Weekly operating cadence

~8 min read

Why it matters

Early companies don’t fail in a dramatic strategic moment. They fail in weeks where nobody can say what was supposed to happen, what shipped, or who talked to a user. A cadence is a short, repeating meeting with yourself (and your cofounder, if you have one) that forces those three facts into the open. It is not a planning framework and not OKRs for a 200-person company. It is how two people stay honest.

Every earlier page dumps into this one: the user quota, the ship, the price conversations, the metric, the cash.

Core concepts

One primary outcome for the week. Not a theme. A result you can point at on Friday: “two treasurers finished a reconciliation” or “price said out loud on three calls” or “runway spreadsheet updated and one cost cut.” If everything is the priority, the semester is the priority and nothing finishes.
A written plan before Monday gets noisy. A few lines: the outcome, the two or three tasks that cause it, who owns each, what you are explicitly not doing. This replaces a backlog tool until the backlog tool would save you time, which is later.
Midweek is for the outside world. Block time for user conversations and for the one distribution motion. Building all week and “doing outreach Friday” means outreach dies every time the build slips. See Distribution.
Friday is a review, not a status performance. What shipped that a user could touch. What you learned (a quote, a number). The primary metric and cash, even if they’re ugly. What you will stop. Then the next week’s one outcome. Thirty minutes. Standing up is fine. A slide deck is not.
Write it where the other person will see it. A shared doc, dated, one section per week. Memory is kind; the doc is not. When you disagree, you are disagreeing with last Friday’s words, which is a better fight than a vibe.
The cadence survives fundraising, exams, and “crazy weeks.” Those are the weeks it exists for. Shrink it (15 minutes, three lines) rather than skipping it. Skipping is how a two-week slip becomes “we should regroup next month.”
If you are an early hire, ask to be in the review or to read the doc. A company that will not show you the weekly truth is telling you there isn’t one. If you are a founder, the first hire should see the ritual in week one so they learn that shipping and users are the culture, not the all-hands speech.

What to do next

Create one running doc titled “Weekly.” Add this week’s date, the one outcome, and the not-doing list. Share it with your cofounder tonight.
Put two recurring holds on the calendar: a Monday 20-minute plan, a Friday 30-minute review. Protect the user-conversation block on a midweek day the same way.
Friday template, copy it every week: shipped (user-visible), learned (quote or number), metric (activation / retention / revenue), cash and months of runway, stop-doing, next outcome.
Run it for four weeks before you change the template. The point is repetition. A new productivity system is procrastination.
If you miss the outcome, write why in one line (scope, fear of talking to users, a class deadline) and cut next week’s outcome in half. Do not write a longer plan.
Once a month, reread the four Fridays. The pattern — always shipping, never learning, or the reverse — is the actual strategy review.

Mental model

flowchart LR
  mon[Monday: one outcome and a not-doing list] --> mid[Midweek: users and the one channel]
  mid --> build[Build only what this week's outcome needs]
  build --> fri[Friday: shipped, learned, metric, cash]
  fri --> mon

The week is a loop with a written residue. Strategy is what the residue says you keep repeating.

Common failure modes

A goal list with eight items. That is a wish list. Force a single outcome or the review becomes “we were busy.”
Reviewing effort instead of evidence. “Worked hard on onboarding” is not a Friday line. “One user finished, two got stuck on the import” is.
Tools instead of the meeting. You do not need a new app to write six lines. Adopt a tool when the doc actually hurts.
Cadence as performance. Sanitized updates that hide a bad cohort or a scary runway date defeat the point. The doc is for the people doing the work. Be blunt in it.

Watch

Setting KPIs and Goals | Startup School — Y Combinator. How early teams pick a primary metric and set a target without drowning in secondary numbers. Use it to choose the one outcome, then come back to the Friday template.
Lecture 14 - How to Operate (Keith Rabois) — YC Root Access. Operating cadence as the company gets slightly less tiny: what you track, what you edit, and how information moves. Steal the “write it down” discipline now; ignore the parts that assume a larger team.