An EOS Scorecard is a weekly one-page report that shows every key number your business needs to be healthy, assigned to a single owner, compared against a weekly goal. That is the complete definition. Everything else is commentary.
The Scorecard is one of the six core tools in EOS (the Entrepreneurial Operating System), a business operating framework created by Gino Wickman and detailed in his book Traction. If you are running EOS at your company, the Scorecard is the tool you review at the top of every Level 10 Meeting. If you are not running EOS formally but want a weekly discipline for tracking what matters, the Scorecard concept transfers cleanly.
This post covers what a Scorecard is, how to build one, what goes on it, and how to avoid the mistakes that make most Scorecards stop working inside 90 days.
What is a scorecard
A Scorecard is not a dashboard. That distinction matters more than most operators realize.
A dashboard shows you everything. A Scorecard shows you the 5 to 15 numbers that, if healthy, tell you the business is healthy. The constraint is the point. When every important number lives on one page and every number has an owner and a weekly goal, you can review the full picture in under five minutes. If a number is red, the conversation starts immediately. If every number is green, you move on.
The EOS Scorecard has five columns: Activity (the name of the measurable), Goal (the weekly target), Weeks 1 through 13 (rolling quarter of actuals), and Who (the person accountable for hitting the goal). Each row is one measurable. Each measurable belongs to exactly one person. No shared ownership, because shared ownership means no ownership.
EOS recommends keeping the Scorecard at 5 to 15 rows. Fewer than five and you are probably missing something. More than fifteen and you are reporting instead of managing.
Scorecard template
A working Scorecard template has three decisions baked into it before you fill in a single number.
Decision one: the measurement unit. Each row on the Scorecard must produce a number, not a status. "Proposals sent: 4" is a number. "Sales feeling good" is not. Every measurable you add needs to pass this test. If you cannot express it as a number at the end of each week, it does not belong on the Scorecard.
Decision two: the owner. Each row has exactly one name next to it. That person is responsible for hitting the goal and for reporting the actual. They are not responsible for every activity that feeds the number, but they are responsible for the number. If the number drops, the conversation is with them.
Decision three: the goal. The goal is a weekly target, not a monthly or quarterly aspiration. If you are trying to book twelve qualified meetings per month, the Scorecard goal is three per week. The weekly cadence is what gives the Scorecard its accountability force. You know within seven days whether the number moved.
A minimal Scorecard template looks like this: one row per measurable, five columns (who, measurable name, weekly goal, 13 weeks of actuals rolling left to right, and a current-week entry). Shade a cell green if the actual meets or exceeds the goal. Shade it red if it does not. The visual pattern over 13 weeks tells you whether a number is trending up, down, or bouncing.
At Sneeze It, our Scorecard started on a Google Sheet. We moved it into OTP so that agents can write their own rows without us updating the sheet manually. More on that below.
Weekly scorecard
The weekly cadence is the mechanism that makes a Scorecard work.
Every week, every owner enters their actual for the week before the Level 10 Meeting. The meeting opens with a Scorecard review. The team walks the sheet from top to bottom, one row at a time. Green rows get acknowledged and skipped. Red rows get flagged and moved to the Issues List for discussion later in the meeting.
The review does not take long when the discipline is in place. If every row is green, the Scorecard section of an L10 finishes in two minutes. If three rows are red, those three issues go to the bottom of the agenda and the team spends real time solving them.
The weekly rhythm builds a pattern that quarterly and monthly reviews cannot build. When a number is red for three weeks in a row, the team has been looking at that problem for three weeks, not discovering it at a quarterly review when it has already cost you a month of lost ground.
The other thing the weekly cadence builds is memory. Thirteen weeks of actuals in one view shows you whether a problem is a blip or a trend. A single bad week on a healthy trendline is different from a number that has never been green. The Scorecard reveals that difference without requiring any analysis.
Measurables
Measurables are the specific numbers that go on each row of the EOS Scorecard.
In EOS, a measurable is an activity or output that a person controls, that can be counted each week, and that is predictive of a business outcome. The word "predictive" is important. A good measurable tells you something is going to happen before it happens. A weaker measurable tells you what already happened.
