master formula

“Our favourite single metric for measuring our corporate performance is the sum of ROIC and Organic Net Revenue Growth.”

Because Constellation’s businesses are capital-light, organic growth doesn’t require incremental capital. You can simply add ROIC to organic growth to estimate value creation. Compensation is tied directly to this metric.

The formula is elegant because it forces trade-offs into the open. A business generating 30% ROIC with 2% organic growth creates roughly the same value as one generating 20% ROIC with 12% growth. Both clear the bar. Neither is inherently better.

Leonard: “The toughest challenge in the software business is intelligently trading off profitability and organic growth. Many entrepreneurs have a huge bias towards growth at the expense of profits. Most private equity owned software firms have the opposite bias. At Constellation we try to find an optimum position where incremental investment still generates good incremental long term returns.”

Vertical market software

Constellation’s businesses share a pattern: they serve specific niches — club management software, marina management, transit scheduling, dealership systems. The markets are small enough that larger competitors ignore them. The software is embedded deeply enough in customer operations that switching costs are prohibitive.

Leonard never articulated grand strategic vision. He was explicit about this: “We are suspicious of ‘vision’.”

Instead, Constellation found a type of business with attractive characteristics — recurring revenue, high switching costs, capital-light operations — and bought hundreds of them. The portfolio emerged from the accumulation of small decisions, not from a master plan.

The culture beneath

Leonard built specific beliefs into how Constellation operates.

“We try to make sure that sycophants, spin-doctors, and mercenaries don’t survive in Constellation’s senior ranks.”

The frugality is legendary. Leonard on business travel: “I can personally afford to fly business class, and I could probably justify having Constellation buy me a business class ticket, but I nearly always fly economy. I do this because there are several hundred Constellation employees flying every week, and we expect them to fly economy when they are spending Constellation’s money.”

The implication: “The standard we use when we spend our shareholders’ money is even more stringent than that which we use when we are spending our own.”

Philosophy made visible.

Decentralisation as strategy

Most serial acquirers centralise over time. Synergies require coordination. Coordination requires control. Control requires headquarters staff.

Constellation went the opposite direction. As the company grew, Leonard pushed decision-making further down. First to six operating groups — Volaris, Harris, Jonas, Vela, Perseus, Topicus. Then to business unit managers within those groups. Each layer gained its own M&A capability.

The result: Constellation can do hundreds of small acquisitions annually instead of being forced into larger, lower-return deals. The constraint on growth shifted from “how many deals can headquarters evaluate” to “how many good small businesses exist.”

Leonard in 2014: “CSI does have a compelling asset that is difficult to both replicate and maintain. We have 199 separately tracked business units and an open, collegial and analytical culture. This provides us with a large group of businesses on which to test hypotheses, a ready source of ideas to test, and a receptive audience who can benefit from their application.”

By 2025, there were over 1,000 business units. The asset kept compounding.

What Leonard built

The track record speaks for itself: 36% compound annual returns since the 2006 IPO, a 156x return while the S&P 500 returned roughly 6x. No share dilution — the same 21 million shares outstanding as at IPO. Over 500 acquisitions completed.

But the numbers only capture part of the achievement. Leonard proved that you could build a compounding machine by pushing autonomy down rather than pulling control up. That small deals done well beat large deals done adequately. That hurdle rate discipline, maintained for decades, creates its own moat.

In 2018, Leonard stopped writing his annual letters. The reason was characteristically blunt: “We are limiting the information that we disclose about our acquisition activity. We believe that sharing our tactics and best practices with a host of Constellation emulators is not in our best interest.”

He’d built something worth copying. He wasn’t going to make it easy.

Connects to Library: Process Power · Optionality · Scale Economies

See also: The Roper Model — Similar focus on cash flow and decentralisation, different sector evolution. The Halma Discipline — Another patient capital allocator, but with more active portfolio reshaping.