Loading page…
A Montreal holding company that bought 130 small vertical software businesses instead of building one, and runs them past $750M of revenue by bundling in payments, offshore dev capacity and a shared AI platform
130 companies owned · 3,000+ employees · first acquisition bought for $5M now throws off $10M+ EBITDA
The first deal is the entire thesis. A $5M purchase of a flat hotel-management product doing about $1–1.5M of EBITDA was at $3M eighteen months later and is above $10M today, with a $50–60M hospitality suite built around it. "All the increases in profitability came from revenue growth."
Fewer bars = easier, cheaper, or faster for an AI-assisted solo builder. Editorial judgments based on the case details.
Sam Youssef calls himself a lifelong entrepreneur — he started by shovelling neighbours' driveways as a teenager, then contracted the shovelling out to other people. Around 2011 he put some of the money from earlier ventures into an investing company called Valsoft Capital. He had read Snowball, the Warren Buffett biography, and wanted a career in investing and deal-making. The first version of that was buying cheap value stocks, and it went badly: "I lost a bunch of money doing that. It wasn't working." The correction was to stop screening on price and start screening on quality — well-managed companies with high revenue predictability, mission-criticality with their customers, and pricing power. In 2011 and 2012 that screen kept landing on software, not by design, and Valsoft Capital rode the SaaS boom. One of its largest positions became Constellation Software, and reading its founder gave Youssef the idea to run a private version of the same machine. So they bought one company to see how it went: a hotel-management software business in South Florida, $5 million. The year after that they bought three more, then five, then eight. By 2022, after taking the business to roughly $200 million on their own capital, some of the largest institutional investors in America backed it. At the time of this interview Valsoft owns 130 companies, has "over $750 million in revenue," more than 3,000 employees, and is marking ten years of the acquisition business. The method has not changed in a decade: buy good companies in rationally structured markets, grow them, take operating leverage on that growth, then redeploy the profit into whichever kind of growth — organic or inorganic — that market rewards.
In a roll-up the customers arrive with the deal, so the real cold start is finding owners who will take your call. Youssef is blunt about how that went: the first year was very hard to get a response, the second year a little easier.
The quick-reference card — acquisition channels, replication playbook, risks & traps, and our verdict — is members-only.
Join the Unicorn Club — $5/moCancel anytime, in one clickPaste into Claude Code / Codex and get a working version of this product end-to-end.
Two production-grade prompts per idea — a full build spec and an SEO growth plan — plus 10 tailored build specs a week. Members only.
Join the Unicorn Club — $5/moCancel anytime, in one clickEvery claim traces back to a listed source — open any of them and check the number yourself. The exact words are quoted below, with a timestamp that jumps to the second each figure is said.
“Today we have uh 130 companies uh over 750 million dollars in revenue.”
15:53 · hear it in the video“A year and a half later this company that we bought for 5 million bucks is making 3 million bucks of EBITDA.”
25:55 · hear it in the videoData credibility: Founder interview on a podcast. Revenue, deal count, employee count, acquisition prices and EBITDA figures are all stated by Sam Youssef himself and are unaudited — Valsoft is private, so none of it is verifiable from filings. The hosts open by describing Valsoft as having a multi-billion-dollar valuation; that is the hosts' characterisation, not a disclosed figure, and no valuation was stated by Youssef. The deal count is given inconsistently in the same conversation: the host says over 130, Youssef says 130 companies in one answer and "we've done it 140 times" in another, so treat it as roughly 130–140 transactions. Revenue is given as "over $750 million" by Youssef after a host says $700 million, and Youssef corrects the host's use of "ARR" to "revenue." The funding rounds mentioned in the show's news segment (Beacon, Chip Agent and others) are financings, not revenue, and are not used here.
The first dollar was somebody else's revenue, bought.
Two loops compound at once. The capital loop is visible in the deal count — one company, then three, then five, then eight — each round funded by the cash flows of the last plus gains from the public-market portfolio, until outside institutions came in at roughly $200 million of self-funded scale.
The genuinely dead stretch came before software: years of buying cheap stocks because that is what Buffett had done, and losing money at it.
Valsoft is not one product, it is a container for 130 of them. The companies it buys are vertical-market software — systems of record for narrow industries like hotel management, forestry, medical practices, revenue cycle management — usually decades old, usually founder-owned, usually with a defensible position in a market too small to attract venture money. Youssef's target band right now is companies doing between $3 million and $100 million. Below $3 million the R&D budget is often too thin to defend the market position, unless the vertical itself is tiny; the bigger the vertical, the less interesting it becomes, because large addressable markets attract capital chasing exponential returns that will "compete with us irrationally." Operationally, the group runs fully decentralised. Each business is led by a general manager, usually the former founder or a long-tenured insider with real vertical expertise, and HQ does not push product strategy down. What HQ builds instead is shared infrastructure that no single portfolio company is big enough to build alone. Our read: that is the actual product Valsoft sells to the companies it buys — payments rails, a bench of engineers who know modern tooling, an AI platform with voice agents as the first module, and seven proprietary integration playbooks that are discussed with the founder before the deal closes, so the buyer and the seller have already agreed what the first year looks like.
