EOS vs OKR: Rocks vs Key Results, and Why Most Companies Under 200 People Pick the Wrong One

By Andreas Pettersson, Founder, Leaders ADAPT

You set eleven priorities in January. Every leader has three. By the end of February, the weekly meeting is a status read. By March, two are done, four are "on track" and five have quietly stopped moving.

Then someone says the problem is the goal system. Rocks are too rigid. Try OKRs. Or the reverse.

That is the EOS vs OKR debate in most companies under 200 people. And most of them pick the wrong one. Not the wrong system. The wrong number.

This page is one spoke of our business operating system comparison. I have run EOS and I have run OKRs, and I run EOS at Leaders ADAPT today.

What is the difference between EOS Rocks and OKRs?

A Rock is a 90 day priority with one owner and a binary finish: at the end of the quarter it is done or it is not.

An OKR is an objective paired with three to five measurable key results that show how far the team got.

Rocks come from the Entrepreneurial Operating System. They live on the Vision/Traction Organizer, get reviewed every week in the Level 10 Meeting and sit next to a Scorecard of weekly numbers. EOS Worldwide describes three to seven Rocks per quarter (EOS Worldwide, read September 2026).

OKRs, short for Objectives and Key Results, come from Andy Grove at Intel. John Doerr learned the method there and introduced it to Google's founders in 1999 (What Matters, read September 2026). They are built for ambition and for measurement, and they are usually written as stretch goals.

Source: Leaders ADAPT comparison, Andreas Pettersson, September 2026. The hybrid column is how I run quarterly goals. It is not a vendor method.

EOS vs OKR under 200 people: why most companies pick the wrong one

The failure is not the system. It is the count.

A huge failure is when organisations try to set ten Rocks and fail at all of them. We were the most successful when we had between one and three per quarter. That was true at Arcules and it is true at Leaders ADAPT now.

My rule is stricter than the books. Every person carries at most one supporting Rock. Get as close as you can to one company wide Rock. Everything else is either a subdivision of it or preparation for a future quarter's Rock.

OKRs make the count problem worse in small companies. Three objectives with four key results each is twelve things to track. Cascade that down one level and a 60 person company is scoring a hundred numbers every week. The leadership team becomes a reporting function.

There is a second confusion underneath. The weekly numbers that tell you whether the business is healthy are KPIs, and in EOS they live on the Scorecard. They are not goals. When a company turns every KPI into a key result, it loses the difference between running the business and changing it.

Keep the Scorecard for running. Keep the quarterly targets for changing.

Motion is not progress. Busy does not equal valuable.

So when a company under 200 people asks me Rocks vs OKRs, or EOS vs OKRs, I ask a different question first: how many did you set last quarter, and how many did you finish?

Can you use OKRs with EOS?

Yes. I do. But one system owns the vocabulary.

I am absolutely on board with Rocks. I push what I call 90 day targets, and I take inspiration from OKRs for the measurement thinking. OKRs are still a very good way to think about how you measure.

Here is the problem I have with running Rocks too hard. There is too much rigidity in EOS in the beginning.

You should be able to change things in between when the business situation changes. OKRs are nicer there. That is the one place I lean away from strict Rock discipline.

How to write a 90 day target with OKR measurement

1. Start from the annual target on the scorecard. The quarter's target should drive a substantial share of it.
2. Write the target as a Rock: one sentence, one owner, a date.
3. Add one to three measures, OKR style, that prove it is done well, not just done.
4. Review the measure weekly in the leadership meeting. No "on track" without the number.
5. If the business pivots mid quarter, change the target in the meeting and log why.

An illustrative example, not a client's. Rock: new customer onboarding runs without the founder by June 30. Measures: 20 new accounts onboarded with no founder call, first value reached within 14 days, onboarding satisfaction of 8 or higher.

One Rock. Three numbers. One owner.

What one company wide Rock looks like

Picture a 45 person services firm. The company Rock for the quarter is a new service line live with five paying clients by September 30. That is the one thing.

Sales carries one supporting Rock: ten qualified proposals for the new line. Delivery carries one: the delivery playbook written and tested on the first client. Finance carries one: pricing and margin model signed off before the first proposal goes out.

Nobody else carries a Rock this quarter. Their job is the Scorecard. Four Rocks, all pointing at the same finish line. The example is illustrative, not a client, and it is what OKR vs Rocks looks like when the count is right.

This is also how OKR vs EOS stops being a debate. The Rock carries the commitment. The key results carry the evidence.

When do OKRs work better than EOS?

OKRs work better when there are many teams that need to align without a single leadership table deciding everything. Product and engineering organisations are the classic case. So are larger companies where the weekly leadership meeting cannot see the front line.

They also fit cultures that want stretch. If your people are energised by a goal they might only partly hit, OKRs give that a language. Rocks do not.

