How to Install an Execution Operating System in a SaaS Company

A founder told us he was scheduling a company-wide meeting to get real about results. Everyone would leave knowing that hanging around was not an option — you perform or you are out. Asked what operating system the company runs on, he had nothing. The question landed flat. Described plainly, it drew the answer every founder gives: yes, we have dashboards and accountability. Asked how, the answer was flat again. That team will churn and the year will miss revenue and profit, and the meeting will be the thing everybody remembers about the quarter.

The experienced managers are the ones most likely to skip this, and they skip it for a reason that is half true. They have run teams before. They have carried a P&L, run a function of a hundred people, sat in somebody else's system for a decade. So they already know how to run the meeting, and they are right about that. What they miss is that they were inside an operating system rather than responsible for installing one, and that everything they now do by instinct was legible to the people around them because somebody had written it down.

Your team cannot read your head. Thirty people cannot each independently derive what matters this quarter from watching you be decisive. The system is not for you. It is for every person who is not in the room when you decide something — which, in any company big enough to need a leadership team, is nearly everyone.

Every company already has an operating system. Work gets prioritized somehow, numbers get watched by someone, people get held to things or they do not. The only question is whether that system is written down and owned, or whether it lives in one person's head and gets rebuilt from scratch every time that person is busy.

The goal: One named framework, installed whole — accountability chart, weekly scorecard, quarterly priorities, a fixed meeting rhythm, documented core processes — running in one platform and cascaded one level below the leadership team.

Background

An operating system is the set of mechanics by which strategy becomes what somebody does on Tuesday. It is not a strategy, not a culture, and not a set of dashboards. Strip any of the three frameworks below and the same eight parts are underneath.

1. A written plan, short enough to repeat. Where the company is going long term, where it will be in three years, what it must do this year, and what it will not do. One page. If the leadership team cannot say it without the page, it is not installed.
2. An accountability structure. Every function of the business is a seat. Every seat has exactly one name in it and a small number of outcomes that name owns. This is not the org chart of reporting lines — it is the map of what must get done and who is answerable when it does not.
3. A scorecard. Five to fifteen numbers reviewed weekly, each with one owner and a target beside it. Weekly and leading, because a monthly number tells you about a quarter you can no longer change.
4. A short list of priorities. Three to five things the company will finish this quarter, and no more than three to five per person. Everything else is explicitly waiting. A list of twelve priorities is a list of none.
5. A meeting rhythm. Same day, same time, same agenda, and it does not move for a customer. Daily where it earns its place, weekly for the leadership team, monthly for the financials, quarterly to reset the priorities, annually for the plan.
6. An issues discipline. A visible, ranked list of everything in the way, and a meeting that spends most of its clock killing the top of it. Identify the real issue, discuss it once, solve it. The alternative is the same four issues resurfacing every week for a year, which is what most leadership meetings are.
7. Documented core processes. The six to ten processes the company actually runs on, written to the twenty percent of steps that drive eighty percent of the outcome, and followed by everyone.
8. One place where all of it lives, visible to the whole company rather than to the leadership team.

Miss any one of the eight and the system degrades in a predictable way. No accountability chart and every issue becomes a negotiation about whose problem it is. No scorecard and the quarter is scored in hindsight. No issues list and the weekly meeting becomes a status report, which is the most expensive way ever devised to send an email.

The three frameworks, and the book behind each. Read the book before you design anything. All three work. The blend you invent yourself before running one of them straight for four quarters does not.

EOS — the Entrepreneurial Operating System. Gino Wickman, Traction: Get a Grip on Your Business. Six components — Vision, People, Data, Issues, Process, Traction — and a small, deliberately unglamorous toolkit: the Vision/Traction Organizer, the Accountability Chart, a weekly Scorecard, quarterly Rocks, the Level 10 weekly meeting, and IDS for working the issues list. It is the most prescriptive of the three and the most complete out of the box, and it was designed so a 10-to-250-person company could run it without a consultant, though certified implementers are widely available and worth it for the first two quarters. It is lighter on strategy and on cash than the others. For a vertical SaaS company installing its first system, this is the default.

