EOS Rocks are the three to seven priorities a person or a company commits to finishing in the next ninety days. The name comes from the jar metaphor — big rocks first, sand fills the gaps — and the whole idea rests on a constraint most leadership teams resist: if everything is a priority, nothing gets finished. This page covers how to shape a Rock so it survives contact with a real quarter, with worked examples of weak versions rewritten into strong ones.
Rocks are a tool of EOS Worldwide, LLC; EOS® is their registered trademark. Trinity One is an independent practice, not affiliated with or endorsed by EOS Worldwide. The examples and failure patterns below come from our own client work.
What Are Rocks in EOS?
A Rock is a specific, measurable outcome with one owner and a ninety-day deadline. It is not a project, a metric, or an area of responsibility. The test is whether two people looking at it on the last day of the quarter would independently agree it is done.
"Improve onboarding" fails that test. "New-hire onboarding checklist live in the HR system and used by all three managers for every hire from 1 October" passes it. The difference is not wordsmithing — the second version forces the decisions that the first one lets you postpone for eleven weeks.
Rocks are the Traction half of the Entrepreneurial Operating System: the mechanism that turns an annual plan into something that actually moves each quarter.
Company Rocks vs. Individual Rocks
Company Rocks are the three to seven things the business must accomplish this quarter. They belong to the whole leadership team and usually come straight out of the one-year plan.
Individual Rocks are what each leadership team member personally commits to. Some derive from a company Rock; others are specific to that person's area.
The relationship between them is where teams go wrong. A company Rock still needs one named owner — "the leadership team" owning something means nobody owns it. And the common failure is arithmetic: seven company Rocks, each cascading into three or four individual Rocks per person, produces a leadership team carrying fifteen priorities each. Every one of them will be off track by week eight.
The workable pattern is three to five company Rocks, each with a single named owner, and each person carrying no more than three to five individual Rocks in total — including any company Rock they own.
How Many Rocks Should You Set?
The stated rule is three to seven. The honest version is that fewer wins, and almost every team we have worked with set too many in their first quarter.
Seven priorities across ninety days, alongside the actual job, means roughly one working day per priority per fortnight. That is enough for a small, well-scoped outcome and nowhere near enough for anything structural. Teams that pick three finish three. Teams that pick seven finish three and feel like they failed at four.
For a first quarter, three is the right number. It feels insufficient in the planning session and correct by week seven.
EOS Rocks Examples
The most useful thing we can hand you is the rewrite. Left column is what teams actually propose in a quarterly session; right column is the version that survives.
| Weak Rock | Strong Rock | What changed |
|---|---|---|
| Improve cash collection | Reduce average days-sales-outstanding from 54 to 45, measured on the 31 Dec ageing report | A number, a baseline, and the report that settles it |
| Hire a sales manager | Sales manager accepted and start date confirmed by 15 Dec | "Hire" is ambiguous — offer out? accepted? started? |
| Roll out the new CRM | All 9 sales reps logging every opportunity in the CRM; zero deals tracked in spreadsheets by 30 Nov | Adoption, not installation. The old version is done when IT says so. |
| Document our core processes | Quote-to-cash process documented in 8 steps, reviewed by the two people who run it, in the shared drive by 15 Dec | Scoped to one process with a named reviewer |
| Improve customer retention | Quarterly business review held with all 12 accounts over $50k, notes filed, by 20 Dec | Retention is an outcome, not a ninety-day action |
| Get the leadership team aligned | One-year plan written, agreed by all 6 leaders, and on one page by 30 Oct | "Aligned" cannot be judged; a signed document can |
| Fix the operations bottleneck | Second shift running in the finishing cell, staffed and producing to standard by 1 Dec | Names the actual intervention rather than the problem |
| Launch the new service line | First paying customer invoiced on the new service line by 31 Dec | Revenue proves launch; a website page does not |
| Better weekly reporting | Scorecard with 12 measures, each with an owner and a goal, reviewed in every weekly meeting from 1 Nov | Specifies the artifact and the point at which it is in use |
| Reduce employee turnover | Stay interviews completed with all 14 people in the two highest-turnover roles by 15 Dec | Turnover moves over years; the interviews are this quarter's work |
The pattern in every rewrite: a date, a number or a named artifact, and a way to settle the argument on the last day of the quarter without anyone's opinion being involved. Getting the underlying measures right is a related discipline — see the EOS Scorecard.
Why Rocks Fail by Week Six
The pattern is consistent enough to predict. Week one to three: energy, some real progress. Week four to five: a customer emergency, a resignation, a system outage. Week six: two Rocks have had no attention in a fortnight, and in the weekly meeting both owners say "on track" because saying otherwise feels like an admission.
That last part is the actual failure. Not the emergency — those are guaranteed. The failure is that "off track" carries social cost, so the team loses six weeks of warning. By week eleven the Rock is visibly not done and there is no time to respond.
Three things fix it. Make "off track" the normal answer rather than the confession — a leadership team where nothing is ever off track is not being honest. Check Rocks weekly in two words, not in a status update. And when something goes off track, put it on the issues list and solve it that week, which is exactly what the sixty-minute IDS segment exists for.
The other structural fix is fewer Rocks. A team carrying three priorities absorbs a bad fortnight. A team carrying seven does not.
Tracking Rocks in the Level 10 Meeting
Rock Review is a five-minute segment in the weekly leadership meeting. Each owner says "on track" or "off track" — two words each, no explanation. Anything off track goes onto the issues list and gets solved during IDS, in the same meeting.
If Rock Review consistently overruns, the Rocks were written too vaguely to be answered in two words, which sends you back to the rewrite table above. The full seven-segment structure is in our Level 10 meeting agenda guide.
At quarter end, count completion honestly. Most teams finish about half in their first quarter and around eighty percent by their third, once they have learned what ninety days actually holds. That learning curve is normal and is covered in the twelve-month view in our EOS implementation guide.
Getting Help
Setting Rocks is easy. Finishing them is a function of whether anyone owns the rhythm between quarterly sessions — which is the gap most companies have. An EOS Implementer facilitates the session where Rocks get set; an Integrator drives them through the eleven weeks afterwards.
Trinity One does both for founder-led companies. Book a discovery call and bring last quarter's priorities — we will tell you why they did or did not land.