Setting Goals That Work: OKRs for Teams Under 20

Why small-team OKRs usually fail

The framework is not the problem. The failures are consistent and all four are avoidable.

Too many. A ten-person company sets five objectives with four key results each, which is twenty things to track. Nobody tracks twenty things, so nothing gets tracked.

Key results that are tasks. "Launch the new pricing page" is work you control completely. Whether it changes anything is the actual question, and a task-shaped key result cannot answer it.

Set and forgotten. Written in week one, opened again in week twelve. In between, the team works on whatever arrives, and the quarterly review is an archaeology exercise.

Tied to pay. As soon as a target affects compensation, people set targets they are confident of hitting. The framework then measures caution rather than progress.

Writing ones that work

1. Three objectives, company-wide, for a quarter

   Not per person, not per team, until you are considerably larger than twenty. Three things the whole company is trying to move.

   Objectives are qualitative and should be readable by anyone: "customers get value in their first week" rather than "improve activation metrics".

   Three, company-wide

2. Two or three key results each, measurable without argument

   A number and a date, where two people looking at the same data would agree on whether it was met.

   If measuring it requires judgement or a meeting, it is not a key result. Rewrite it until the measurement is mechanical.

   No judgement in the measurement

3. Name one owner per key result

   A person, not a team. They do not do all the work, but they own knowing where it stands and raising it when it slips.

   Unowned key results are the ones that quietly become nobody's job by week four.

   One name each

4. Write down what you are not doing

   Three objectives implies things you have decided to defer. Name two or three of them explicitly.

   This is the part that gives the framework teeth. Goals without stated trade-offs are aspirations, and everything continues as before.

   The part everyone skips

5. Check them against last quarter

   Before finalising, look at what you set last time and what actually happened. Repeating an objective is fine; repeating it without acknowledging why it did not move is how it fails again.

   Look back before looking forward

The cadence that keeps them alive

This is where most attempts die, and the fix costs almost nothing.

Weekly, five minutes, in a meeting you already have. Attach it to the review slot in your weekly operating rhythm. Each owner says on track, at risk, or off track, and off track gets a decision rather than a discussion.

Monthly, thirty minutes. Is anything clearly not going to happen? Say so in month two rather than discovering it in month three. Killing a key result mid-quarter is a legitimate outcome and far better than pretending.

Quarterly, an hour. Score, discuss why, and set the next set. Combine this with a retrospective rather than running two separate meetings.

Keep them visible without a dashboard project. One page, wherever the team already looks. The effort spent building elaborate tracking is usually effort not spent on the goals.

Scoring without theatre

Score each key result 0 to 1 based on what actually happened. Average them for the objective.

Around 0.7 is the commonly cited target for genuine stretch goals. The number matters much less than the conversation, and two patterns are worth watching for.

Consistently scoring near 1.0 means the targets are safe. That is not a moral failing, but it does mean the exercise is documenting the plan rather than stretching it.

Consistently scoring below 0.4 usually means the key results were not within the team's control, which is more demoralising than useful. Goals dependent on things you cannot influence produce learned helplessness rather than focus.

Discuss why, not who. A missed key result is information about the plan, the market, or the capacity, and treating it as a personal failure guarantees safer targets next quarter.

When OKRs are the wrong tool

Worth saying plainly, because the framework is often applied where it does not fit.

Under about eight people, everyone already knows what everyone is doing. A ranked list of three priorities on one page usually beats the ceremony.

In a genuinely reactive business, where the work is largely inbound and unpredictable, quarterly goals fight reality. Service levels and capacity measures fit better than objectives.

In the first months of a company, when what you are doing changes monthly, quarterly goals are obsolete before they are scored. Shorter cycles or a simple priority list serve better.

When the real problem is that priorities keep changing from above, OKRs will not fix it. They will just create a document that gets contradicted, which is worse than having none.

If you do use them, the connection that makes them real is between the quarterly objectives and what the team actually does on a Monday. Goals that never appear in weekly planning are decoration, and the fix is mechanical: the Monday priorities should visibly ladder up to one of the three objectives, or you should be able to say why not. Pod keeps those priorities and check-ins in the same place as the work, so the connection is visible rather than asserted in a document nobody opens.