OKRs: ambitious objectives and measurable key results, one quarter at a time

An objective may stretch you; a key result must be countable.

In this topic: Execution and control 7

An objective may stretch you. A key result must be countable.

OKRs pair one ambitious objective with a few measurable key results per quarter. How they work, how to score them and the pitfalls that Grove and Doerr already warned about.

What are OKRs?

A team's first OKR session ends with one objective and eleven key results: "launch the new website", "hold a weekly stand-up", "improve onboarding". Three months later every box is ticked and nothing has changed for a single customer. That is a task list with a new name.

OKRs (objectives and key results) are a goal-setting method for a short period, usually a quarter. The objective says in words what you want to achieve, and it may be uncomfortable. The key results make it measurable: two to five numbers that show, without discussion, whether you got there. They describe outcomes, not activities. OKRs are visible to everyone in the organisation, so teams can see what others are working on and where they depend on each other.

Where OKRs come from: Intel, Doerr and Google

Andy Grove developed the system at Intel as a form of management by objectives and described it in High Output Management (Random House, 1983). His logic was simple: the objective tells you where to go, the key results show how close you are getting, and the whole system only brings focus if you keep the number of objectives small.

John Doerr learned the method at Intel in 1975 and introduced it to Google's founders in 1999. In Measure What Matters (Portfolio, 2018) he describes what he calls the superpowers of OKRs: focus on a few priorities, alignment between teams, tracking of progress and stretching beyond what seems possible. The book made OKRs popular far beyond Silicon Valley, and also made them easy to copy badly.

How an OKR works: objectives, key results and scoring

  • Objective: one short, qualitative sentence that is ambitious and fits on one line. "Make our support the reason customers stay" works; "Support KPIs" does not.
  • Key results: each one a number with a starting value, a target and the end of the quarter as deadline. If you cannot score it, rewrite it.
  • How many: Google's guide advises three to five objectives with around three key results each, because more leads to over-extended teams (source: Google re:Work, 2016).
  • Scoring: key results are graded from 0.0 to 1.0, and the sweet spot for ambitious OKRs lies around 60 to 70% (source: Google re:Work, 2016). A team that always scores full marks is setting goals that are too safe.

Score per key result = (result achieved - starting value) / (target - starting value), kept between zero and one. The objective's score is the average of its key results.

Aspirational or committed: choose per objective

Aspirational objective

Deliberately set beyond what seems feasible. Partial achievement counts as success. Suits growth, new markets and process improvement, where a team that reaches two-thirds of a bold target has still moved further than with a safe one.

Committed objective

Must be achieved in full, and resources are adjusted until it is. Suits deadlines, legal obligations, audits and customer deliveries. Falling short here is a problem, not a learning experience.

Google's own OKR playbook draws the same line: committed OKRs are meant to be delivered in full, while aspirational ones are expected to fall short on average. Keep committed objectives few. Most of them describe the normal running of the business and belong on the KPI dashboard.

The OKR cycle in one quarter

  1. 01

    Before the quarter: choose

    Two or three weeks ahead, leadership sets a few company OKRs that follow from the strategy or the OGSM. Teams then propose their own.

  2. 02

    First week: align

    Teams compare drafts, find where they depend on each other and agree who needs what from whom. This is where most OKR programmes quietly fail.

  3. 03

    Every week: a short check-in

    A quarter of an hour per team: status of each key result, how confident you are and what stands in the way.

  4. 04

    Mid-quarter: adjust

    If an assumption no longer holds, change or drop the key result. Holding on to a goal that has lost its meaning only produces a poor score.

  5. 05

    End of quarter: score and learn

    Score each key result, discuss what worked and what did not, and carry the lessons into the next set of OKRs.

Weak and strong key results: examples

WeakProblemStronger
Launch the client portalA task, not an outcomeShare of samples registered through the portal rises from 20% to 70%
Improve turnaroundNo measure and no end valueAverage turnaround for standard tests falls from 4.0 to 2.5 working days
Call 50 prospectsMeasures effort, not effectEight prospects start a three-month trial package
Revenue 3% above last yearBusiness as usual; belongs on the dashboardRevenue from the new dairy package grows from zero to €120,000 this quarter
Faster, cheaper and better reportingThree results in oneThree separate key results, each with its own number
More followers on social mediaA vanity measure with no link to the objective30 qualified enquiries from food producers, of which 10 receive a proposal

Fictitious Firmfocus examples.

A simple test: if a team can tick off a key result without anything changing for a customer or for the business, it is a task. Tasks belong in the team's planning, not in its OKRs.

