OKRs vs. KPIs: How They Work Together

OKR vs KPI is often framed as a choice, and it is a false one. KPIs measure the ongoing health of the business: the dials you watch every month to be sure the machine is running. OKRs, which stand for objectives and key results, are a goal-setting method for driving change: where you want to go this quarter and how you will know you got there. A company needs both. KPIs tell you where to look, OKRs tell you what to change, and many key results are simply KPIs with a new target. This guide compares the two, works through an OKR with a scoring example and checked arithmetic, shows how the two fit together, covers cadence and common mistakes. It belongs to business analytics for finance and operations, follows leading vs. lagging indicators and operations KPIs, and leads to the KPI tracking template and KPI examples by department. OKRs often tie into the targets set in budgeting and forecasting.

The difference at a glance

How they fit together

KPIs describe the baseline. When a KPI is persistently off target or needs a step change, it becomes the subject of an OKR. The key result states the movement: "Raise on-time in-full delivery from 91 percent to 96 percent by the end of the quarter." After the objective is achieved, the KPI stays on the scorecard to make sure the gain holds, and the OKR retires. Some key results are not KPIs at all, such as "launch the new replenishment process in all three warehouses," which is a milestone and belongs to the OKR alone.

A simple rule: put the things that must keep working on the KPI scorecard, and put the things that must change on the OKR list. Companies that confuse the two end up with OKRs that restate routine targets or KPIs that change every quarter.

A worked OKR

Objective: Make our delivery reliable for our top customers.

Key results:

1. Raise on-time in-full delivery from 91 percent to 96 percent.
2. Cut stockout days on the top 100 items from 14 per month to 7.
3. Complete the safety stock reset for the top 100 items by week 6.

At the end of the quarter the results are 94 percent on-time in-full, 10 stockout days and the safety stock reset completed in week 7 (hypothetical). Scoring each key result as progress from the starting value toward the target:

• KR1: (94 - 91) / (96 - 91) = 3 / 5 = 0.60
• KR2: (14 - 10) / (14 - 7) = 4 / 7 = 0.57
• KR3: milestone completed a week late: scored 0.8 by agreement for late completion

The objective score is the average: (0.60 + 0.57 + 0.80) / 3 = 1.97 / 3 = 0.66. In many OKR practices, 0.6 to 0.7 on a stretch objective is a good result: it shows real progress and that the target was ambitious. The discussion afterwards matters more than the score: what worked, what did not and whether to carry the objective into the next quarter. The on-time in-full KPI remains on the scorecard at the target of 96 percent, and the gap of two points becomes the carry-over.

What makes a good key result

• Measurable. A number and a deadline, with a starting value.
• Outcome-focused. Prefer "reduce stockout days" over "hold weekly meetings," which is an activity. Milestones are acceptable when the outcome cannot be measured within the period.
• Within the team's influence.
• Ambitious but not absurd. A target that is certain to be hit is not a stretch, and one that is impossible is demoralizing.
• Few. Two to five per objective, and only a few objectives per team.

Cadence

Running both on the same cadence reduces meeting load: the monthly business review looks at the KPI scorecard first and then the OKR progress.

Alignment

OKRs cascade from company priorities to teams, but not mechanically. A company objective such as "Improve profitability of core products" may lead the operations team to an objective on freight and inventory, the sales team to one on discount discipline, and finance to one on pricing analytics. Teams propose their own OKRs in light of company priorities, and the leadership team checks for conflicts and gaps. Avoid linking OKR scores directly to pay, which tends to make people choose safe key results. Use them as a basis for conversation about performance.

Common mistakes

• Treating OKRs as a to-do list.
• Duplicating KPIs as OKRs every quarter.
• Too many objectives.
• Key results that are activities, not outcomes.
• No baseline, so progress cannot be scored.
• Linking scores to compensation and then sandbagging.
• Setting and forgetting: no check-ins.

A finance and planning OKR

OKRs are not only for sales and product teams. Consider a planning team that wants to make its forecast something leaders trust.

Objective: Make the forecast reliable enough to run the business on.

Key results:

1. Cut the three-month revenue forecast error (mean absolute percentage error) from 12 percent to 8 percent.
2. Reduce days to close from 9 working days to 6.
3. Publish a rolling 13-week cash forecast by week 4 of the quarter.

At the end of the quarter the results are 9.5 percent error, 7 days to close, and the cash forecast published in week 4 (hypothetical). Scoring: KR1 is (12 - 9.5) / (12 - 8) = 2.5 / 4 = 0.625; KR2 is (9 - 7) / (9 - 6) = 2 / 3 = 0.667; KR3 is complete, 1.0. The objective score is (0.625 + 0.667 + 1.0) / 3 = 2.292 / 3 = 0.76. That is above the usual 0.6 to 0.7 range for stretch goals, which prompts a question: were the key results ambitious enough? If a team scores 1.0 every quarter, it is probably setting safe targets. A good retrospective asks what the team learned, what blocked the remaining progress and what the next objective should be. The KPIs behind these key results, forecast accuracy and days to close, stay on the scorecard.

Mid-quarter check-ins

OKRs lose their value when they are set in week one and revisited in week thirteen. Run a short weekly or biweekly check-in. Each key result owner reports the current value, their confidence on a scale of 1 to 10 that the target will be met and any blockers. A drop in confidence is an early warning, and the conversation should be about what help is needed. At the halfway point, hold a longer review to decide whether to adjust plans or drop a key result that has become irrelevant. Changing a key result is acceptable when circumstances change materially, but the change should be recorded and explained, not made silently to improve the score.

When to use a KPI instead

Not everything belongs in an OKR. Use a KPI when the goal is to keep a measure in a healthy range, such as keeping days sales outstanding under 45 days. Use an OKR when you want to change the level, such as bringing it from 52 to 45 this quarter. Ask: is this a state to maintain or a change to make? If the team is already at target, adding it as a key result produces a trivial 1.0. If the team is far from target and needs a project to close the gap, it is a candidate for an OKR. Review the list each quarter, and retire OKRs that have been achieved.

FAQs

What is the difference between OKRs and KPIs?

KPIs monitor the ongoing performance of the business. OKRs set ambitious goals for change over a period, with measurable key results.

Can a key result be a KPI?

Yes. A key result is often a KPI with a new target and a deadline.

Should I use OKRs and KPIs together?

Most companies benefit from both: KPIs for monitoring and OKRs for focusing effort on improvement.

How are OKRs scored?

Each key result is scored from 0 to 1 based on progress from the starting value toward the target, and the objective score is the average. Scores of 0.6 to 0.7 on stretch goals are often considered good.

How many OKRs should a team have?

Typically one to three objectives, each with two to five key results.