Panic Early

The heuristic

"If you're going to panic, panic early." Eric Glyman's rule for the moment a business starts dying, formed during the crisis at his earlier company, Paribus, when retailers began sending cease-and-desist letters over its price-adjustment refund service.¹

The situation it came from

Revenue did not vanish overnight, which is what makes the heuristic useful rather than merely dramatic. It showed up on the dashboard as a slow slide, refunds slowing by 2%, then 5%, then 10%, as individual retailers quietly withdrew their price-adjustment policies, and support agents started telling customers the policy no longer existed. Roughly 80% of revenue disappeared over the episode, with about a dozen people depending on Glyman at the time. He is candid that the hardest part was psychological and partly performative: "I'm nervous, but how will I face my team? Should I be nervous about this and inspire focus? Should I be confident, inspire action? Even the psychology of working through that was tough."¹

The content of the heuristic

Panicking early is not about the emotion, it is about refusing the delay. Be honest immediately rather than hoping the number resolves on its own. Tell the team the entire situation rather than protecting them from bad news, a practice consistent with the transparency Glyman later built into Ramp's own idea-meritocracy. Convert the fear into problem-solving: "a lot of building a business is being in a business of problem solving. You're going to be put in scenarios you didn't anticipate, and you have to figure it out, and there are paths." Then keep going, a sentiment he pairs with Churchill's line about going through hell.¹

The implicit failure mode is the opposite pattern: seeing the number move, telling yourself it is noise, shielding the team from the news, and arriving at the actual crisis with less time and less trust than an earlier, honest reaction would have preserved.

The arithmetic that made it survivable

Worth recording alongside the psychology, since it is why the response could be strategic rather than desperate: the company was growing more than 40% a month at the time, "which means you're doubling in less than every two months. You're actually not that far from being right back to where you were." An 80% drawdown against that growth rate works out to roughly five months of lost progress rather than an ending. Fast growth functioned as its own form of crisis insurance.¹

Elsewhere in the same practice

The same instinct reappears at Ramp as a deliberate habit: Glyman and Karim Atiyeh argue over metrics three to twelve months out on purpose, the agony that sits underneath what looks from outside like a smooth growth curve.² Glyman describes the earlier Paribus loss, in a separate telling, as a "known-but-unmanaged risk," meaning the actual failure was not the surprise itself but having seen the risk coming and not acted on it in time.