// okr glossary · 101 terms

OKR terms, in plain English.

Every word that gets thrown around OKRs, defined honestly in one line, then linked to a full reference page. Where an idea belongs to someone else, we say whose it is. Where we coined it, we say that too.

      Written by Matt Roberts, ZOKRI co-founder
      Last reviewed August 2026
      101 terms · 20 of them ours
    

Find

ZOKRI’s own vocabulary The core framework Metrics and measurement Why OKRs fail: the named anti-patterns Cadence, review and decision Teams, roles and alignment The other goal frameworks, and who wrote them The strategy the OKRs execute

ZOKRI’s own vocabulary

Terms we coined or sharpened. Each has one canonical definition, used verbatim everywhere we write. If you quote them, quote them from here.

Value SequencingZOKRI term#: Value Sequencing is quarterly planning with the ability to accelerate or slow down: goals are ordered by value and pulled at the speed they close, finishing early when the outcome is reached, stopping when stopping is the right call and the reasons are kept, or running longer when the outcome genuinely needs it, while the quarterly planning and reporting rhythm stays fixed. It answers which goal comes next, and why that one. Read more →
Continuous Goal SettingZOKRI term#: Continuous Goal Setting decouples setting from reviewing: reviewing keeps the fixed quarterly clock, setting has none, so a goal is pulled the moment capacity opens and the next review reports on it. It answers when a goal may start. The quarter stays. The waiting goes. Read more →
Finish, then flowZOKRI term#: When a goal reaches its outcome, the team banks the learning and pulls the next most valuable goal immediately, with no waiting for a ceremony. No idle time between goals. Read more →
Goals at the speed of the workZOKRI term#: Goal duration is set by the outcome, not by the reporting period: a goal can close in five weeks or run for five months while the quarterly heartbeat stays fixed. Read more →
Grade, don’t scoreZOKRI term#: Grade a quarter on the evidence, from Excellent to Bad, rather than computing a number from 0.0 to 1.0 and calling 0.7 a success. Grades decide; percentages inform. Fifty per cent progress can be excellent, and a clean 100% is often a sandbag exposed. Read more →
Aligned, not cascadedZOKRI term#: Teams write their own goals against a shared strategy; goals are not decomposed and handed down as arithmetic slices. The proof is older than the argument: catchball in Hoshin Kanri ran it at industrial scale decades before OKRs arrived. Read more →
The SHOP ModelZOKRI term#: The SHOP Model is the four lanes of an OKR portfolio, Strategic, Health, Operational Excellence and Personal, each with its own logic, where only Strategic OKRs ever touch a bonus. It is the answer to “what types of OKRs are there”, and to most compensation arguments. Read more →
Initiatives, Process Commitments and ExperimentsZOKRI term#: The three kinds of work under a Key Result: Initiatives are scoped deliverables, Process Commitments are recurring behaviours, and Experiments are tests of a hypothesis. Most teams call all three “activities”, then wonder why the plan cannot be read. Read more →
Brain vs RuntimeZOKRI term#: Your brain is your linked concepts, evidence and principles, owned forever in portable formats; the runtime is whichever AI carries it this year, rented and replaceable. Confusing the two is how a company ends up renting its own thinking. Read more →
Statements vs SystemsZOKRI term#: A generic-sounding strategy backed by compounding machinery is real; a distinctive-sounding claim backed by nothing is an impostor. Judge systems, not sentences. Read more →
Time to truthZOKRI term#: Time to truth is the lag between a problem existing and leadership knowing about it, measured retrospectively against documented past incidents. Operational rather than attitudinal, and a fair test of whether your rhythm is real or ceremonial. Read more →
Issues and obstacles logZOKRI term#: The standing, owned, dated list of what stands between a team and its Objective: seeded by pre-mortem at creation, kept alive by check-ins, drained by rule-based escalation. Read more →
The initiative graveyard ratioZOKRI term#: The ratio of initiatives quietly abandoned to initiatives formally closed with a reason, counted across the last twenty-four months. The best single measure of an organisation’s honesty about endings, and arithmetic from your own artefacts, so it is not arguable. Read more →
No permanent residentsZOKRI term#: A metric earns an OKR only when it is broken beyond routine fixing or is being pushed to a new level; when the work lands it returns to the scorecard with a new threshold. The promotion and demotion lifecycle between scorecard and goals. No metric lives in the goal set forever. Read more →
Discovery is never an OKRZOKRI term#: Product discovery runs as its own continuous track with its own confidence scoring, and is never written as an OKR, because grading learning stops the learning. Read more →
A reasoning system, not a scoreboardZOKRI term#: OKRs exist to decide what matters and to test whether the bet is working, not to tally output. If your OKR review is a reading of numbers rather than an argument about beliefs, you have a scoreboard. Read more →
Impossible to catch up with, not impossible to copyZOKRI term#: Advantage is not being uncopyable; it is compounding faster than a copier can close the gap. Read more →
The Advantage PlanZOKRI term#: The Advantage Plan is the plan for how your company wins, built with you and installed as the operating system that runs it: goals, metrics, rhythm and the AI thread. The plan never travels alone. A plan without the machine that runs it is a document. Read more →
Dig the new bed firstZOKRI term#: Before AI reclaims an hour, name the judgment work that hour will do, and assign it publicly the same week. Redeployment before reclamation. Read more →
The human layerZOKRI term#: AI augments everything that is written down; humans own everything that is at stake. The line that decides what you automate and what you never hand over. Read more →

