Running a Portfolio on EOS: The L10, Rocks & Scorecards That Keep Businesses Aligned

L10 meetings, quarterly rocks, and weekly scorecards are the three things that keep a family-office portfolio aligned without the ownership group becoming the bottleneck. Done right, they make every operator's week faster, not slower. Done wrong, they become the corporate meetings everyone secretly hates. Here's the version that actually works.

The L10 meeting, as it actually runs

L10 stands for Level 10 — the meeting's job is to rate a perfect 10 every week. It runs the same agenda every time, which is the entire point. The leadership team stops spending energy figuring out what the meeting is and starts spending it on the work.

What kills L10s

• The owner monologuing. If the owner talks more than 25% of the meeting, the team will stop bringing real issues.
• No issues list. Without a running list, the same problem gets re-debated every week.
• Solving in the scorecard segment. Off-track numbers move to IDS. Don't solve them in the review.
• Skipping weeks. Three skipped meetings in a row and the muscle is gone. Reschedule, don't cancel.

Quarterly rocks — the part that makes the company actually change

Rocks are the 3–5 things each company commits to finishing in the next 90 days. They're not goals; they're done-or-not-done deliverables with one named owner.

What a good rock looks like

• Specific outcome, not a verb: "Launch new CRM with first 30 accounts loaded" — not "improve CRM."
• One owner. Not a team. One person whose name is on it.
• Doable in 90 days by a person doing their actual job, not a hypothetical one.
• Visible to everyone in the company, not just leadership.

Why we cap at five

Year one in a new portfolio company, leadership wants to fix fifteen things at once. We've watched it happen. Twelve weeks later, two are done, eight are half-done, and five never started. Five rocks, all finished, beat fifteen rocks, half finished, every quarter of every year.

Scorecards — five to fifteen weekly numbers that tell the truth

A scorecard is the leadership team's dashboard for the week. Not the month, not the quarter — the week. Each row is a number with a goal and an owner. If the number is on track, you move on. If it's off, it becomes an IDS item.

The leading-vs-lagging test

Revenue is a lagging indicator — by the time it's down, the cause is two months old. Sales calls booked, demos run, proposals out, pipeline coverage — those are leading. A scorecard built on lagging numbers tells you the past. A scorecard built on leading numbers gives you a chance to change it.

What keeps a family-office portfolio aligned

Across the portfolio, three things stay constant — and that's what makes the family office's job possible.

1. Same meeting structure. Every L10 runs the same agenda, regardless of industry or size. An operating partner can walk into any portfolio company's L10 and know exactly where they are in the meeting.
2. Same rock format. Three to five, one owner each, 90 days, done-or-not-done. Quarterly rock reviews look the same across the portfolio, which means comparison is real.
3. Same scorecard skeleton. The rows differ by company, but the format — weekly, with goal and owner per row — does not. That's what lets a family office review every scorecard in one sitting.

How the family office plugs in

• Weekly L10: not invited. The operating team owns it.
• Monthly: reporting package — P&L, balance sheet, cash, scorecard summary, rock status.
• Quarterly: rock review and planning. Family office in the room, asking questions, pushing on scope.
• Annually: V/TO refresh — values, 3-year picture, 1-year plan, big rocks for the year.

Installing EOS in a new acquisition — the 90-day plan

This is the EOS-specific overlay on our broader first 100 days post-close playbook, focused on the operating-system pieces.

1. Days 1–14: Listen tour with leadership, observe an existing weekly meeting if there is one, draft scorecard v1.
2. Days 15–30: Draft V/TO with operator, install scorecard, hold first L10 (it will be rough — that's expected).
3. Days 31–60: Run weekly L10s, sharpen scorecard rows based on what's actually moving, pick Q1 rocks.
4. Days 61–90: First full rock cycle, first monthly close with the new package, family office quarterly with the operator.

The five pieces that make the cadence work

Strip away the vocabulary and an operating system is five practical commitments. This is the same structure described on our platform operating system page, written here at the level of a single leadership team.

1. A meeting cadence that does not move. Weekly leadership meeting, same day and time; monthly financial review after close; quarterly planning; annual reset. Reschedule, never cancel.
2. One accountable owner per number and per commitment. Not a department, not a committee — a name. If two people own it, nobody does.
3. A defined source for every KPI. Each row on the scorecard names where the number comes from, who pulls it, and when. Two people arriving with two different revenue figures is a source problem, not a performance problem.
4. An exceptions list, worked in order. Only off-track items get discussion time. Everything on track gets a nod and the meeting moves.
5. A decision log. One running list: date, decision, who decided, and the reason. It ends the re-litigation of settled questions and gives a new operator the history in an afternoon.

Where AI fits — drafting, not deciding

Used carefully, AI removes preparation work from this cadence without touching who is accountable. The boundary we hold is simple: AI drafts, a human decides.

• Reasonable AI drafting: assembling the scorecard packet, summarizing last week's notes, drafting an agenda from the open exceptions, turning a discussion recording into candidate action items, drafting the first version of a decision-log entry.
• Human decision, always: what the number means, whether a commitment is met, who owns the fix, pricing, hiring and firing, capital spending, customer commitments, and anything that goes into the financial statements.
• Verification rule: a drafted number is not a reported number until someone ties it to the defined source. The accountable owner signs off, not the tool.

The honest part

EOS is not magic. It's a structure that forces leaders to make choices, finish things, and have the same conversation every week so the company compounds instead of resets. Across a portfolio of businesses, that's how one small ownership group stays out of the way and still stays informed.