A practical guide to OKRs for mid-sized companies

A mid-sized company sits in an awkward place: too large for coordination to happen by overhearing it, too small to absorb heavy process. That gap is where goals quietly stop connecting to each other, and it is the problem OKRs were designed for.

What they are

An objective is a qualitative statement of what you are trying to achieve. It should be ambitious, directional and memorable enough that people can say it without opening a file. Key results are the measurable outcomes that would tell you the objective happened. If you hit all the key results and the objective still has not been achieved, the key results were wrong.

Four rules make the difference between OKRs and a rebranded to-do list. Everyone can see everyone’s, which is what makes the system work at all. The goals are ambitious enough that hitting 70 to 80 per cent counts as success. They are kept away from compensation, because the moment a goal decides a bonus, people set easier goals. And they run on a quarterly cycle with check-ins in between.

Why the middle size feels it most

It is always the same story. Marketing launches a campaign that sales was not told about. Engineering ships something that does not touch the problem customer success has been escalating for a month. Each team optimises its own metric perfectly and the sum is worse than any of the parts.

None of that is a communication failure exactly. It is the absence of a visible structure that connects one team’s work to another’s.

Setting them up

Start with three to five company objectives per quarter and nothing else. For example, an objective about accelerating revenue growth with key results such as reaching a specific ARR figure and raising the average deal size from one number to another. Or an objective about customers, with a target NPS and a churn rate you are trying to bring down.

Then let departments write goals that explicitly support one of those, and individuals write goals that mix an outcome with something they are trying to learn.

The four mistakes everyone makes first

Too many. Three to five per level, enforced, or the word priority stops meaning anything.

Business as usual dressed as an objective. If it describes what the team does every quarter anyway, it is not an OKR, it is a job description.

Sandbagging, which is the predictable consequence of judging people on their scores. Ambitious attempts have to be safe to make.

And set and forget: goals written in the first week of the quarter and reopened in the last. Put the review inside meetings that already exist, including the 1:1.

On scoring, the standard is a scale from zero to one, where 0.7 to 1.0 is a good outcome for an ambitious goal. Consistently scoring 1.0 means the goals are too easy, which is a nicer problem to have than the opposite and still a problem.

Sign up for a free TalentRewards trial.