Most roll-ups fail to compound because they are built around acquisition velocity, not integration capacity. The builders who get this right look very different from the ones who do not.
The roll-up thesis is seductive. Buy fragmented operators in a fragmented sector at single-digit multiples, integrate, sell the platform at a higher multiple. The math on a spreadsheet is almost embarrassing.
The math in reality is harder. The platforms that actually compound share three traits. They invest in integration infrastructure — shared services, a real ERP, a single operating cadence — before they need it. They are willing to pause acquisitions for a quarter or two when integration debt accumulates. And they pay for operators, not just deals.
The platforms that fail share the opposite traits. They optimize for deals closed per year. They run each acquisition as a standalone for too long. And they staff the corporate function with deal people rather than operators. The result is a portfolio, not a platform.