What OKRs Are (and Aren't)

An OKR consists of two parts:

Objective: A qualitative statement of what you're trying to achieve. Ambitious but grounded. "Become the go-to tool for freelance designers" is an objective. "Increase revenue" is not — it's too vague to drive decisions.

Key Results: 2–4 measurable outcomes that tell you whether you achieved the objective. Key Results are not tasks. "Launch referral program" is a task. "25% of new signups come from referrals" is a Key Result.

The distinction matters: tasks describe input, Key Results describe output. You can complete tasks and still not achieve outcomes. OKRs force you to define what success looks like, not just what you'll do.

The Mistake That Makes OKRs Bureaucratic

Large companies implement OKRs at the company, division, department, team, and individual level with elaborate alignment processes. The result: days per quarter spent writing and aligning goals, with limited connection to day-to-day work.

Early-stage startups don't need this. With a team of 5–15 people, you need:

• Company-level OKRs: 2–3 per quarter, shared by the whole company
• Nothing else

Individual OKRs, team OKRs, and department OKRs add overhead without adding clarity at small team sizes. The whole company should be able to see the company OKRs on a single screen and understand immediately what the company is optimizing for this quarter.

Setting Good OKRs as a Startup

Three common mistakes:

Sandbagging. Setting Key Results you're 95% sure you'll achieve. Good OKRs should have roughly 70% confidence of achievement — ambitious enough to stretch the team. If you always hit 100%, your targets aren't ambitious enough.

Lagging metrics only. Revenue, signups, and churn are lagging indicators — they tell you about the past. Good OKRs include leading indicators (metrics that predict future outcomes). "User activation rate above 60%" is a leading indicator for future retention. Include both.

Too many OKRs. The power of OKRs is focus. If you have 6 objectives with 4 Key Results each, you have 24 numbers to track. Nobody will track them. Limit to 3 objectives maximum, 3 Key Results each.

A Practical OKR Example

Objective: Achieve initial product-market fit in the freelance design tools category

Key Results:

1. 40% of week-2 users are still active in week-8 (retention signal)
2. Net Promoter Score above 50 from first 100 users (satisfaction signal)
3. At least 3 users willing to be paid reference customers (sales signal)

These Key Results describe outcomes, not activities. They're measurable. They tell you whether the objective was actually achieved. And you can't achieve them by doing theater — you can only achieve them by building a product users actually love.

The Cadence That Works

Quarterly planning (1 day): Set OKRs for the quarter. This includes reviewing the previous quarter (did we hit our KRs? What did we learn?), and setting the next quarter's objectives.

Weekly check-in (15 min): Review progress on Key Results. Are the numbers moving? If not, why? What's the team doing about it? This isn't a formal review — a quick async Slack update works fine.

Mid-quarter adjustment: OKRs can be adjusted if circumstances change dramatically. This should be exceptional, not routine. If you're adjusting OKRs every few weeks, the problem is in your planning process.

When OKRs Don't Fit

OKRs are quarterly. They're not appropriate for goals that need to change faster than quarterly (most pre-PMF startups) or for operational work that's ongoing and doesn't have clear completion criteria.

At the very early stage (pre-product, pre-validation), a simpler instrument works better: a hypothesis board or a learning roadmap. "This quarter, we'll test whether X is true by building Y and measuring Z." This is the spirit of OKRs without the formalism that doesn't fit a team still finding direction.

Move to formal OKRs when you have enough stability to set meaningful 12-week goals. Most teams find this natural around the post-MVP, early growth stage.