Scaling Up — the Rockefeller Habits. Verne Harnish, Scaling Up: How a Few Companies Make It…and Why the Rest Don't. Organized around four decisions every growing company has to get right — People, Strategy, Execution, Cash — with the One-Page Strategic Plan at the center, the 7 Strata of Strategy behind it, a Critical Number and theme each quarter, the Rockefeller Habits checklist, and a daily-weekly-monthly-quarterly-annual rhythm. It carries far more strategic and financial instrumentation than EOS, which is exactly why it suits a company whose binding constraint is strategy or cash rather than accountability. It has more surface area, so it is easier to half-install.

OKRs — Objectives and Key Results. John Doerr, Measure What Matters, and behind it Andy Grove's High Output Management, where the method started at Intel before Doerr carried it to Google. An Objective is a qualitative, ambitious statement of what you are going after; three to five Key Results are the measurable outcomes that prove you got there. Set quarterly, made transparent across the company, graded honestly at the end. It is the best of the three at making ambition specific and at aligning a product organization.

Be clear-eyed about the last one. OKRs are a goal-setting framework, not an operating system. They say nothing about who owns which seat, which meeting the numbers are read in, how an issue gets resolved, or what the rhythm is. Adopt OKRs and you have installed part four of the eight and still owe yourself the other seven. Most failed OKR rollouts are exactly this: goals adopted, system never built, and by the third quarter nobody scores them.

Worth reading alongside any of the three: Chris McChesney, Sean Covey and Jim Huling, The 4 Disciplines of Execution — the discipline of one wildly important goal, lead measures rather than lag measures, a scoreboard the team keeps itself, and a weekly cadence of accountability. It is not a full operating system either, but it is the sharpest short treatment of why a team moves a number, and it grafts onto all three.

Choosing. Under a hundred people with no system today, pick EOS. Strategy or cash is the constraint and the leadership team can carry the instrumentation, pick Scaling Up. A product-led company with a strong existing goal culture can run OKRs, provided you accept that you are also building the accountability chart, the weekly meeting and the issues discipline yourself. Whichever you choose, commit to four quarters before you judge it. Two quarters is not a trial; it is the installation.

