Don't Interrupt Compounding

The math is thirty for thirty

Eric Glyman frames the durability of the old banking families as a single arithmetic point. "It's not who grew 100% or 200% this year. It's which businesses can grow 30% for 30 years, and if you do that you will be a giant business."¹ His claim for Ramp is that the runway for that kind of uninterrupted compounding is unusually long: the company is roughly doubling annually at scale and holds only about 1.5% of the US corporate and SMB card market, so even a severe deceleration to thirty percent a year for decades is, in his words, physically possible because the market is so big.¹ The point is not to keep doubling. It is that the compounding runway is enormous, and the enemy is anything that cuts it short.

What the families actually do

Glyman reads three behaviors into the families that have been "around for 200 years."¹ The first is to do simple things well, consistently, for a very long time. The second is not to sell, which he states as the direct instruction: "You want to find a business where you can just compound for a long time." The third is to stay in through the cycle. A significant US recession, on his count, arrives every seven to eleven years, and most people sell at the bottom because they cannot absorb more pain or the risk of further downside. The families, he argues, "figure out how to stay in." The interruption, not the downturn itself, is what destroys the compounding.

He names the asymmetry plainly. Early on, with real costs and savings measured in weeks, risk forces a founder to sell when things get frightening; the families had the balance sheet and the temperament never to be forced out. This is the sibling relationship to Don't Sell Your Baby (Founder Control), which is the structural-control version of the same refusal, and to Freedom Number, the personal floor that makes holding on psychologically survivable; Glyman notes his own Paribus exit gave him that floor.

Compounding as an output, and the inputs that produce it

Glyman resists treating the growth number as the goal. "All my stock is in Ramp. It's just a certificate. It's only become valuable because we've built something that makes a lot of people a lot better off. My whole obsession is how do we keep doing that for a very long time."¹ He says openly that he hopes Ramp is the last company he works on, and that his co-founder Karim Atiyeh feels the same.

The earlier of his two accounts supplies the behavioral mechanism. People, he argues, "overestimate 6 months and underestimate 10 years"; a decade feels foreign, but one to two percent compounded daily is how Ramp "kept growing at a consistent rate for a long period."² He points to the Y Combinator weekly ritual of three questions, how much did you grow this week, what is your single biggest real problem, and how will you solve it, as the operational form of the discipline: "You don't need to solve the problems of 10 years from now."² The athlete framing is the same idea, that Super Bowl winners practice fundamentals and focus on the controllable inputs, and the outcomes follow as lagging indicators.

The limit the principle carries

The counterpoint appears inside Glyman's own source, voiced by Sam Parr: "Every time I listen to founders I think I'll own this for 100 years, then a pain-in-the-ass issue comes up and I think, set it up so I can flip it."¹ Glyman does not dispute that the emotions are powerful; his answer is that the long-term intent is a discipline held against them, not a feeling that survives on its own. The principle also assumes a market deep enough to reward decades of compounding, which is the specific condition Ramp's low market share supplies and not a general guarantee. As a claim it is easiest to affirm once the compounding has already happened, which is why Glyman keeps returning it to the controllable inputs rather than the score.