A founder's long hours need a different employee bargain

A founder can choose to spend most waking hours on a company. That choice does not establish what an employee owes it. Equity, control over priorities and the ability to stop change the bargain. If an organization expects extreme hours, it should describe that expectation before hiring rather than letting attendance become an unstated test of commitment.

Looking back from November 2026, Armin Ronacher's 4 September 2025 essay, “996”, challenged a culture being promoted around startups. The shorthand describes twelve-hour days on six days each week. Ronacher specifically questioned giving 72 hours a week to somebody else's startup and distinguished a founder's position from an employee's. That is his argument about the arrangement, not a finding from a study of employee outcomes.

Compare the choices behind the same schedule

Two people may leave the office at the same time while making very different choices. A founder may decide what the company builds, negotiate funding and benefit directly from ownership. An employee may have some equity but little authority over the direction that determines its value. Describing both as equally committed because both stayed late erases information the hiring conversation needs.

Ownership is not a guaranteed payout. Founders can lose money and employees can rationally value speculative upside. The distinction is not that one role always wins. It is that a salary, an equity grant and control are different parts of compensation and responsibility. A shared vocabulary of sacrifice cannot make those parts interchangeable.

Control over stopping matters too. Someone who can postpone a launch or reduce scope has an option unavailable to someone expected to meet the date. Even founders face customers and investors, so their control is limited. But limited control differs from having no authority over the deadline while carrying the expectation to absorb its consequences through longer hours.

A sprint needs an ending someone can enforce

A bounded period of extra work can be a reasonable choice. An employee may want concentrated experience, a meaningful bonus or a specific launch to succeed. The strongest counterargument to criticism of intense schedules is that adults can knowingly choose such an arrangement. A company need not impose everybody else's preferred pace on a person who wants it.

That defense requires more than enthusiasm during recruitment. Name the expected schedule, its duration and how compensation works. Explain who can reduce scope, what happens after the deadline and how someone raises a concern without first proving loyalty. A supposedly temporary sprint with no enforceable end becomes an ordinary schedule by repetition.

Voluntary participation also has a practical test. If declining an optional evening changes promotion prospects or access to meaningful work, the formal option may communicate less than the manager's behavior. This is a management question about incentives, not a conclusion about the legality of a particular contract. No employee outcomes or workplace investigation are claimed here.

Hire for an agreement people can describe

A useful hiring discussion separates normal working expectations from exceptional events. Explain the ordinary availability window, any support duties and the kinds of deadline that may require a separate agreement. State whether the role can be performed well within that ordinary arrangement. Applicants should not have to infer the answer from photographs of late-night meals.

Describe equity with the same precision. A percentage without vesting conditions or an explanation of what the holder can influence leaves the economic bargain unclear. This article does not value a particular grant. The management principle is to explain the relevant terms rather than use ownership language to imply that every employee has a founder's authority and upside.

Interview behavior should match the agreement. If the advertised role promises predictable hours but every interviewer celebrates exhaustion as the ideal, candidates receive two conflicting messages. Resolve that conflict before asking someone to join. Explicit demanding expectations allow an informed choice; vague expectations leave the most consequential term to be discovered after acceptance.

Judge contribution without turning output into another shortcut

Ronacher favors attention to output over hours at the desk. That observation supports asking what the work achieved, but it does not supply a universal productivity formula. Counting commits or closing tickets can be as misleading as counting evenings if the counts reward fragmented work or ignore prevention, review and support.

A team's contribution includes decisions that avoid unnecessary work. A developer who removes a risky requirement before implementation may produce less visible activity than one who builds and repairs it. Managers need enough context to recognize that difference. Replacing one simplistic attendance score with another simplistic output score would preserve the original problem.

There are limits to the argument. Some work genuinely requires coordinated presence or a response during an incident. An employment agreement can include those responsibilities explicitly. The issue is treating an undefined amount of extra availability as moral evidence about the person. Operational needs can be stated and planned; a loyalty contest cannot be reliably scoped.

For custom software development, realistic scope and named decision makers help make time expectations credible. They do not guarantee an easy project, but they give someone authority to change the work when constraints collide. The founder's personal appetite for intensity can remain personal. Employees deserve a bargain they can understand before they decide whether to share it.