The weekly rhythm

SaaS Founder OS pays off through cadence, not marathon sessions. The whole system is designed so that ~an hour a week keeps a multi-product portfolio honest.

The loop

1. Open the Dashboard. Start with Needs Attention — overdue reviews, incomplete scorecards, blocked gates, high unmitigated risks. That list is your agenda.
2. Work next actions, not whims. Each product's Overview names the 1–5 highest-leverage moves. Do one. Experiments are the unit of progress: open the Experiments tab, mark it RUNNING, and give it a metric and a target before you start.
3. Record evidence where it was produced. Filled a playbook section? Ran an interview (a note in the Build Workspace)? Collected a new data point? Enter it immediately — gates read evidence, and evidence you didn't write down doesn't exist.
4. Advance gates when — and only when — they're ready. The bell tells you when a gate's requirements are satisfied. If you override anyway, write the reason; the override and its missing evidence are logged forever.
5. Close the week with a sweep of Reviews. For each product due, record a decision — Build, Validate More, Scale, Maintain, Sell, Park, or Kill — with a rationale. "No decision" is the one outcome the system refuses to normalize.

The Reviews queue — each product due shows its recommendation and the specific gaps behind it. Decide, with a rationale, and move on.

Let the machine carry the numbers

Once a Live Data Feed is connected, MRR, customers, and churn flow in on their own and the dashboards, gates, and reports read the real numbers. You never hand-update a spreadsheet again — the review becomes a decision meeting, not a data-entry session.

When you fall behind

Everyone does. The recovery path is built in: open Reviews, sort by overdue, and make decisions in a single sitting — even a rough one. The decision history forgives imperfect calls; it never forgives silence. A well-reasoned PARK or KILL is a win — it returns hours to the products that can use them.