The core framework

The words every OKR conversation runs on. If a definition here differs from the one you were taught, the difference is deliberate and explained on the linked page.

OKR#: Objectives and Key Results is a goal method for choosing what matters and knowing whether you are making progress: the Objective is the goal, and the Key Results prove it. A reasoning system, not a scoreboard. Read more →
Objective#: An Objective is a clear, qualitative statement of what you want to achieve this cycle, memorable and worth chasing, and specific enough to guide without deciding the solution. Read more →
Key Result#: A Key Result is a measurable outcome that proves an Objective has been met, written from X to Y. Describe results, not tasks. Usually two to four per Objective. Read more →
Key Result types#: The three shapes a Key Result takes: metric (from X to Y), milestone (a dated binary event) and baseline (measure now, target next cycle). Milestones are legitimate occasionally and a warning sign in bulk. Read more →
OKRs as bets#: Reading each OKR as a deliberate commitment of scarce resources to a goal you believe will matter, made knowing you might be wrong, so it can be tested and learned from. Read more →
Committed vs aspirational OKR#: Committed OKRs are expected to be fully achieved; aspirational OKRs are deliberately ambitious, where strong partial progress counts as success, provided the stretch is labelled. Unlabelled stretch is how an honest miss becomes a broken promise. Read more →
Moonshot and roofshot#: A moonshot is a goal set far beyond current capability, so that even a serious miss moves the business; a roofshot is hard but achievable. Moonshots need cover from consequences to work at all. Read more →
OKR cycle#: The period over which a set of OKRs is planned, run and reviewed, conventionally a quarter. We keep the review clock quarterly and let each goal take the time its outcome takes. Read more →
OKR owner#: The single named person accountable for an Objective or Key Result, who reports the truth on it every week. Ownership is one name. A committee is not an owner. Read more →
Company, team and individual OKRs#: The three levels OKRs are commonly written at. We use company and team goals, and give personal goals their own lane rather than making them a third tier of the same document. Read more →
Nested OKRs#: OKRs written at several levels, so that a team’s Key Result becomes a parent Objective one level down. Standard in most OKR literature, and the fastest route to arithmetic goals and sign-off queues. Align instead. Read more →
Transparency#: Every OKR, its progress and its confidence, visible to everyone in the company by default. Transparency without psychological safety produces careful goals, not honest ones. Read more →

Metrics and measurement

KPI#: A Key Performance Indicator is a metric you monitor continuously to keep the business healthy. KPIs run always-on; OKRs are the focused changes you choose to make. Read more →
Metric#: A named, calculated value combining measurements, often a ratio or a rate, such as activation rate or monthly recurring revenue. Read more →
Measurement#: A timestamped record of a single event, the raw data that metrics are calculated from. Read more →
KPI vs OKR#: KPIs measure ongoing health; OKRs target a specific change. A metric that breaks, or that you want to push to a new level, graduates into an OKR and later returns to the scorecard. Read more →
Metric tree#: A map of measures arranged by cause and effect, from a north-star outcome down to the inputs a team can move, making leverage and assumptions visible. Read more →
Input vs output metric#: An output measures the result you want and is often slow; an input measures something you can move now that you believe drives it. Good Key Results usually target inputs and report outputs. Read more →
Leading and lagging indicators#: A leading indicator moves before the outcome and gives early feedback; a lagging indicator confirms the result after the fact. Choose numbers by the story they test. Read more →
Proxy metric#: A measure that tracks the outcome closely enough to act on when the real outcome is too slow to steer by. A leading indicator claims “this causes that” and owes you a tree; a proxy claims “this tracks that” and owes you a correlation. Read more →
Baseline Key Result#: A Key Result used when you cannot yet set a credible target: you measure to establish the number this cycle, then set a real from-X-to-Y target next cycle. Read more →
North star metric#: The single measure a company treats as the best available proxy for customer value received, used to orient everything below it. One number cannot carry a strategy. It can order a metric tree. Read more →
Vanity metric#: A number that reliably goes up, feels good and changes no decision. Read more →
Aspirational (stretch) target#: A deliberately ambitious target, labelled as such, so an honest miss is not mistaken for a broken commitment. Attach cash to a stretch target and the floor becomes the target. Read more →
KPI scorecard#: The home of business-as-usual: the critical few metrics with owners and thresholds, watched continuously so OKRs can stay focused on change. Read more →
Target setting#: Deciding the number a Key Result moves to, from a stated baseline, with the reasoning recorded. A target without a baseline is a wish with a decimal point. Read more →