Every seat on your Accountability Chart should have at least one measurable on the Scorecard. EOS's Accountability Chart defines who owns what function in the company. The Scorecard is where you verify that each seat is producing what it is supposed to produce, every week.
At Sneeze It, our seats include humans like Bogdan (our COO) and Janine (accounting) and agents like Radar (chief of staff), Dash (analytics), Tally (scorecard agent), and Dirk (sales). Each of those seats has measurables on the same Scorecard. Bogdan's rows are things like delivery projects on track and open Accelo issues resolved. Dirk's rows are cold emails drafted and pipeline stage transitions. Tally's row is KPI values pushed to the chart without error.
The seats use different metrics because the work is different. The discipline is identical.
Leading vs lagging indicators
This is the most common mistake I see on Scorecards that stop working.
A lagging indicator measures an outcome that has already happened. Revenue collected is lagging. Proposals won is lagging. Clients retained is lagging. These numbers are important. They are not what goes on a Scorecard.
A leading indicator measures an activity that predicts the outcome. Qualified discovery calls booked is leading. Proposals sent is leading. Client check-in calls completed this week is leading. The relationship between leading and lagging is causal: if you hit the leading indicators consistently, the lagging outcomes follow.
EOS recommends building your Scorecard primarily with leading indicators because leading indicators are what the seat owner can actually control week to week. Your salesperson can control how many proposals they send. They cannot control how many prospects say yes in any given week. Putting "revenue closed" on a salesperson's Scorecard row gives them a number that reflects the market as much as it reflects their effort. Putting "proposals sent" on their row gives them a number they own.
The practical test is simple: can the person in that seat change this number through their own actions this week? If yes, it is probably leading. If the number is determined by external decisions outside their control, it is probably lagging.
Lagging indicators belong on your company P&L and your monthly leadership dashboard. Leading indicators belong on your weekly EOS Scorecard.
Frequently asked questions
What is the difference between a Scorecard and an Accountability Chart in EOS? The Accountability Chart (EOS's version of an org chart) defines who owns which function and seat in the company. The Scorecard shows whether each seat is producing what it is supposed to produce, week over week. The Accountability Chart is structure. The Scorecard is performance.
How many measurables should be on an EOS Scorecard? EOS recommends 5 to 15 rows. The number should be large enough to show whether the business is healthy and small enough to review in under five minutes. Most teams starting out should aim for 7 to 10 rows and add only when a gap shows up that no existing row covers.
Can AI agents have rows on an EOS Scorecard? Yes. An agent seat works exactly the same way as a human seat on a Scorecard, provided the measurable is expressed in business-outcome language (not runtime or technical metrics). At Sneeze It, Dash reports leads touched per day, Dirk reports emails sent and meetings booked, and Arin reports call center appointment rate against a 30 percent target. The Monday review covers those rows the same way it covers Bogdan's rows. For more on this, see Humans and agents on the same scorecard.
How do you handle a row that is red for three or more weeks in a row? In EOS, three red weeks on the same row becomes a priority issue. It goes to the top of the Issues List with a structured IDS session: Identify the root cause, Discuss the options, Solve with one clear owner and a due date. A chronic red row usually signals one of three things: the goal is wrong, the inputs feeding the number are broken, or the wrong person is in the seat.
Does the EOS Scorecard work if you are not fully running EOS? Yes. The Scorecard concept works independently. You need three things: a list of 5 to 15 numbers that represent business health, a single owner per number, and a weekly review cadence. The Level 10 Meeting format that EOS recommends is where the Scorecard gets reviewed, but even a simple weekly team call with a shared sheet produces the core benefit. See how to run a Level 10 Meeting for the meeting format that makes the Scorecard actionable.
Run your operating system in OTP
OTP is a scorecard where humans and agents share the same rows, the same weekly goals, and the same accountability conversation. Bogdan, Janine, Radar, Dash, Dirk, Arin, and Tally all live on one chart.
In Claude Desktop or Cursor or any MCP client, add this block:
"otp": {
"command": "npx",
"args": ["-y", "@orgtp/mcp-server"]
}
Restart the client. Then ask: "Use OTP to show me my scorecard and flag any measurables that are below target this week."