How to build an OKR presentation: the setting deck, the review deck and the scoring math

Disclosure: NOXI publishes this guide and is one of the tools it mentions; the method works in any presentation software.

The worked example is fictional: Brightmoor Analytics, a 180-person analytics software company setting and then reviewing its Q1 2027 company OKRs. Every number is invented for teaching, and every score is calculated in the open so you can check it or rerun it with your own figures.

What an OKR is: objective, key result and score

An OKR (objectives and key results) is a goal written in two parts: an objective that states, in words, what should be achieved, and two to five key results that measure, in numbers, whether it was achieved by the end of a fixed period, usually a quarter. An OKR presentation is the slide version used to agree those goals and later to report how they scored.

The method has a short and well-documented history. Google's re:Work guide traces it to Andy Grove at Intel, who framed goals around two questions: where do I want to go, and how will I pace myself to see if I am getting there. John Doerr learned the method at Intel and introduced it to Google's leadership in early 2000 (Google re:Work, Set goals with OKRs). Doerr later wrote the standard book on it, Measure What Matters, published by Portfolio on 24 April 2018 (Penguin Random House), and gave the TED2018 talk most people find when they search for "John Doerr OKR presentation" (TED, Why the secret to success is setting the right goals).

Three properties from the primary sources shape every slide in this guide:

• Few objectives. re:Work advises three to five objectives, because more leads to over-extended teams and diluted effort, and around three key results per objective. What Matters describes three to five key results under each objective.
• Graded with a number. At Google, OKRs are usually graded from 0.0 to 1.0, where 1.0 means fully achieved; each key result is graded and the objective takes a rough average (re:Work).
• Public, and not a performance review. re:Work says OKRs are public so everyone can see what others are working on, and that OKRs are not synonymous with employee evaluations. What Matters describes them as divorced from compensation (What Matters, What is an OKR).

Setting deck vs review deck: two jobs, two decks

An OKR presentation is really two presentations. The setting deck, shown before the quarter starts, asks the room to agree the objectives, baselines, targets and owners. The review deck, shown after the quarter ends, reports the scores, explains every missed commitment and says what changes next quarter. Mixing the two in one deck is a common reason OKR meetings run long.

The two decks share one artefact: the OKR table, a single spreadsheet with one row per key result. The setting deck fills its baseline, target, owner and type columns; the review deck adds actual and score. If both decks are built from the same table, nobody can quietly change a target between January and April. The quarterly review deck sits naturally next to a QBR presentation, and the annual objectives that feed it usually come from the annual plan and budget presentation.

The 8-slide OKR setting deck, slide by slide

A setting deck for three company objectives needs eight slides: a title, the baselines, the reason for these objectives, one slide per objective, an alignment slide and a slide on how the OKRs will be tracked and reviewed. Add one slide per extra objective; never go above five objectives.

Three rules make the setting deck decidable in one meeting:

• Baselines before targets. Slide 2 exists so that every target on slides 4 to 6 has a number to start from. A key result without a baseline cannot be scored with the formula later in this guide.
• Titles that state the commitment. Write slide titles as full sentences ("We will make onboarding self-serve: first dashboard in 5 days, not 14"), often called action titles.
• Say what you are not doing. An objective list is also a list of exclusions. One line on slide 7, such as "No new modules this quarter", prevents the meeting from adding a sixth objective.

For a company-wide audience, the setting deck shrinks to three or four slides inside the January meeting: the theme, the objectives on one slide and the review date. Our all-hands meeting presentation guide shows where that block fits in a 45-minute run of show.

Anatomy of a good OKR slide

A good OKR slide has six parts: the objective as the title, a type label (committed or aspirational), and one row per key result showing the metric, its baseline, its target and its owner. During the quarter, add a seventh column for the latest confidence status; at the end of the quarter, replace it with the actual value and the score.

Keep the slide a table, not a chart. Ten numbers in three rows read faster as a table, and a table keeps baseline, target and actual side by side, which is what a reviewer compares. When a key result needs a trend, put the line chart on a backup slide; our guide to presenting data in a presentation covers when a chart beats a table.