Why OKRs fail: the named anti-patterns

Named failure modes leaders recognise on sight. Naming one in a room is usually faster than diagnosing it.

Activity trapZOKRI term#: Mistaking motion for progress: a team busy shipping work while the outcome that matters does not move. Read more →
Task-based Key ResultZOKRI term#: An activity dressed as an outcome, such as “launch the feature”, which can be completed in full while the result you wanted never moves. The most common OKR mistake, and the easiest to spot: if finishing it teaches you nothing about the customer, it is a task. Read more →
Measurement mirageZOKRI term#: Choosing easy metrics over meaningful ones, confusing leading and lagging indicators, and neglecting baselines. Read more →
Solution biasZOKRI term#: Falling in love with a specific solution: skipping problem validation and ignoring the alternatives. Read more →
Sandbagging#: Quietly setting targets you already expect to beat, usually because goals are wired to consequences. The cure is consequence design and labelled stretch, not exhortation. Read more →
The 0.7 convention#: The widespread practice of scoring OKRs from 0.0 to 1.0 and treating 0.7 as success. It teaches teams to calibrate targets to be hit at 70%, which is sandbagging with a formula. Grade, don’t score. Read more →
Watermelon reporting#: Status that is green on the outside and red on the inside, where the report says fine and the work does not. Measure your time to truth instead of arguing about honesty. Read more →
OKR theatre#: Running the full ceremony, planning, check-ins and reviews, while no decision anywhere in the company changes as a result. The test is not whether the meetings happen. It is whether anything was stopped. Read more →
Strategy impostor#: A statement that looks and sounds like strategy while containing no choice a sane rival would reject. Read more →

Cadence, review and decision

Operating cadence#: The heartbeat that keeps goals moving: weekly check-ins and signals, plus quarterly reviewing and reporting. The heartbeat is fixed; goal duration is not. Read more →
Check-in#: A short, structured update where a team reports progress and confidence against its Key Results, centred on the plan and the beliefs, not just the percentage. Read more →
Confidence score#: A quick judgement of how likely a Key Result is to land, used to prompt conversation and surface risk early, never to grade people. Read more →
Retrospective#: The end-of-cycle review that grades the quarter on evidence and asks what the team now believes, turning a cycle into learning. Read more →
CFR#: Conversations, Feedback and Recognition: the continuous performance-management practice intended to run alongside OKRs. Read more →
Pre-mortem#: A structured session, run at goal creation, that imagines the goal has failed and asks why, then seeds the issues and obstacles log with owners and dates. Read more →
Changing or killing an OKR mid-quarter#: The rule for a goal that no longer makes sense: say so out loud, record the reason, and stop it deliberately rather than letting it decay. Stopping well is a skill, and a quietly abandoned goal teaches the whole company that endings do not matter. Read more →
Dual-track discovery#: Running a continuous discovery track alongside delivery, so learning work happens outside OKRs rather than being graded as a goal. Read more →

