Standards Over Rules

The distinction

Eric Glyman describes standards over rules as his mechanism for keeping a company fast and high-quality as it grows from what he calls "you and Karim by inspection" to thousands of people. "A lot of people try to run their company through rules, under these circumstances you can do this, you can't do that. The best sports teams and the best organizations have really high standards. Craft is kind of a proxy or code word for this."¹

In his framing a rule is a floor. It enumerates the permitted and forbidden, scales poorly because you cannot write a rule for every case, and produces compliance to the minimum, since people clear the bar the rule sets and then stop. A standard is a ceiling that everyone internalizes. When "people's expectations of everyone in the organization are high, for design, for the speed we'll get back to customers, the speed we ship products, the clarity we communicate," the bar is not enforced case by case but expected peer to peer, and it rises on its own.¹

The Steve Jobs example

Glyman's exemplar is late-career Steve Jobs, and the point he draws is about anticipation rather than correction. "Late in Steve Jobs's career, there was very little work that got to him that was truly bad. It was to a certain level, because you knew that if you were going to present to Steve, you needed to be on."¹ The standard does its work before the work reaches the top, because everyone pre-filters to the expected bar. In his contrast, a rule-based organization catches bad work at review, while a standards-based organization rarely produces it in the first place.

How Ramp carries the standard

Glyman describes two mechanisms that let the standard travel. The first is craft as the carrier. "Craft is a proxy or code word for this," so hiring obsessive craftspeople installs the bar in the people rather than in a document. The second is a common rubric that lets autonomous teams share a yardstick. Ramp gives teams latitude over what to build but a shared measure of how good, which he describes as money and time saved per customer: "we can go team by team and look at how much money and time you saved, it allows people on different teams to have a common conversation across very different products."¹

Relationships to adjacent concepts

Standards over rules is closely paired with Quality is Fractal, which supplies the reason the bar shows up at every level at once, and it depends on the kind of high-craft hiring that lets an internalized standard exist at all. It is also distinct from Process vs Bureaucracy in a way Glyman is careful about: process and rules can move a below-average team up to average, while standards are how an above-average team stays at the frontier. And it relates to Simplicity as Distillation, since a standard is only as useful as the taste that defines what great looks like, and when craft is commoditized that judgment becomes the scarce input the standard propagates.

Where it strains

Glyman's account carries its own limits. A standard everyone internalizes presupposes high talent density, because among average people "high standards everyone internalizes" collapses back into needing rules, so the model is not a substitute for hiring well. Standards are also harder to make legible to newcomers and auditors than rules, which are teachable and defensible, and regulated functions may still need explicit rules precisely where an internalized bar is too soft. There is a further failure mode: the Steve Jobs bar can curdle into fear-driven perfectionism if the standard becomes a particular person's approval rather than a shared expectation.