The Founder’s Guide to Setting OKRs
That Actually Work
Most startup OKR implementations fail within two quarters — not because OKRs don’t work, but because founders apply an enterprise framework to a 10-person company and create bureaucracy instead of focus. Here’s the lightweight version that actually creates alignment.
- Why most startup OKRs fail
- What OKRs actually are — and aren’t
- How many OKRs you actually need
- Writing objectives that create direction
- Writing key results that measure what matters
- The quarterly OKR rhythm
- When to abandon an OKR mid-quarter
Why Most Startup OKRs Fail
They fail for three reasons. Too many OKRs — a team of 8 with 24 OKRs has no priorities, it has a list. OKRs that are actually tasks rather than outcomes — “launch the new pricing page” is a task, not a key result. And no regular review rhythm — OKRs set in January and reviewed in December are not OKRs, they’re aspirations with better formatting.
The fix is not a better OKR tool. It’s a more honest conversation about what actually matters this quarter, and the discipline to say no to everything that isn’t on that list.
If your team can’t remember their OKRs without looking them up, you have too many. Every person on the team should be able to say their top objective from memory at any point in the quarter.
What OKRs Actually Are — and Aren’t
OKRs are a focus and alignment tool, not a performance management tool. This distinction matters enormously. If people are evaluated and compensated based on hitting their OKRs, they’ll set easy OKRs. If OKRs are ambitious targets that create directional alignment, people will set meaningful ones.
Objectives are qualitative, directional, and inspiring. They answer “where are we going?” not “what are we doing?” A good objective makes people want to work toward it.
Key Results are quantitative, specific, and time-bound. They answer “how will we know we’ve arrived?” Not tasks completed, but outcomes achieved. “Reach $50k MRR” is a key result. “Write 10 blog posts” is not.
How Many OKRs You Actually Need
For a company under 20 people: 1-3 company-level OKRs per quarter. Each with 2-3 key results. That’s it. No team OKRs, no individual OKRs — at this stage, company OKRs are everyone’s OKRs.
For a company 20-50 people: 2-3 company OKRs, and 1-2 OKRs per team that directly contribute to a company OKR. No team should have an OKR that doesn’t connect clearly to a company OKR.
The trap: adding OKRs to cover every important area. If everything is equally important, nothing is. Pick the 2-3 things that, if achieved, would make this quarter genuinely great. Everything else is either supporting work or it waits.
“Help me set OKRs for my startup for the next quarter. Context: [describe your stage, current metrics, what you shipped last quarter, what the biggest challenges are]. The 2-3 most important outcomes this quarter are: [describe what genuinely needs to happen]. For each one: write an inspiring qualitative objective (1 sentence), then write 2-3 specific, measurable key results that would tell us we’ve achieved it. Make sure the key results are outcomes not tasks — measured in numbers, not checkboxes.”
Writing Objectives That Create Direction
A good objective is ambitious, qualitative, and answers the question “what does winning look like this quarter?” It doesn’t need to be measurable — the key results handle measurement. It needs to be directional and motivating.
Test for a good objective: can someone who joins the company on day one understand what it means and why it matters? If it requires internal context to interpret, it’s not clear enough. If it doesn’t feel like a stretch, it’s not ambitious enough.
Examples of weak objectives: “Improve our product” (too vague), “Keep the team happy” (unmeasurable in a useful way), “Hit our revenue targets” (this is a key result, not an objective).
Examples of strong objectives: “Become the go-to tool for Series A hiring teams,” “Make our onboarding so good that customers are successful before we talk to them,” “Build the sales foundation that scales to $1M ARR.”
Writing Key Results That Measure What Matters
The most common key result mistake: writing tasks dressed up as outcomes. “Launch the referral programme” is a task. “Generate 20 referral signups from existing customers” is a key result. The difference is that the task can be completed without producing the intended outcome, but the key result can’t be faked.
Each key result should answer: how will we know this objective is achieved? It should be specific enough that at the end of the quarter, everyone agrees whether it was hit or not. Ambiguous key results produce ambiguous accountability.
A good set of 3 key results covers leading indicators (what do we need to do to make the outcome possible?) and lagging indicators (did the outcome happen?). Relying only on lagging indicators tells you what happened but not how to course-correct mid-quarter.
“Here are our OKRs for this quarter: [paste your objectives and key results]. Stress-test them: (1) Which key results are actually tasks rather than outcomes — and how should I rewrite them? (2) Are there important outcomes we’re not measuring that could let us hit these key results while missing the actual objective? (3) Are any key results too easy — achievable even if the quarter goes badly? (4) What’s missing from this set that we’d regret not measuring?”
The Quarterly OKR Rhythm
Week 1: Set OKRs. Share them with the whole team. Make sure everyone understands how their work connects to at least one OKR.
Every 2 weeks: A 30-minute OKR check-in. Not a status update — a signal check. For each key result: on track, at risk, or off track? If at risk or off track, what’s the one thing we’re changing?
End of quarter: Score each key result 0-1. 0.7 is considered a strong result for ambitious OKRs — if you’re consistently hitting 1.0, your OKRs aren’t ambitious enough. Write a one-paragraph retrospective per objective: what happened, what we learned, what changes next quarter.
When to Abandon an OKR Mid-Quarter
Sometimes a quarter changes — a new opportunity appears, a market shifts, an assumption turns out to be wrong. When that happens, clinging to OKRs set on old information is worse than acknowledging reality and adjusting.
The test for whether to abandon an OKR mid-quarter: is this OKR still the most important thing for the company, given what we now know? If yes, stay the course. If no, have an explicit conversation about what’s replacing it and why — don’t just let it quietly die. The conversation matters as much as the decision.
Get 50 more prompts for ops, planning, and running a lean startup — free.