OKRs vs KPIs: definitions that survive a real quarter
A KPI, or key performance indicator, is a standing measure of whether the marketing function is healthy. It has no expiry date and no finish line. It has a range the team agrees is acceptable, and it stays on the dashboard whether it's green, amber or red. Pipeline coverage of 3x is a KPI. So is a 14-month CAC payback, a 24% win rate, a marketing-sourced share of pipeline.
An OKR is a different instrument. Objectives and Key Results came out of Intel under Andy Grove in the 1970s and were written up in his book High Output Management (1983). John Doerr carried the method to Google in 1999 and documented it in Measure What Matters (2018), where he reduces the format to one sentence: "I will [objective] as measured by [key results]." The objective is the change you're committing to. The key results are the two to four numbers that prove it actually happened, inside a fixed window, usually a quarter.
So the honest one-liner is this. KPIs describe the state of the function. OKRs describe an intervention in it. Confusing the two is how a marketing team ends up with a beautifully maintained dashboard and no answer to the question of what, specifically, it is trying to change before December.
| Dimension | KPI | OKR |
|---|---|---|
| Question it answers | Is the function healthy right now? | What are we deliberately changing this quarter? |
| Time horizon | Ongoing, no expiry | Time-boxed, typically one quarter |
| What good looks like | Sitting inside an agreed healthy range | Movement from a stated baseline to a stated target |
| How many | As few as let a marketing lead make decisions, ideally one screen | Three to five objectives, roughly three key results each (Google re:Work guidance) |
| What happens when it's met | Nothing. It keeps being watched. | It's scored, retired, and replaced next quarter |
| Marketing example | Marketing-sourced pipeline, CAC payback, win rate, pipeline coverage | Objective: make inbound the default first touch for enterprise deals. KR: enterprise inbound-sourced pipeline from 1.2M to 2.5M. |
| Owner | The function, agreed with finance and sales | One named person per key result, negotiated up and down |
| Failure mode | Watching so many numbers that none of them drive a decision | Rewriting last quarter's dashboard with a deadline attached |
They overlap on purpose, and that's fine
The confusion is understandable, because key results are usually expressed in the same units as KPIs. Marketing-sourced pipeline shows up on the dashboard and inside the OKR doc. Teams see the duplication and conclude that one of the two systems is redundant.
It isn't. The metric is the same; the contract around it is not. A KPI has a range and no end date. The same number becomes a key result the moment it gets four things attached: a baseline, a target, a deadline, and one named owner. Marketing-sourced pipeline is a KPI when it's reviewed monthly and nobody's job depends on where it lands. It's a key result when someone has said out loud that it goes from 1.2M to 2.5M by 30 September and their quarter is scored on it.
Which is why the useful question is never "should this be a KPI or an OKR". It's "is anyone changing their behaviour because of this number". If yes, it belongs in the OKR doc for as long as that's true. If no, it belongs on the dashboard, where it's doing the quieter and equally necessary job of telling you when something breaks.
Why marketing teams turn OKRs back into KPIs
Here's the pattern I see most often. Quarterly planning starts, someone opens the dashboard, and the OKR doc gets populated by copying the metrics that were already there and adding a number to each one. Objective: improve marketing performance. Key result: 400 MQLs. Key result: 1.5M pipeline. Key result: 12 published articles.
None of that is an objective. It's a forecast wearing a costume. The diagnostic is simple and slightly uncomfortable: if the key result would land in roughly the same place whether or not anyone changed how they work, it's a KPI, not a key result. A team that runs 400 MQLs a quarter every quarter and then writes "400 MQLs" as a key result has committed to nothing except continued existence.
The second tell is the score. Google's re:Work guidance treats 60% to 70% as the expected landing zone for aspirational OKRs, on the logic that consistently hitting 100% means the bar was set where the team already knew it could reach. A marketing team scoring green across the board four quarters running isn't excellent. It's sandbagging, usually because somewhere in the building OKR scores are quietly feeding bonus conversations. (Doerr is explicit about this in Measure What Matters: decouple OKRs from compensation, or the system teaches people to aim low.)
And the third tell is the output key result. "Publish 24 articles" is a task list with a number on it. It can be hit in full while organic pipeline stays exactly flat, at which point the team has proof of effort and no proof of effect. Where the team genuinely controls the lever inside the quarter, write the key result as the outcome, not the activity that's supposed to cause it.