Teams, roles and alignment

Strategy alignment#: Everyone being able to read, understand and recall the strategy, which is what makes autonomous, aligned decisions possible. Read more →
Cascade: choices vs goalsZOKRI term#: Strategic choices cascade logically; goals must not cascade bureaucratically. The distinction that resolves the cascading argument. Read more →
Catchball#: The Hoshin Kanri practice of passing a draft goal up and down the organisation until the people who must deliver it have shaped it. Alignment without cascading, invented on factory floors in the 1960s. Read more →
Empowered OKR team#: A cross-functional, single-threaded team resourced to match the importance of its goal, with autonomy, de-risked collaboration and invested leadership. Read more →
Single-threaded team#: A team with one clear owner and genuine focus on a single important goal, rather than many people borrowing time from other jobs. Read more →
Push and pull managementZOKRI term#: Managers proactively pushing their skills into goal work, and being pulled into business-as-usual only where they are needed. The reverse of the usual default, where the goal gets whatever attention is left. Read more →
Psychological safety#: The shared belief that you can admit a miss or challenge a view without punishment. The foundation that keeps confidence scores and stretch targets honest. Read more →
OKR champion#: A trained internal expert who guides their team on goals and keeps quality high after external help has left. Read more →
Executive sponsor#: The senior leader who backs a goal with attention and air cover. Whether they show up is the real signal of whether the goal matters. Read more →
Business-as-usual#: The ongoing work that keeps the business healthy, managed with a scorecard rather than with OKRs. Confusing the two causes most failed rollouts. Read more →
Wildly important focus#: Limiting goals to the few things that matter most, because attention is finite. One or two goals per team, worked in serial, beat a long list worked in parallel. Read more →
Creating capacity#: Deliberately delegating, deferring or stopping existing work before a team is asked to take on a goal. A goal added to a full plate is a goal that will lose to the plate. Read more →
OKRs and compensation#: The design question of whether goal attainment touches pay, and where. Cash on team goals does not add motivation, it converts candour into arithmetic. Only Strategic OKRs ever touch a bonus, and only through a gate. Read more →

The other goal frameworks, and who wrote them

We are not framework partisans. This shelf is where we go shopping, with receipts, and every idea we took is credited.

Who invented OKRs#: Andy Grove developed the practice at Intel in the 1970s, calling them iMBOs, and John Doerr took it to Google in 1999, where it became OKRs and the name stuck. Grove’s own account is in High Output Management (1983); Operation CRUSH is the engagement that made the method famous inside Intel. Read more →
Management by Objectives (MBO)#: MBO is the practice of agreeing objectives between manager and employee and reviewing performance against them. The direct ancestor of OKRs, and the source of most of the habits OKRs were meant to fix. Read more →
SMART goals#: SMART is a checklist for writing a goal statement: specific, measurable, achievable, relevant and time-bound. It tests the sentence, not the choice. A perfectly SMART goal can be the wrong goal, and “achievable” quietly argues against ambition. Read more →
BHAG#: A Big Hairy Audacious Goal is a ten-to-twenty-five-year ambition intended to focus and galvanise a company. Read more →
V2MOM#: V2MOM is a five-part annual plan: Vision, Values, Methods, Obstacles and Measures, written top-down and published company-wide. Its best idea is that obstacles are a mandatory section of the plan. Read more →
4DX and WIGs#: The 4 Disciplines of Execution: choose a Wildly Important Goal, act on lead measures, keep a visible scoreboard, and hold a cadence of accountability. Read more →
Balanced Scorecard#: A performance framework that measures a business across four perspectives: financial, customer, internal process, and learning and growth. Its causal strategy map is the ancestor of the metric tree, and its research gave the field its founding receipt about strategies failing in execution. Read more →
Hoshin Kanri#: A Japanese strategy deployment method that connects long-term direction to annual plans through catchball, the practice of negotiating goals up and down the organisation. Read more →
EOS and Rocks#: The Entrepreneurial Operating System runs a business on six components, with quarterly priorities called Rocks and a weekly Level 10 meeting. Strong on rhythm and ownership, lighter on measurement design than OKRs. Read more →
NCT goals#: NCT replaces OKRs with three layers per quarter: a Narrative of strategic context, three to five Commitments, and the Tasks kept visibly separate from them. Its contribution is making written context a structural part of the goal rather than a preamble people skip. Read more →
Goal-setting theory#: The research finding that specific, difficult goals produce higher performance than vague or easy ones, given feedback, commitment and ability. It also documents the dark side: goals plus incentives produce tunnel vision, gaming and ethical drift. Read more →
Implementation intentions#: A goal is far more likely to be acted on when it is written as if-then: if this situation occurs, then I will do that specific thing. Read more →
Mental contrasting and WOOP#: Wish, Outcome, Obstacle, Plan: pairing a vivid picture of success with an honest look at what stands in the way, which outperforms positive visualisation alone. The fluent, obstacle-free quarter plan is running the condition that loses. Read more →

The strategy the OKRs execute

Goals only inherit their meaning from a strategy. These are the strategy terms we use, credited to the people who wrote them, and free to read.

