An EOS Scorecard is a short list of weekly numbers — usually five to fifteen — each with a named owner and a goal, reviewed in the same five minutes of every leadership meeting. Its purpose is not reporting. It is early warning. A scorecard built correctly tells you in week two that a problem is forming; a scorecard built the way most companies build one tells you in month two that a problem already happened.
The Scorecard is a tool of EOS Worldwide, LLC, and EOS® is their registered trademark. Trinity One is an independent practice, not affiliated with or endorsed by EOS Worldwide. The layout and examples below are our own.
What Is an EOS Scorecard?
It is a single table. Rows are measures. Columns are the last thirteen weeks. Each row has one owner and one weekly goal, and each cell is a number that is either at goal or not. That is the whole artifact — and its discipline comes from what it excludes. No commentary, no monthly roll-ups, no charts.
The Data component of the Entrepreneurial Operating System rests on this one table. Companies with a working scorecard argue about what to do; companies without one argue about what is true.
Leading vs. Lagging Measures
This is the concept most teams get wrong, and getting it wrong is why first-draft scorecards feel useless.
A lagging measure reports an outcome after it is fixed. Revenue, profit, customer churn, project margin. They matter enormously and they are almost useless weekly, because by the time revenue is down the causes are six weeks old.
A leading measure counts an activity or a condition that predicts the outcome. Not "revenue" but "proposals sent". Not "churn" but "accounts with no contact in 30 days". Not "project margin" but "hours logged against estimate at the 50% mark".
The test: if this number is bad this week, can somebody do something about it this week? If not, it is a lagging measure and belongs in a monthly review rather than the weekly scorecard.
A reasonable scorecard runs about two-thirds leading. Keep a small number of lagging measures — cash in bank is worth seeing weekly even though you cannot change it by Friday — but a scorecard that is all lagging is a report, and a leadership team will stop reading it within a quarter.
How Many Numbers Should Be on a Scorecard?
Five to fifteen. Below five you are not covering the business; above fifteen you cannot read it in five minutes, which means it stops being reviewed weekly, which means it stops being a scorecard.
Fifteen rows read aloud as "at goal / at goal / off / at goal" takes about ninety seconds. Fifteen rows where each off-track number gets explained takes twenty-five minutes and destroys the meeting. The number of rows is rarely the real constraint — the discipline of not explaining them is.
Start with one measure per seat on your EOS Accountability Chart, plus two or three company-level numbers. That naturally lands between seven and ten for most small leadership teams.
EOS Scorecard Examples by Function
Concrete measures that behave well weekly, by area:
| Function | Leading measures that work | Common mistake |
|---|---|---|
| Sales | New conversations booked · proposals sent · pipeline value added this week · days since last contact on top 20 accounts | Tracking closed revenue only, which reports a decision made weeks ago |
| Operations | On-time delivery % · jobs at or under estimated hours · open jobs past promised date · rework incidents | Tracking output volume, which rises and falls with demand rather than performance |
| Finance | Cash in bank · AR over 60 days · invoices issued within 2 days of completion · unbilled work in progress | Waiting for the monthly close, by which point everything is history |
| Service / Support | First-response time · tickets open over 7 days · accounts with an unresolved escalation · scheduled reviews held | Average satisfaction score, which moves too slowly to act on weekly |
| People | Open roles past 45 days · one-to-ones held vs. scheduled · new hires past 30 days without a check-in | Headcount, which is a status not a signal |
The Scorecard Layout
Copy this structure into a spreadsheet. Thirteen weekly columns give you a quarter at a glance, so a slow drift is visible without anyone charting anything.
| Measure | Owner | Goal | W1 | W2 | W3 | W4 | …W13 |
|---|---|---|---|---|---|---|---|
| New conversations booked | Dana (Sales) | ≥ 12 | 14 | 11 | 13 | 9 | … |
| Proposals sent | Dana (Sales) | ≥ 5 | 6 | 5 | 4 | 5 | … |
| On-time delivery % | Marcus (Ops) | ≥ 95% | 96% | 97% | 92% | 94% | … |
| Jobs at/under estimate | Marcus (Ops) | ≥ 80% | 84% | 79% | 81% | 77% | … |
| AR over 60 days | Priya (Finance) | ≤ $40k | $38k | $41k | $47k | $52k | … |
| Cash in bank | Priya (Finance) | ≥ $250k | $281k | $274k | $262k | $255k | … |
| Tickets open > 7 days | Sam (Service) | ≤ 3 | 2 | 4 | 3 | 2 | … |
| One-to-ones held | Sam (Service) | 100% | 100% | 86% | 100% | 100% | … |
Read the AR row across. It went off goal in week two and worsened for three consecutive weeks. On a monthly report that surfaces once, in arrears, as a $52k problem. Here it is a conversation in week two, while it is $41k and one phone call.
Who Owns Each Number
One person per measure. Not a department — a person, by name. The owner is not necessarily the person who can single-handedly move the number; they are the person accountable for it being at goal and for raising it when it is not.
Two rules worth enforcing. No number without an owner, ever: an unowned row gets skipped within a month. And no more than about three numbers per person, or the ownership becomes nominal. If a seat needs five measures to describe it, that is usually a sign the seat is carrying two jobs — worth checking against your People Analyzer assessment of whether the right person is in the right seat.
Reviewing the Scorecard in Your L10
Five minutes, at the top of the weekly leadership meeting. The owner of each row says the number and whether it is at goal. Nothing else. Every off-goal number goes straight onto the issues list, and the ones that matter get solved in the sixty-minute IDS segment later in the same meeting.
Explaining a number during the scorecard segment is the single most common way leadership teams destroy their own meeting — the detail is in our Level 10 meeting agenda guide.
Expect to rebuild the scorecard at least once. Almost every first draft is too lagging, too long, or measures what is easy to extract rather than what predicts. Rebuilding it in month two is normal, not a false start, and it is one of the milestones in our EOS implementation guide.
Getting Help
Choosing measures that genuinely predict is the hard part, and it is easier with someone who has seen scorecards in businesses like yours. An EOS Implementer builds this with your leadership team and holds the discipline of the weekly review.
Trinity One does this with founder-led companies, most often in distribution, manufacturing, construction and professional services. Book a discovery call and bring whatever you review weekly today.