Steps

1. Pick one framework and buy the book for the whole leadership team. Everyone reads it before the launch session — not the summary, not the podcast episode. Give it fourteen days. If the team will not read one book to install the system the company will run on for the next five years, you have found the real problem and it is not the framework.
2. Name the person who owns the system, and it is not you. The founder is the visionary; someone else keeps the rhythm — EOS calls the seat the Integrator, Scaling Up gives it to whoever owns the cadence. That person owns the calendar, the agenda, the scorecard population, the issues list and the quarterly session. If the only person who can run the system is the founder, it will die the first quarter the founder is busy, which will be the next one. Decide at the same time whether to bring in an outside implementer or coach for the first two quarters. Most companies should.
3. Run a two-day launch session off-site with the whole leadership team and phones in a box. Half a day does not produce this. You come out with four artifacts: the one-page plan, the accountability chart, the first draft of the scorecard, and the first quarter's priorities. Book it at least three weeks out so the reading actually happens.
4. Draw the accountability chart before you argue about people. Lay out the seats the business needs — sales, marketing, product, engineering, customer success, finance and operations at a minimum — and write the three to five outcomes each seat owns. Then put exactly one name in each seat, including your own name in three of them if that is the truth today. Two names in a seat is zero owners. Whether each person is the right person for the seat is the next conversation, and it goes much better once the chart exists, because it is a conversation about a seat rather than about a person.
5. Build the scorecard to five to fifteen weekly numbers, each with one owner and one target. Leading over lagging: demos booked beats bookings, tickets reopened beats CSAT, onboarding milestones hit beats go-live date. Pull from the KPI Dashboard Creation play — the scorecard is the weekly subset of that dashboard with a name and a number beside every line. Test it against the clock: if populating it takes one person more than an hour on Monday morning, it is the wrong scorecard and it will quietly stop being populated in six weeks.
6. Set three to five quarterly priorities for the company, and no more than three to five per person. Where the company has a Meaningful Exit Plan, the benchmark gaps on it are the first candidates for the list — that is the mechanism by which the exit stops being a hope and becomes this quarter's work. Each has one owner, a due date inside the quarter, and a written definition of done agreed before the quarter starts — not "improve onboarding" but "the last ten customers went live inside thirty days." Write down what is explicitly being deferred, and say it to the company, because the list of what you are not doing is the half of prioritization that teams never hear.
7. Install the weekly leadership meeting and never move it. Ninety minutes, same day, same time, and a fixed agenda that gives the first thirty minutes to reporting and the last sixty to solving: scorecard read on-track or off-track only, priorities on-track or off-track only, customer and employee headlines, last week's to-dos, then the issues list worked from the top down. Anything off-track becomes an issue rather than a discussion in the reporting section. Close by capturing to-dos, agreeing the one message that cascades to the company this week, and rating the meeting out of ten.
8. Put the rest of the rhythm on the calendar today, for four quarters. A five-to-fifteen-minute standing daily huddle for the teams that need one; a monthly financial review against budget; a one-day quarterly session to score last quarter, reset priorities and refresh the one-year plan; a two-day annual to rebuild the plan and revisit the three-year picture. The quarterly is the load-bearing beam. Systems almost never die at the weekly meeting — they die the first time a quarterly gets postponed and never rescheduled.
9. Cascade one level down within thirty days. Each department runs the same meeting with the same agenda on its own numbers, its own priorities and its own issues list, and issues move up and down between the levels rather than dying at the boundary. A system that stops at the leadership team is a leadership team habit, and the rest of the company will correctly read it as something that happens to them on Mondays.
10. Document the core processes. Name the six to ten that the company actually runs on — lead to close, close to onboarded, onboarded to supported, hire, release, collect cash — give each one an owner, and write each to the twenty percent of steps that drive eighty percent of the result. Then get them followed by all, which is a different and harder job than writing them. See the Quality Management System play, which owns the documentation and the conformance evidence.
11. Run the first quarter in a spreadsheet and a shared document, then buy the platform. The tool is not the system, and a company that buys software first usually ends up with a configured tool and no rhythm. Once the rhythm has survived a full quarter, move it into a platform against the specification below.
12. Score the system itself every quarter, at the quarterly session, before you set the next priorities. Five numbers: percentage of quarterly priorities completed, target eighty; percentage of scorecard weeks populated on time; leadership meeting attendance; average meeting rating; and the count of issues on the list older than ninety days. Two consecutive quarters below eighty percent on priorities means you are setting too many, setting the wrong ones, or letting the quarter be eaten — and the fix is a conversation about capacity, not a speech about accountability.

The Software

No platform will make a system exist, and the system works on paper. The reason to buy one is that discipline decays wherever the artifacts are scattered — the scorecard in a spreadsheet, the priorities in a project tool, the issues in somebody's notebook, the meeting agenda in nobody's hands at all. A platform puts the artifacts in one place and wraps the meeting around them, so the meeting cannot be run without the numbers being present.

Golden Section recommends Ninety.io for most companies. It is purpose-built for this job — the one-page plan, the accountability chart, the weekly scorecard, quarterly rocks, the issues list, to-dos, process documentation and cascaded team meetings, with the weekly meeting running inside the tool against a timed agenda. Running the meeting in the tool rather than beside it is the feature that matters, and it is the one most alternatives get wrong. It is priced per user, which lets you extend it past the leadership team, and it was built around EOS while now supporting other frameworks.

Others worth a look, depending on the framework you chose:

• Bloom Growth (formerly Traction Tools) — the same category and also EOS-native, particularly strong on running the meeting itself, with support for a Scaling Up rhythm.
• Align — built around the Rockefeller Habits and Scaling Up: the One-Page Strategic Plan, daily huddles, cascading KPIs. The first place to look if you chose Scaling Up.
• Rhythm Systems — Scaling Up-adjacent, heavier and more consultative, aimed at larger teams with a dedicated operations resource.
• OKR platforms — Quantive, Perdoo, WorkBoard, Profit.co, Lattice, 15Five — genuinely good at objectives, key results, scoring and alignment, and largely silent on seats, issues and the meeting. The same shape as the framework: plan for a second tool or a serious build to cover the rest.
• General work tools — Asana, Notion, Monday, ClickUp, Linear — you can build a competent operating system in any of them, and plenty of companies have. The cost is that you now own the build forever, and it drifts the month the person who made it gets busy. Ideal for proving the rhythm in quarter one; a poor place to leave it.