EOS works better when the leadership team is one table, the company is under about 250 people and the problem is focus rather than ambition. Most of the companies I advise are there.

A quick test. Ask each leader to name the company's top priority for the quarter without looking it up. If the answers differ, you have a focus problem, and Rocks fix focus. If the answers match but nobody can say how far along it is, you have a measurement problem, and that is where key results earn their place.

Whichever you pick, give it time. An EOS implementer I spoke with put it at a year and a half to two years of 90 day cycles before a team really learns the tools. Switching systems after two quarters resets that clock.

Rocks vs OKRs vs SMART goals

People mix these three up constantly. They are different tools.

You can write a Rock to a SMART standard. You can measure a Rock with key results. Our guides to OKRs for leadership teams, OKR vs KPI and company goals go deeper on each.

Where does AI fit with Rocks and OKRs?

Neither Rocks nor OKRs were designed for AI. EOS was created before AI, and AI is now a massive culture change. EOS needs to evolve so it does not miss that.

The measurement side is where AI helps first. Collect Rock status from each owner before the weekly meeting.

Pull the key result numbers from the systems they live in. Flag any target whose measure has not moved in two weeks. Draft candidate measures for next quarter from the scorecard history.

The temptation is to let AI write the goals themselves. Do not.

A goal is a commitment a person makes in front of their peers. If a model drafted it, nobody made it. Use AI to sharpen the measure after the team has chosen the target, never to choose the target.

What stays with the leadership team:

• Which target makes the cut this quarter
• The call on an issue blocking a Rock
• What happens when a commitment is missed
• The annual plan the targets serve
• The conversation with an owner who is behind
• Any date promised outside the company

Read what not to delegate to AI for the full list and AI for companies running EOS for the workflows.

What a goal system does not fix

A goal system tells you what to finish. It does not tell a leader how to get a team to finish it.

This is where EOS is lacking. It lacks the how when it comes to leadership. Companies running EOS understand how to do it from an ops perspective, but not the details: how to delegate, how to make people accountable, how to run 1:1s, how to give feedback. OKRs have the same gap.

When a Rock misses, the post mortem usually finds a delegation problem, not a goal problem. The owner never had the authority. Or nobody asked in week three why the number had not moved.

That is why I built the 5 Minute Leader to sit on top of EOS. Our guide to leadership team development covers the people problems every goal system surfaces.

For the independent view of EOS as a whole, read our independent EOS review.

How many quarterly priorities did your leadership team set last quarter, and how many got finished? Reply and tell me the two numbers.

Common questions about EOS vs OKR

What is the difference between a Rock and an OKR?

A Rock is a 90 day priority in the Entrepreneurial Operating System with one owner and a clear done or not done finish. An OKR pairs a qualitative objective with three to five measurable key results, and progress is scored on each result. Rocks favour a few committed outcomes, while OKRs favour measurement and ambition.

Can a company use OKRs and EOS at the same time?

Yes, if one system owns the vocabulary. A common pattern keeps Rocks as the leadership team's quarterly priorities inside EOS and borrows OKR style key results to define how each Rock is measured. Running two separate goal lists, one of Rocks and one of OKRs, usually splits attention and doubles the reporting.

How many Rocks should a leadership team set each quarter?

EOS Worldwide describes three to seven Rocks per quarter at the company level. Many operators run fewer. A practical rule is one to three company Rocks, ideally one company wide Rock that the others support, and at most one supporting Rock per person. Setting ten or more usually means none of them are finished.

Are OKRs better than Rocks for a small company?

Usually not as the main system. Companies under about 200 people often lack the layers that make cascaded OKRs useful, and the scoring overhead lands on a small leadership team. Rocks are simpler to run weekly. Small companies still benefit from OKR thinking when they define what a finished Rock must measurably achieve.

Rocks vs OKRs vs SMART goals: which should a leadership team use?

They answer different needs: Rocks set the leadership team's few quarterly commitments, while OKRs add ambitious, measurable outcomes and suit larger or product led organisations. SMART goals are a format for writing any single goal clearly. Many leadership teams set Rocks, write each one to a SMART standard and borrow key results for measurement.

Is EOS vs OKR a choice between two full systems?

No. EOS vs OKR compares a full business operating system with a goal setting method. EOS includes vision, people, data, issues, process and a weekly meeting, and Rocks are one part of it. OKRs cover goals and measurement only, so a company using OKRs still needs its own meeting rhythm, scorecard and accountability structure.

Your next question

If EOS vs OKR was your question, these are the natural next reads:

• The 4 Disciplines of Execution: WIGs, lead measures and the scoreboard
• EOS vs Scaling Up, compared by an operator
• The 5 Minute Leader protocols for delegation and accountability

Or see all six frameworks in the business operating system hub.