Mark Leonard

There is one operator alive who has built a great public-market compounder almost entirely without anybody outside the software industry knowing his name. His name is Mark Leonard. The company he founded in 1995 with $25 million of Canadian venture money is now one of the largest software businesses in the world, and he has accomplished this by acquiring more than 900 tiny vertical-market software companies, one at a time, holding them forever, and refusing essentially every interview anybody has ever asked him to give.

For the operator who suspects the great fortunes are still being built quietly in industries nobody is watching, Leonard’s career is the most direct evidence. The Constellation playbook is in plain sight. He publishes it himself every year. Almost nobody outside the software industry has actually read it.

A Canadian software fund

The pre-history is unusually thin. Mark Leonard worked in venture capital in Canada in the 1980s and early 1990s, and somewhere along the way he developed the conviction that the most underpriced asset in technology was the small, niche vertical-market software business. The thesis went like this. Almost every industry, no matter how small, runs on a piece of dedicated software written specifically for it. The companies that wrote those programs were typically tiny, founder-owned, sticky, and trading at low single-digit multiples of earnings when they traded at all. Most of them were not for sale. But every year, some of them were.

In 1995 he raised about $25 million from a syndicate of Canadian venture investors and founded Constellation Software with the explicit purpose of buying these companies, one at a time, forever.

The first decade, in private

For its first ten years, Constellation was a quiet private company. It made its first acquisitions in industries most people in technology had not heard of. Software for school bus dispatchers. Software for hotel front desks. Software for parking lot operators. Software for pawn shops, automotive parts distributors, dental practices, marina managers. The acquisitions were typically in the $1 million to $20 million revenue range, bought for roughly five to seven times EBITDA, and they kept their existing names, their existing managers, and their existing customer relationships.

The model worked because nothing else interfered with it. Constellation did not centralise sales, did not rebrand the acquired companies, did not extract synergies in the conventional sense. It did three things: provided capital, provided a slow and gentle improvement on operating practices, and held forever.

Going public, by accident

The IPO on the Toronto Stock Exchange in 2006 was unusual. Leonard had not particularly wanted to go public. The early venture investors needed liquidity, and the cleanest path was a small listing. Constellation went public at about CA$17 per share. By 2024 the same share was trading at over CA$4,000. The compound annual return since IPO is among the highest of any publicly listed company in the world.

The IPO had a useful side effect for the operator audience. It forced the company to publish quarterly results and annual reports. From 2007 onward Leonard wrote the most candid and useful capital-allocation letters of any public-company CEO since Buffett. The letters explain, in detail, the hurdle rates Constellation uses, the multiples it pays, the way it allocates capital between its operating groups, and the reasoning behind every major decision.

Reading them is the closest you can come to spending a year inside the world’s most disciplined software rollup.

The operating-group structure

By 2009, with more than 100 acquisitions complete, Leonard had divided Constellation into operating groups: Volaris, Harris, Jonas, Vela, Perseus, and Topicus. Each group runs its own pipeline, makes its own acquisition decisions, runs its own portfolio companies, and has its own head office. The Constellation parent does only two things: allocates capital between the groups based on returns on capital and growth opportunities, and enforces a small set of corporate-culture standards.

The result is structurally interesting. Constellation has roughly 900 acquired businesses, organised into six operating groups, with effectively no centralised operations. It is a holding company in the cleanest sense of the word. The advantage is scalability. The disadvantage is that almost nobody else can replicate it, because almost nobody else is willing to truly leave acquired companies alone.

The Leonard discipline

Across thirty years and 900 acquisitions, a small set of principles surface repeatedly in the shareholder letters.

The acquisition discipline is mechanical. Every potential acquisition is screened against a hurdle rate. If the projected after-tax return does not clear the hurdle, the deal does not happen, regardless of how attractive the target looks otherwise. Leonard has walked away from large transactions that other software acquirers were happy to bid up.

The operating discipline is similarly mechanical. Acquired companies keep their name, their leadership, and their autonomy. The head office’s job is capital allocation, not operating decisions. Operating groups can spend their own retained earnings without parent approval; new equity allocations from the parent are reviewed against the hurdle.

The communication discipline is unusual. Leonard has chosen written communication over almost any other form. The annual letters are long, careful, often self-critical. Quarterly conference calls were eventually eliminated, replaced by written Q&A. The only direct contact most institutional investors have ever had with him is through these documents.

The lifestyle discipline matches. He has refused virtually every interview request. He has cancelled his own salary when his board’s compensation committee proposed an arrangement he considered excessive. He lives, by all credible reporting, modestly relative to his fortune. The contrast with the 1980s raider archetype is total and intentional.

What it teaches the next operator

The Constellation playbook is unusually clean to extract because Leonard has been so explicit in writing.

Choose an industry with thousands of small, recurring-revenue businesses that almost nobody else is studying.

Set a hurdle rate for acquisitions and refuse, every time, to violate it.

Buy small. Keep the management. Compound.

Use public capital markets to fund the platform but stop trying to manage the share price. Reinvest retained earnings into more acquisitions and let the price take care of itself.

Write the letters. Communicate seriously with shareholders. The right ones will find you and stay.

It is, in retrospect, the most boring possible roll-up strategy. It is also one of the most lucrative ever executed. The operator who wants to build a long-duration compounder, rather than a high-velocity trade, has no better single source than Constellation’s annual letter archive. Start with the 2007 letter and read forward. The playbook is right there.