How to run both without drowning the team in numbers
Most marketing teams don't have a measurement problem. They have a count problem. The instinct when reporting feels shaky is to add a metric, and the result after two years is a dashboard nobody reads and an OKR doc with eleven objectives in it. Four rules keep both systems usable.
- Cap the OKRs harder than the framework tells you to. Google's re:Work guidance says three to five objectives per team per quarter. For a B2B marketing function that also has to keep the standing engine running, two objectives with two or three key results each is usually the real ceiling. Anything beyond that is a planning artefact, not a plan.
- Keep the KPI set to one screen. If the marketing lead can't read the whole dashboard without scrolling, the dashboard has stopped being an instrument panel and started being an archive. The full argument for which numbers earn that space is in the marketing KPIs that matter for B2B.
- Give the two different cadences. KPIs get reviewed monthly, with two or three pipeline lines watched weekly. OKRs get a mid-quarter honesty check and a score at quarter close. Reviewing OKRs weekly turns them into a status meeting; reviewing KPIs quarterly means you find out about a broken channel eight weeks late.
- Keep them in two separate documents. The dashboard is the instrument panel. The OKR doc is the route. Merging them into one master sheet feels tidy and reliably produces a document that answers neither question.
One more thing worth saying plainly, because it's where OKR programmes usually go to die: do not cascade a personal OKR set onto every individual. Team-level objectives with named owners per key result is enough structure for a marketing function of any realistic size. Individual OKR sheets for six people generate roughly forty numbers a quarter, most of which exist to be reported rather than to be changed.
What to do when the OKR is green and the KPI is red
This is the moment the whole thing gets tested, and it happens more often than the frameworks admit. The content key result landed at 100%. Organic pipeline, the KPI sitting underneath it, didn't move. Or the demand-gen objective scored 0.8 while CAC payback stretched from 14 months to 19.
The rule is that the revenue-side KPI wins, always. A key result is a hypothesis about what would improve the health of the function. When the key result lands and the health metric doesn't follow, the hypothesis was wrong, and the correct response is to change next quarter's objective rather than to celebrate the score. This is why the trust order between numbers has to be written down before the numbers disagree, which is the core of a marketing measurement framework that survives a board meeting.
The failure worth naming: teams that treat the OKR score as the report. A green quarter with a flat pipeline is not a green quarter. It's a well-executed guess about the wrong lever, and knowing that is more valuable than the score was. Objectives that follow directly from a settled GTM strategy rather than from last quarter's dashboard tend to guess better, because the lever has already been argued about once.
KPIs keep you honest about where you are. OKRs keep you honest about where you said you'd go. Run only KPIs and you manage a dashboard. Run only OKRs and you'll score a beautiful quarter that nobody can find in the revenue line.
Keep reading: The marketing KPIs that matter for B2B · Marketing measurement framework for B2B · The first marketing hire
Frequently asked questions
What is the difference between an OKR and a KPI in marketing?
A KPI is a standing measure of how healthy the marketing function is, watched continuously and judged against an agreed range: CAC payback, win rate, pipeline coverage. An OKR is a time-boxed commitment to change something, usually within one quarter, written as an objective plus two to four key results that prove the objective happened. KPIs tell you where you are. OKRs tell you where you said you would go.
Can a KPI be used as a key result?
Yes, and often it should be. The difference is not the metric, it is the contract around it. A KPI has a healthy range and no end date. The same metric becomes a key result when it gets a baseline, a target, a deadline, and one named owner. Marketing-sourced pipeline is a KPI when it is watched every month and a key result when the commitment is to move it from a stated number to a stated number by a stated date.
How many marketing OKRs should a team set per quarter?
Google's re:Work guidance recommends three to five objectives per team per quarter, with roughly three key results each. For most B2B marketing teams that is still generous. Two objectives a quarter, each with two or three key results, is usually the honest ceiling for a function that also has to keep the standing engine running. A long OKR list is a planning artefact, not a plan.
Should marketing OKRs be tied to performance reviews and bonuses?
No. In Measure What Matters (2018), John Doerr argues for decoupling OKRs from compensation, because tying stretch goals to pay teaches teams to set targets they already know they can hit. Score OKRs to learn what happened, and evaluate people on a wider view that includes judgement, collaboration, and the standing KPIs they own.
Not sure whether your OKRs are goals or just last quarter's dashboard?
Every Focus4ward engagement starts with an audit. Two weeks to map what marketing is actually measured on, which numbers drive decisions, and the two objectives worth committing to next quarter. A diagnostic, not a pitch.
Book a call