A permanent owner should write down when it would sell
Buy-and-hold acquirers win founders with a promise never to sell. I argue that the promise becomes more credible, not less, when the owner publishes the narrow conditions under which it would break it, and that holding on to a business that needs a different owner is not loyalty.
For seventeen years I told founders that Sorrelgate would never sell their company. I meant it, and so did Lucien Archambeau, who founded Sorrelgate and had put that sentence in every letter he sent a prospective seller since 1999. In 2024 I recommended that we sell nine businesses, and our board agreed. My position now is that a company which buys software businesses to own them permanently should publish, in advance, the few conditions under which it would sell one, and that doing so strengthens the promise rather than weakening it.
I know how that sounds to a founder. An owner that announces exceptions to its loyalty seems to be preparing to use them. But an unconditional promise is not really a promise; it is a description of the present mood of the people making it. Boards change, chief executives retire and markets move. A founder who sells to a permanent owner deserves to know what would happen to the business if the owner's view of it changed, and in what circumstances that could happen.
What the promise is worth to a founder
The promise matters because of who sells to companies like ours. Most of the businesses we buy employ between twenty and two hundred people and were built over decades by someone who knows every customer by name. Many founders could get a higher price from a private-equity fund. They sell to us, often for less, because they believe their employees and customers will not be passed on again in four or five years, merged into something unrecognisable or loaded with debt to fund someone else's return. In our own records, roughly two in five sellers over the past decade said at their first meeting that permanence was the main reason they were talking to us at all.
That makes the promise an economic asset. It lowers our price, shortens negotiations and brings us founders who are not running auctions. It would be foolish to dilute it casually. Yet our hospitality decision showed me what the unconditional version costs. Our nine restaurant point-of-sale and reservation businesses had served independent restaurants well for years. From 2021 their customers began moving to systems whose providers earned most of their money from card processing and gave the software away cheaply to win it. Competing would have required us to become a payments business, with the licensing, fraud exposure, hardware financing and capital that implies. Organic revenue fell in 2022 and fell faster in 2023.
We could have held on. Our model tolerates slow decline in a business that still produces cash, and the hospitality group still did. But holding on meant asking 640 people to work in a business we would never invest in properly, while customers drifted to competitors we had decided not to fight. It meant telling restaurants we would maintain their software without telling them we had no plan to improve it. That is not loyalty. It is a slow sale to nobody.
Three conditions, written before they are needed
The policy our board adopted in March 2024 has three conditions, and any one is enough for management to propose a sale. The first is that the business needs a capability we have decided as a company not to build, so that its customers would be better served by an owner who has it. The second is that the business cannot earn a return above its cost of capital under our ownership for the foreseeable future, not because it is badly run but because its market has changed. The third is that a law, licence or customer requirement makes our ownership itself the obstacle, for instance where a buyer must be domiciled in a particular jurisdiction.
Just as important are the reasons that do not appear. We will not sell because a business is growing slowly, because someone offers a high price or because a vertical has gone out of fashion with investors. We will not sell to fund a large acquisition elsewhere. Those exclusions are what keep the promise valuable, because they cover the reasons most owners actually sell.
The policy also governs how we sell. Price is one criterion among several. In 2024 we accepted C$212 million from Pellara Commerce, a payments and restaurant technology company that intended to keep the businesses running, rather than C$231 million from a financial buyer that would have merged them and closed most of the offices. Two years later Pellara employs about 560 of the 640 people who transferred and has kept seven of the nine offices. Our board was entitled to prefer the higher figure, and one director argued for it. We concluded that a permanent owner selling once in twenty-five years would be judged by future sellers on what happened to the people, and that this judgement had a value of its own.
The cost of keeping a business out of loyalty
The strongest objection came from Lucien, and I take it seriously. He argued that founders do not read policies; they hear that you sold something. He was partly right. Four founders we were speaking to in late 2024 asked about the sale, and one withdrew. I had underestimated how fast news travels among founders who know each other through trade associations and customer conferences.
My answer to him, then and now, is that founders also hear when an owner keeps a business it has stopped investing in. Our general managers talk to each other, and so do the families who sold to us. A portfolio with a few businesses being quietly starved tells every founder what permanence means in practice. I would rather they hear that we sold nine businesses once, under written conditions, to an owner that kept most of the jobs.
There is a second cost of loyalty that boards tend to miss. Capital and management attention held in a declining business are not available anywhere else. The proceeds funded eleven acquisitions in our Civic and Education groups in 2025, from founders who chose us partly because we could explain the hospitality sale rather than deny it.
I do not think every permanent owner should adopt our three conditions. A company that owns businesses in fewer, steadier markets may never need them. But every such company should be able to answer a founder who asks directly what would make it sell. If the answer is never, the founder should ask what happens when the market moves. If the answer is that it depends, the founder should ask on what. The honest reply is a short list, written down, approved by the board and shown to sellers before they sign. We now send ours with the first letter.