Why roll-ups fail

A roll-up is simple to describe and brutal to execute: buy many small companies in one industry, combine them under one roof, and run them as a single larger business. On a spreadsheet it looks like free money. In practice most roll-ups underperform, stall, or come apart, and the reason is almost never the math. It is everything the math quietly leaves out.

I am building toward this model on my own account, so I read the failure literature as an operator, not a spectator. The pattern is consistent enough to be uncomfortable. The buyers who blow up are rarely bad at arithmetic. They are good at arithmetic and bad at the day after the close, which is where a roll-up is actually won or lost.

What a roll-up actually is

The mechanics are plain. You pick one industry, usually fragmented and full of small owner-run businesses, and you acquire them one at a time. Each one keeps operating. You bolt them together into a group that shares a back office, purchasing power, and a name. Do it enough times and you have turned a dozen small companies into one medium-sized one.

The appeal is that these industries are everywhere. Trades, services, local operators nearing retirement with no succession plan. Good businesses, quietly profitable, run by people who never built a second layer of management. On paper a buyer who can absorb ten of them looks like a genius. The trouble starts the morning after the first deal closes, when the spreadsheet ends and the operating begins.

Roll-up

Buying many small companies in one industry and combining them into a single larger group under shared ownership.

Multiple arbitrage

The gap between the low earnings multiple you pay for a small company and the higher multiple a larger, more durable group is worth.

Integration debt

The unmerged systems, pricing, and processes every acquisition adds. Invisible on the balance sheet, and the thing that actually sinks most roll-ups.

The arbitrage everyone shows you

The engine every roll-up runs on is multiple arbitrage. Small companies trade cheap because they are risky to own: one key person, a concentrated customer list, no systems. A buyer might pay three or four times earnings for one. But a larger, diversified group with real management is worth more per unit of earnings, so it gets valued at a higher multiple. Buy low, combine, and the whole is suddenly worth more than the sum you paid. The gap is the arbitrage, and it is real.

That is the pitch, and it is not wrong. The mistake is treating the re-rating as automatic. The higher multiple is a reward for having built something genuinely more durable than the parts. If you have only stapled the parts together, the market eventually prices you exactly as what you are: a pile of small risky companies carrying a lot of debt. You do not get the arbitrage for buying. You get it for integrating, and integrating is the part nobody puts in the deck.

The arbitrage is the last thing that actually happens, not the first.

Where roll-ups actually die

There is no single cause of death. There are five, they compound, and any two of them together are usually enough. Go through them slowly, because this is where the money is actually lost.

Integration never happens

Two companies doing the same thing still run on different software, different pricing, different quoting habits, different ways of paying people. Merging that is slow, unglamorous work that no spreadsheet models and no press release celebrates. So it gets deferred. The team is busy, the next deal is more exciting, and the integration slips a quarter, then a year. What you own is a holding company full of silos that happen to share a logo, carrying all the cost of scale and none of the benefit.

The founder walks out with the business

In a small company the owner often is the company. The relationships, the reputation, the judgment on the hard jobs, all of it lives in one head. You buy the business, the founder cashes out, and eighteen months later he is gone and the customers he personally held are drifting to whoever he plays golf with now. You paid for earnings, and the earnings had legs. This is owner dependence priced at acquisition scale, and it is the single most expensive thing most buyers fail to check.

The debt that made the math sing

Leverage is what makes the arbitrage look spectacular on paper. It is also what turns one soft quarter into a solvency problem. Debt does not care that integration is behind schedule or that a founder left early. The payment is due either way. A roll-up financed to impress a lender has no room to absorb the ordinary bad luck that every operating business eventually meets, and roll-ups meet more of it than most, because they are ten businesses stacked together.

Buying faster than you can digest

Speed is the quiet killer. Closing the next deal before the last one is absorbed feels like momentum, and to outsiders it looks like progress. Underneath, the integration debt is compounding faster than the earnings, and every new logo makes the backlog worse instead of better. The acquirer confuses a growing revenue line with a growing business. They are not the same thing. One is a number. The other is a system that can carry the number.

No operating layer underneath

Underneath all four of the above sits the real gap. Nobody built the shared system that was supposed to make the group worth more than the pieces. There is a name, a lawyer, a lender, and a spreadsheet, but no operating layer: no single way to quote, schedule, bill, report, and run the work. Without it, integration cannot happen even when someone finally tries, because there is nothing to integrate the companies into. The re-rating was always downstream of a system that was never built.

Two roll-ups, same industry

Picture two buyers going after the same fragmented trade in the same year. Same deals on the table, same lenders, same retiring owners. One compounds into something durable. The other spends three years looking busy and quietly unwinds. Nothing separates them at the start except how they behave after each close, and that difference is the whole game.

The failing one is not run by fools. It is run by people who believed the deck, treated the arbitrage as the starting point instead of the finish line, and mistook the ease of buying for the ease of owning. Everything they did was rational given a wrong assumption: that the re-rating was theirs the moment they signed. It never was. It has to be earned, one integrated company at a time, and the earning is slow.

What the survivors do differently

The roll-ups that work are boring on purpose. They do four things in a specific order, and the order is not optional.

1. Install before you buy

   Build the shared operating system first, so there is something real for each company to plug into. Not the acquirer's ego stamped on a business, a genuine spine: one way to quote, schedule, bill, report, and pay people. The failure mode is buying first and hoping to build the system later, under deal pressure, which never happens.

2. Go slow on purpose

   Digest one company fully before buying the next, so the operating layer stays ahead of the deal flow instead of drowning under it. The failure mode is treating a fast deal count as success and letting integration debt compound in the dark.

3. Keep the people

   The relationships and the judgment are the asset, so structure every deal to keep the operator in the seat and motivated, not cashed out and gone. The failure mode is buying the earnings and letting the person who generated them leave with the customers.

4. Hold one industry

   Focus is what lets a single operating system fit every company you buy, so a tenth acquisition looks like the ninth. The failure mode is chasing cheap deals across unrelated industries until no shared system can fit any of them.

Notice that going slow and holding long are the same discipline seen from two angles. A buyer forced to sell inside a fund's clock cannot afford to digest one company before the next, because the clock is ticking on a return. A buyer with permanent capital can let the operating layer set the pace. Structure is upstream of behavior. Fix the structure and the patient behavior gets easier to hold.

Before you buy the first company

You can screen yourself out of a doomed roll-up before you ever sign, and it is cheaper to do it now than after the second close. If you cannot honestly tick all of these, you are not ready to buy the first company, let alone the tenth.

• You have run something messy and made it run better, not just modelled it in a sheet
• You own one operating system worth plugging a company into before you buy the first
• The industry is fragmented enough that a tenth deal still looks like the first
• Every deal is structured so the operator stays and stays motivated
• The debt survives a bad quarter, not just a good spreadsheet
• You can genuinely go slow, because nobody is forcing you to sell in five years

This is the same reading I bring to how Orevida buys: the operating layer comes first, the deals come at a pace the layer can carry, and the people who run each business are the asset I am paying to keep. A roll-up that skips any of those is not a faster version of the same strategy. It is a different, worse strategy wearing the same name.

You earn the right to buy by having something worth plugging companies into, and you earn the higher multiple by holding long enough to prove the group is more durable than its parts.

An honest note on where I stand

This is the model I am building, deliberately. I have spent ten years making businesses run without the owner in the room, and that operating layer is the whole reason to attempt a roll-up at all, not a nice-to-have I will add later. The order matters more than anything else in the strategy: build the system, then buy into it, then hold long enough for the market to agree the group is durable. Anyone who tells you the arbitrage is the easy part has never tried to integrate the second company.