What OKRs are

OKRs (Objectives and Key Results) are a goal-setting framework for deciding what a team will achieve over a set period, usually a quarter, and how it will know. An objective is where you're going: qualitative and worth caring about. Key results are how you'll know you got there: two or three specific, checkable outcomes per objective.

When people say “our OKRs”, they mean the whole set: a handful of objectives, each with its key results. The two parts do different jobs.

The objective is one sentence a new starter would understand. “New sign-ups get to a real result in their first session” is an objective. “Q3 onboarding improvements” is a label. Ambitious enough to change what the team works on; plain enough to remember.

The key results answer “how will we know?”. The classic guidance is three to five per objective. For a product team, two or three is better. Every key result needs someone watching it, work underneath it, and a few minutes of attention each week. Five key results across three objectives is fifteen things to track before you've written down any work, and the fourth and fifth usually restate the first two. Fewer, sharper key results are the ones you actually steer by.

The cadence is usually quarterly. Thirteen weeks is long enough to change a number or a behaviour, and short enough that a bad bet gets found out while there's time to correct it. Some companies also set annual OKRs; the quarterly ones should read as steps toward them.

Where OKRs came from

Andy Grove developed OKRs at Intel in the 1970s, building on Peter Drucker's idea of management by objectives. John Doerr learned the method working under Grove, and in 1999 he introduced it to Google, then a young startup. Google's long, public use of OKRs is a big part of why the framework spread so widely. Doerr later wrote the best-known book on the subject, Measure What Matters.

Why OKRs stall: the missing layer

Most PMs have lived this. The OKRs are agreed in week one, they look sharp in the planning doc, and by week four nobody can say how they're tracking. Not because the team isn't working hard. Because objectives and key results describe outcomes, and nobody wrote down the work.

The work lives somewhere else: the sprint board, a roadmap slide, a list in your notebook, your head. So when your manager asks how the quarter is going, you answer from memory.

The fix is a third tier: initiatives. An initiative is a bet you place to move a key result. “Cut the setup wizard from seven steps to three.” “A day-two email based on what the user actually did.” Each one:

  • Sits under the key result it's meant to move. If a key result has no initiatives, it's a wish. If an initiative serves no key result, ask why you're doing it.
  • Has one owner. A person, not a team.
  • Can span quarters. Big bets often do. The initiative carries on while the key results around it change.
  • Carries a short, current note on what just happened and what's next, so status gets written as you go instead of reconstructed the night before a review.

Friyay users often have 30 or more initiatives in flight in a quarter. That's where the quarter is won or lost, and it's the layer plain OKRs leave out. Objectives, key results and initiatives together are the Friyay framework.

Good ideas with great execution are how you make magic.

Larry Page, co-founder of Google

An OKR example, worked through

A product team at a self-serve software company gets plenty of sign-ups, but too many never come back after day one. Here's one objective, three key results, and two initiatives under each. The numbers are illustrative; the shape is the point.

ObjectiveNew sign-ups get to a real result in their first session

KR 1
Sign-ups who complete the core action in their first session: 31% → 45%
  • Replace the empty start state with a guided first task · Designer
  • Move profile and settings questions to after the first task · PM
KR 2
New sign-ups still active in week two: 38% → 46%
  • Day-two email based on what the user actually did on day one · Growth engineer
  • Save-and-resume, so an abandoned setup picks up where it left off · Eng lead
KR 3
New users get set up without contacting support (qualitative)
  • In-product checklist for the three setup steps people ask about most · PM
  • Review every onboarding ticket on Fridays and fix the top cause each fortnight · Support lead

Notice what's there: one objective, three ways of knowing, six pieces of owned work. And what isn't: a date on everything, a Gantt chart, a slide per key result. KR 3 has no number because the team doesn't have a trustworthy one yet.

A shorter example: recurring revenue

ObjectiveGrow recurring revenue

KR
Monthly recurring revenue: $180k → $250k
  • Convert 1,000 mid-tier users to the Pro package
  • Convert 1,000 free users to mid-tier subscribers

This one comes from an earlier version of this page. We've given the key result a starting point, which the original lacked. Without one, nobody can tell whether $250k is a stretch or a rounding error.