OKR example: a laboratory's quarter with dairy producers

Kestrel Analytical, a testing laboratory with 85 staff, is moving from one-off tests to monitoring contracts with food producers. Contract sales started slowly, because buyers wanted proof of fast results before signing. For the third quarter, the commercial and microbiology teams share one aspirational objective: make Kestrel the obvious choice for dairy producers who want results they can plan around.

Key resultStartTargetAchievedScore
Dairy producers on a trial package or contract21280.6
Dairy samples registered through the portal20%70%60%0.8
Average turnaround for dairy microbiology, working days4.02.53.10.6
Objective score (average)0.67

Fictitious Firmfocus example. Scores: (8 - 2) / (12 - 2) = 0.6; (60 - 20) / (70 - 20) = 0.8; (4.0 - 3.1) / (4.0 - 2.5) = 0.6; average (0.6 + 0.8 + 0.6) / 3 = 0.67.

Alongside it ran one committed objective: pass the accreditation surveillance audit without major findings. It was met and is scored separately, as done or not done.

The score of 0.67 sits in the 60 to 70% range that Google calls the sweet spot for ambitious goals (source: Google re:Work, 2016). The review showed why the first key result lagged: producers on a trial waited for a full month of results before signing. Next quarter's key result therefore measures conversion from trial to contract. Turnaround has stabilised and moves to the dashboard as a regular KPI with a norm.

Known OKR pitfalls, and what Grove and Doerr warned about

  • Too many objectives: Grove's whole point was focus. Six objectives per team means nothing was chosen. A small firm does well with one to three company objectives.
  • Tasks as key results: key results describe outcomes; the work to get there belongs in the planning.
  • Linking OKRs to bonuses: Google states that OKRs are not synonymous with performance evaluation, and Doerr argues for keeping them apart from pay decisions. Research by Lisa Ordóñez and colleagues shows that hard goals tied to rewards can distort risk-taking and encourage unethical behaviour (Academy of Management Perspectives, 2009). Paid on the score, people set safe goals.
  • Everything aspirational: a stretch goal for customer deliveries is a risk, not an ambition.
  • OKRs without a strategy: OKRs make a choice executable but do not make the choice. Check first what good strategy is.
  • Imposed from the top: let teams propose most of their own OKRs, linked to the company goals, or commitment stays thin.
  • Set and forget: without the weekly check-in, the quarter ends with a surprise instead of a score.

OKRs or OGSM: which suits you?

OKRs work better when

  • your market or product changes quickly;
  • teams can act fairly independently;
  • partial results are acceptable;
  • people are used to sharing goals openly.

OGSM works better when

  • you need a direction for several years;
  • coherence between departments matters most;
  • owners or lenders must be able to rely on the goals;
  • a quarterly rhythm is new to the organisation.

The two combine well: OGSM for the course, OKRs for quarterly pushes, and the Balanced Scorecard or a dashboard for the measures that must stay on target all year.

OKRs in a small business and a larger organisation

Small business

Keep it small: one to three company objectives, reviewed in the weekly meeting you already have, on a whiteboard or a shared document. Be careful with ambitions that cost money: an ambitious sales target is fine, an ambitious cash assumption is not. Check any investment behind an OKR in the Business Planner.

Larger organisation

The value lies in visibility and alignment across teams. The risk is a cascade of hundreds of goals that mainly produces paperwork. Keep company OKRs few, let teams derive their own and discuss dependencies every quarter. Turn OKRs into an appraisal system and the ambition disappears.

OKRs are one of the goal-setting tools in execution and control. For the multi-year course, see the OGSM model; for organising the big initiatives, see strategy implementation. All strategy topics are in the strategy overview.

What is the difference between an OKR and a KPI?
A KPI tracks the ongoing health of the business, such as margin or delivery reliability. An OKR describes what you want to change in a period. A KPI that needs fixing can become a key result for a quarter and then return to the dashboard.
How many OKRs should a team have?
Google's guide advises three to five objectives with around three key results each. For a small business, one to three objectives for the whole company is usually wiser.
Should OKRs be linked to bonuses?
Better not. When pay depends on the score, people set goals they know they can reach, and the ambition that makes OKRs useful disappears. Discuss performance separately.
What is a good OKR score?
For aspirational OKRs, Google puts the sweet spot at around 60 to 70% (source: Google re:Work, 2016). For committed OKRs, only full achievement counts.
Firmfocus Advisor

Test whether your draft key results are outcomes or tasks


Sources

Firmfocus jaarlidmaatschap voordelen

Maak gebruik van onze ondernemerstools en de laatste branche-informatie

Verkrijg direct toegang tot alle stappen, downloads en tools voor ondernemers.

Verkrijg direct toegang