The Whiteboard Never Lies. That Is the Problem.
Every roll-up pitch contains the same three lines. Add-ons trade at lower multiples than the platform, so every acquisition is accretive on day one. Shared infrastructure cuts costs nobody could cut alone. And a bigger, more diversified group deserves a bigger multiple at the end. All three lines are true. That is exactly what makes the strategy dangerous: the math keeps working on paper long after the business underneath has stopped working in fact.
A buy-and-build, a roll-up, a platform consolidation: the names change, the physics do not. What follows is the failure catalogue we underwrite against, learned the only way this catalogue can be learned.
The Six Ways a Roll-Up Dies
1. Buying revenue, not capability
The easiest number to grow in a consolidation is revenue; you can simply purchase it. The question that separates an acquisition from an accumulation is what the add-on brings besides its invoices: customers you could not reach, a capability you did not have, density in a geography you wanted. Scale that adds no advantage is just size, and size alone has never earned a premium multiple for long.
2. Integration debt
This is the one that kills quietly. Every acquisition closed before the previous one is absorbed leaves a residue: a second invoicing system, a third price list, two sales teams calling the same client. Like technical debt, integration debt accrues interest invisibly and comes due at the worst moment, when the market turns, or the one person who understood both systems resigns. The platforms that survive treat integration capacity, not deal supply, as the real constraint. There are always more add-ons to buy. There is rarely more organization to absorb them.
3. Fragile leverage
Debt makes the arithmetic prettier and the company weaker. A capital structure that only works if every add-on performs, every synergy lands, and no customer leaves is not a structure; it is a bet on perfection. We structure for the year something goes wrong, because in a ten-year consolidation there is always such a year.
4. Losing the people who were the value
In an SMB, the moat usually has a first name. The founder who knows why every client stayed, the operations manager everyone trusts. Roll-ups that treat acquired companies as line items watch that value walk out quietly over eighteen months, and the spreadsheets notice two years too late. Keeping the people who made each business work is not a soft preference. It is asset preservation.
5. Multiple-paid drift
The entry-price discipline that makes the whole thesis work erodes one small exception at a time. A competitor shows up, an auction starts, a founder hears what the platform paid last time. Each add-on bought one turn too high is a quiet transfer of the strategy's margin of safety to a seller. Walking away is part of the playbook, which is easy to write and hard to do with a pipeline to feed.
6. The exit clock
Consolidation runs on a long clock: find, close, integrate, repeat, for years. A fund runs on a short one. When the two clocks disagree, the fund clock usually wins: integration gets compressed, add-ons get rushed to bulk up the story, and the group is sold as a collection of deals rather than a company. We wrote about this at length in what no fund clock changes; nowhere does it matter more than here.
What the Disciplined Few Do Differently
The successful minority is not smarter. It is more boring, in five specific ways.
- They pick the platform like it is the whole deal. Resilient cash flows, a defensible position, and an operator with the judgment to lead consolidation, not just the appetite to acquire. A mediocre platform with a great pipeline is still a mediocre company.
- They map the fragmentation before buying into it. A repeatable universe of add-ons with one clear logic, instead of a string of opportunistic one-offs.
- They protect the entry multiple like a covenant. The arbitrage is the thesis. Pay it away and what remains is an expensive collection of small companies.
- They integrate like it is the product. Systems, data, culture, sequenced at a pace the organization can absorb. The deal is the easy part; everyone forgets that until the third one.
- They let time do the heavy lifting. The compounding in a consolidation shows up in years five through ten, which is precisely when most owners are preparing to leave. Patient capital is not a luxury here. It is load-bearing.
The Reading That Holds Up
Two books say the quiet part well. Lessons from the Titans is the closest thing to a textbook on serial acquisition done right, and every failure mode above appears in it with a name and a date attached. The Outsiders explains the capital-allocation temperament underneath. Both are on our bookshelf, next to the rest of what we hand to operators.
Where We Sit
Santa Marta backs operator-led platforms in Spain and the United States, in minority or control positions depending on governance, and the team has helped build these businesses from the inside: sourcing, financing structure, add-on execution, and the unglamorous work of integration. Our capital is permanent, which means the integration calendar belongs to the platform, not to a fund's countdown.
If you are an operator pursuing a buy-and-build, or seriously considering one, the long version of our playbook is here, and the door is open early, before the platform, before the thesis is fully formed. We have made these calls ourselves.
"There are always more add-ons to buy. There is rarely more organization to absorb them. Every roll-up that died believed the opposite."