Three ways, stacked. First, entry price. Valsoft has paid anywhere from 1x ARR to 10x ARR, and Youssef is explicit that the multiple is a function of what he is buying: a product he badly wants in an industry where he already owns a large stable of customers is worth paying up for, while a declining business that will hurt the organic growth profile and needs a rewrite gets a very low multiple because the risk sits with the buyer. Second, operating improvement. The first acquisition went from about $1–1.5 million of EBITDA to $3 million in eighteen months and above $10 million today, and Youssef insists all of that profitability increase came from revenue growth rather than cost-cutting — new products sold into an existing base, plus a heavier development investment than a retiring owner was willing to make. The hospitality suite built out from that single core is now $50–60 million of revenue on its own. Third, the multiple spread between what small private vertical software costs and what a $750 million diversified group is valued at. Youssef pushes back on the easy version of the second lever, though: selling more products into the same people is the most profitable use of an R&D dollar, especially if you own the system of record, but it is "the lazy way a little bit" and it erodes market share over time if you are not also winning new customers.
The cost structure is the interesting part, because it is what an individual portfolio company could never afford. Valsoft carries over 100 people whose full-time job is finding deals, spread across 25 countries, plus coverage from the US and Canada — that is a permanent fixed cost justified only by volume. It runs large development centres in Sri Lanka, India and elsewhere staffed with engineers who know current tooling, and can put 20–25 of them onto one acquired application to strip out tech debt so the product can iterate again. It owns payment infrastructure that gets bundled into portfolio companies as a new revenue line. It runs an internal AI lab with a core platform that lets a general manager launch voice agents for their industry without building the plumbing, and Youssef says GMs have shipped those in several different industries already. Our read: none of these are things the $3–10 million software company being acquired could have built, and all of them are things it immediately needs, which is why Valsoft can honestly present itself as an operator rather than a financial buyer. On the AI question Youssef gives a counterintuitive answer about R&D budgets: set up properly, they should go down, because you can do more with less — and if your stack cannot support that, "you better get there quickly."
The copyable part is the screen, not the scale. Youssef looks for narrow verticals where the market is structured rationally, meaning nobody is dumping venture money into it chasing a 50x outcome, and where a system of record has enduring economics because the software is deeply integrated into workflows and the vendor employs actual industry experts who function as consultants. At a solo scale that translates into buying a $200K–$500K ARR vertical tool from a retiring owner rather than starting from a blank page — the same screen, three zeros down. Change three things. First, forget the 1x–10x range and buy at the low end, because you have no ecosystem to bolt on and the spread is your only margin for error. Second, replace the seven playbooks with the one or two you can actually run: modernising the product so it can ship features again, and adding a second revenue line the incumbent never bothered with. Third, take Youssef's integrity point literally — the seller knows far more about the business than you do, and gauging the quality of your counterparty matters more than the model, because "people of high integrity build businesses of high integrity and vice versa." Our read: the underrated detail is that the playbooks are discussed with founders before they sell. That conversation is the diligence.
Youssef names the failures himself: 25–35% of acquisitions have landed below expectations, driven by subscale companies that deteriorated and by customer concentration where losing one large account broke the economics. At small scale that failure rate is fatal rather than annoying, which is the single biggest reason not to copy this with one deal and all your capital. The AI question is genuinely open. The bear case, which Youssef attributes to a widely heard argument that vertical SaaS is dead because end customers will build their own tooling, he rejects for vertical software specifically: the software is only part of what the customer buys, alongside integrations, workflows and vendor expertise drawn from serving hundreds of similar customers, and the software's share of the customer's cost base is too small for replacement to pay. He concedes the risk for thin pieces of software and for horizontal tools. The real danger he flags is tech debt: businesses that cannot adapt get challenged because the rate at which technology is produced over the next ten years will be significantly higher than the last ten, and a vendor that cannot ship the AI-enabled features its customers now need will watch those customers become competitively disadvantaged and switch. Our read: the roll-up structure makes that a portfolio problem rather than a single-company problem, which is exactly the protection a solo acquirer does not have.
The four founder playbooks — first customers, first dollar, the flywheel, the silent stretch — are in the Unicorn Club, on this idea and on all 1,435 others. The story, the numbers, the sources and every later section stay open to everyone.
Join the Unicorn Club — $5/moCancel anytime, in one click