Strategy Choice Cascade#: Five linked choices that make up a real strategy: winning aspiration, where to play, how to win, must-have capabilities and management systems. OKRs are its execution layer. Read more →
Winning aspiration#: What winning looks like for this company, defined in terms of the customer whose life it changes rather than the revenue it produces. Read more →
Where to play#: The choice of field: which customers, segments, geographies and channels you compete in, and, just as loudly, which you do not. Read more →
How to win#: The theory of advantage: why the chosen customer chooses you over every alternative on your chosen field. Read more →
Must-have capabilities#: The specific things a company must be distinctively good at for its way of winning to hold. Built deliberately, and never the same list as generic excellence. Read more →
The kernel#: A good strategy has three parts: a diagnosis of the challenge, a guiding policy for dealing with it, and coherent action that carries the policy out. Read more →
The crux#: The pivotal, solvable challenge where focused effort will actually cash out, and the point a good strategy concentrates on. Read more →
Bad strategy#: Bad strategy has four hallmarks: fluff, failure to face the challenge, mistaking goals for strategy, and bad strategic objectives. A wall of well-written OKRs is a goals list awaiting a diagnosis. Read more →
The Opposite Test#: A fast filter for a real choice: would a sane rival choose the reverse? If not, it is an operating imperative, not a strategy. Read more →
Strategic pillar#: One of the three to five named battlegrounds a company has chosen to win on, the headline layer that goals turn into quarterly progress. Read more →
Proximate objective#: A goal close enough to be achievable now, the bridge from a multi-year pillar to a single quarter’s work. Read more →
Compound advantage#: A system where today’s success makes tomorrow’s easier and more defensible, so the moat grows while rivals copy it. Read more →

Still fuzzy on any of these?

We define them properly and put them to work in your business, through consulting, coaching and the AI Business OS.

See OKR consulting →

// the questions behind the words

Six questions the definitions do not settle.

QWho invented OKRs?

Andy Grove developed the practice at Intel in the 1970s, where it was called iMBOs, building on Peter Drucker’s Management by Objectives. John Doerr learned it at Intel and introduced it to Google in 1999, where the name OKRs took hold and the method spread.

QWhat is the difference between an OKR and a KPI?

A KPI is a metric you monitor continuously to keep the business healthy. An OKR is a specific change you have chosen to make this cycle. KPIs run always-on and live on a scorecard; OKRs are focused bets with a start and an end. A KPI that breaks, or that you want to push to a new level, is promoted into an OKR, and returns to the scorecard with a new threshold when the work lands.

QHow many OKRs should a team have?

One or two Objectives, each with two to four Key Results. Attention is finite, and teams that run several goals in parallel usually finish none of them well. Fewer goals worked in sequence beat a long list worked at once.

QShould OKRs be linked to pay or bonuses?

Only with great care, and only in one lane. Cash attached to team goals does not add motivation, it converts candour into arithmetic: targets get calibrated to be hit and confidence scores stop being honest. In the SHOP Model only Strategic OKRs ever touch a bonus, through a company gate, while health and operational excellence work is rewarded with recognition and enablement.

QShould OKRs cascade?

Strategic choices should cascade logically. Goals should not cascade bureaucratically. Decomposing a company goal into arithmetic slices and handing them down produces sign-off queues, borrowed numbers and no ownership. Teams should write their own goals against a strategy they can read and recall, which is alignment rather than cascade.

QIs 70% a good OKR score?

Treating 0.7 out of 1.0 as success teaches teams to calibrate targets so they will be hit at 70%, which is sandbagging with a formula. Grade the quarter on evidence instead, from Excellent to Bad. Fifty per cent progress against a genuine stretch can be excellent, and a clean 100% is often a target that was set low.

// go deeper

OKR Examples →

Nine functions, good and bad

OKR Consulting →

Full implementation

OKR Coaching →

Learning by doing

OKR Training →

Create OKR champions

The Knowledge System →

Every term, with the reasoning

// that worked

That Worked

Short reports from the gap between plausible and works.

Definitions are the free layer. What we're learning about using them is the fortnightly one.

Issue #7: Calibrated to be hit. Read previous issues →

Fortnightly. Under 500 words. Every claim carries a receipt. Unsubscribe is one click and no hard feelings. Privacy.

Ready to put the words to work?

Start with a look at your goals. A couple of weeks, small fixed fee, honest answers.

Meet Us

// the install pack

You know the words. Now test the sentences.

The Coach, our free install, checks whether your actual OKRs live up to these definitions, and rewrites them when they don’t.

Get the book and pack →