Qualitative key results

Textbook OKRs insist every key result is a number. In practice, a lot of what matters in a quarter can't be counted honestly yet. Maybe the event isn't tracked. Maybe it's a capability nobody has measured: can the sales team demo the new pricing without an engineer in the room?

When the number doesn't exist, write what done looks like, in a sentence you can judge honestly at the end of the quarter. “New users get set up without contacting support.” “Finance signs off the new billing model.” Then track it by confidence (on track, at risk, in danger) and by the work underneath. What you shouldn't do is invent a percentage to satisfy the format. A made-up metric gets reported every week and believed by nobody.

Friyay lets both objectives and key results be qualitative, and you can add a number to a key result later, once a real one turns up.

OKRs for individuals, teams and companies

Company OKRs set direction. Team OKRs are where most of the useful work happens: each team picks objectives that move the company ones, and the initiatives underneath are theirs. Once you have several teams or workstreams, group OKRs by team or by area of focus (Payments, Onboarding, Platform) so each group can read its own tree without wading through everyone else's.

Personal OKRs work well for a PM, as long as they're about outcomes you own rather than your job description. One objective for your product area and one for your own craft, say “Run discovery the team trusts”, is plenty. If personal OKRs start to feel like a performance review in disguise, drop them.

OKRs in agile and scrum

OKRs and agile don't compete; they work at different heights. Key results are the quarter's outcomes. Initiatives map roughly to epics: chunks of work big enough to move a key result. Sprints deliver slices of initiatives. A 13-week quarter is about six two-week sprints, so each sprint goal can name the key result it's moving, and the sprint review becomes a natural moment to check it. Keep OKRs at outcome level and the backlog at ticket level, and you won't end up planning the quarter ticket by ticket.

OKRs vs NCT

NCT (Narratives, Commitments, Tasks) swaps the objective for a short written narrative about the problem and why it matters now, with measurable commitments and the tasks that deliver them underneath. It's the same shape as OKRs plus initiatives, with more weight on the story, and Friyay runs either; see running NCTs.

Common OKR mistakes

  • Key results that are tasks. “Launch the new onboarding” is an initiative. The key result is what changes when it lands.
  • Too many objectives. Seven objectives means none of them is a priority. Keep it to one to three per team.
  • Set and forget. OKRs written in week one and opened again in week thirteen are a record of what you hoped for. Look at them weekly, even if it's five minutes.
  • Metrics with no baseline. “Increase activation to 45%” means nothing if nobody knows it's 31% today. Find the starting point first, or write the key result as a statement until you can.
  • No owner. Work owned by “the team” is chased by nobody. Put one name on every initiative.

OKR FAQ

What does OKR stand for?

Objectives and Key Results. An objective is what you want to achieve; key results are how you'll know you've achieved it. Most teams set them quarterly.

How many OKRs should a team have?

One to three objectives, each with two or three key results. Any more and the week's work spreads too thin to move any of them properly.

Are OKRs the same as KPIs?

No. KPIs are the health metrics you watch all the time, like revenue, churn or uptime. OKRs are the few things you're deliberately trying to change this quarter. A KPI that has drifted the wrong way often becomes a key result for a quarter or two until it's fixed.

How often should you review OKRs?

Look at the work under them weekly, check the key results properly at mid-quarter, and score them at the end of the quarter. The weekly look matters most, because it's where you spot a stalled initiative while there's still time to act.

Do OKRs work for one person?

Yes, if you keep them small: one or two objectives, a couple of key results each, and the initiatives you'll actually work on. For a solo founder or a PM setting personal goals, the value is less in the format and more in deciding, in writing, what you're not doing this quarter.

Where Friyay fits

Friyay holds the whole structure in one place: your goals and key results, the initiatives under them, and a weekly list you pick on Monday and close on Friday. Objectives and key results can be qualitative, initiatives can run across quarters, and status gets written as you go, so you can clock off on time. The framework page explains the structure, and quarterly planning and how to set your priority areas for next quarter are the companions to this guide.

Set up your OKRs in Friyay — free to start