Whatever you buy, hold it against this specification. A tool that misses the first six lines is a task manager.

1. Holds the one-page plan and makes it visible to every employee, not just to the leadership team.
2. Holds the accountability chart with seats, owners and the outcomes each seat owns, and links each seat to the numbers and priorities it carries.
3. Weekly scorecard: one row per measurable, one owner, a target, automatic red or green against that target, and at least thirteen weeks visible at once so trend is readable without exporting anything.
4. Quarterly priorities with owner, due date, status and milestones, at company level and cascaded to each person.
5. Runs the weekly meeting: a timed agenda that holds section time, pulls the scorecard and the priorities into the meeting rather than sending you elsewhere for them, captures issues as they surface, and ends by generating to-dos.
6. A ranked issues list that persists between meetings and can move an issue down to a department or up to the leadership team without retyping it.
7. To-dos with an owner and a default seven-day clock, resurfacing next week whether or not anyone remembered.
8. Automatic minutes and a meeting rating, archived and searchable. This is the documentation that raises an exit price — see the KPI & Strategic Meetings play — and it is worth real money later that it accumulates without anybody maintaining it.
9. Cascading teams: departments running the same structure on their own numbers, rolling up to the leadership view.
10. Integrations that populate scorecard lines from the source systems — accounting, CRM, product analytics, support desk. Any number a human retypes every Monday is a number that will eventually stop being retyped.
11. Somewhere to hold the documented core processes, attached to the seat that owns each one.
12. Quarterly and annual session support, with history: last quarter's priorities and their completion rate available without archaeology.
13. Per-user pricing you are willing to extend to every employee. If cascading is expensive, you will not cascade, and a system that never leaves the leadership team is half a system.
14. Export. You should be able to take your plan, chart, scorecard history and priorities with you if you leave.

Troubleshooting

My leadership team are experienced operators and they find this juvenile. They ran bigger companies that already had an operating system, and they were inside it rather than responsible for building one. Experience makes an executive good at running a meeting; it does not produce a scorecard, and it does not tell the twenty-nine people outside the meeting what good looks like this week. Put it to them as a question rather than a mandate: name the three company priorities this quarter, and the owner of each. If four experienced managers give four different answers, the argument is over.

We already have dashboards and accountability. Then answer four questions, each with a name and a date. Which meeting is the scorecard read in? Who owns each number on it? What are the company's three priorities this quarter? What happened after the last number that went red? If any of the four takes more than a sentence, you have instrumentation and not a system. Dashboards are the part of an operating system that is easiest to buy, which is exactly why it is so often the only part present.

We tried EOS and it did not stick. It is almost always one of three things. The founder stopped attending the weekly meeting, which tells everyone the meeting is optional. Or a quarterly session got postponed once and then never rescheduled, and the priorities went stale inside six weeks. Or it was never cascaded, so the company experienced it as something the leadership team did on Mondays. Before you conclude the framework was wrong, check which of the three happened, because the answer is usually in the calendar rather than in the method.

We are eight people. Is this too early? Install three of the eight parts now: a weekly meeting with a fixed agenda, five numbers with owners and targets, and three priorities a quarter. Add the accountability chart the moment a second person manages anybody, and the rest as the leadership team forms. The full system before there is a team to run it is cost without leverage.

This will slow us down. It costs the leadership team ninety minutes a week and one day a quarter. Set that against the last three decisions this company made twice, the last hire who spent a quarter working on the wrong thing, and the last number nobody owned. Then price the version where you hold the all-hands about performance and change nothing structural: the good people update their résumés, because performance pressure without a system reads to a strong employee as a leader who does not know what is wrong.

I am the founder, and I am the one who will not follow it. That is the most useful sentence in this play, and the answer is to hand the system to somebody else and submit to it in public. A founder who cannot be measured has taught the whole company that measurement is for other people, and no framework survives that.