Writing key results as outcomes, not tasks

A key result must describe an outcome that can be measured, not an activity that can be completed. Google's OKR playbook, published on What Matters, says key results must describe outcomes, not activities, and that words such as "consult", "help", "analyze" or "participate" signal an activity (What Matters, Google's OKR Playbook).

The test is simple: could the team complete the key result and still fail the objective? If yes, it is a task. Rewrite it as the change the task is supposed to cause.

Tasks still belong somewhere: in the team's plan, on a backup slide or in a project tracker. They do not belong in the key result column, because a quarter of finished tasks can score 1.0 while the objective has not moved. If you need to show the project work behind a key result, our project status report presentation guide covers milestone slides and RAG rules.

Committed vs aspirational OKRs, and how to show them on the slide

A committed OKR is a goal the team agrees will be achieved in full, so its expected score is 1.0. An aspirational OKR is a stretch goal the team does not yet know how to reach, so its expected average is about 0.7. Label every objective with its type on the slide, because the same score means opposite things for the two types.

The playbook names "failing to differentiate between committed and aspirational OKRs" as the first of its classic traps (What Matters, Google's OKR Playbook). On a slide, the type has to be visible before anyone reads a score:

The green, amber and red bands above are this guide's working convention, not a published standard. What Matters notes that many organisations grade with red, yellow and green but gives no numeric thresholds; the bands here are set so that the aspirational green starts at the playbook's 0.7 expectation and so that committed OKRs have no amber at all. Pick your bands once, print them in the deck's footer, and do not move them between quarters.

How to score OKRs: the formula and a worked score

Score each key result as the share of the planned change that actually happened: (actual − baseline) ÷ (target − baseline), capped between 0.0 and 1.0. The objective's score is the average of its key results' scores. The same formula works for metrics where lower is better, because the numerator and denominator are then both negative.

The formula agrees with Google's own example. re:Work describes a key result "launch six new features" that ends with three launched and scores 0.5. With a baseline of 0, the formula gives (3 − 0) ÷ (6 − 0) = 0.5. Starting from a baseline matters as soon as the metric does not start at zero, which is almost always.

The scoring table: Brightmoor's first objective

Two of the three key results are "lower is better" metrics. The formula needs no special case: days fell by 7 out of a planned fall of 9, so the score is 7 ÷ 9. Writing the formula as (actual − baseline) ÷ (target − baseline) everywhere means one spreadsheet column works for every row, and nobody has to remember which metrics run backwards.

Edge cases the formula has to handle

• Worse than the baseline. If activation fell to 36%, the raw score is (36 − 38) ÷ 17 = −0.12. Record 0.0 and explain the fall in the review; a negative score adds nothing to the average except confusion.
• Beyond the target. If tickets fell to 3, the raw score is (3 − 9) ÷ (4 − 9) = 1.2. Record 1.0 and say the target was beaten, then ask whether it was set too low.
• Yes-or-no key results. A launch date or a deadline scores 1.0 if met and 0.0 if not. re:Work describes such binary key results; most committed key results are of this kind.
• Weights. re:Work notes that some key results may be weighted differently. If you weight, print the weights on the slide (for example 50/25/25) and set them in the setting deck, never after the results are in.

An OKR score calculator in six spreadsheet columns

If you searched for an OKR scoring template, this is all it needs. One row per key result; the objective score is an average of the rows that belong to it.

For binary key results, enter baseline 0, target 1 and actual 1 or 0; the same column F then returns 1.0 or 0.0. For the objective score, use =AVERAGE(F2:F4), or =SUMPRODUCT(F2:F4,G2:G4) with weights in column G that add up to 1.

What is a good OKR score, and what 1.0 every quarter tells you

For aspirational OKRs, a good score is about 0.6 to 0.7. Google's re:Work guide calls 60% to 70% the "sweet spot" and says the expectation is an average of 0.6 to 0.7 across OKRs graded on its 0.0–1.0 scale. For committed OKRs, the only good score is 1.0.

The re:Work guide also explains the two failure modes. Consistently scoring lower means too little is being achieved; consistently scoring 1.0 means the OKRs are not ambitious enough (Google re:Work, Grade OKRs). Google's playbook calls the second pattern sandbagging and adds that a committed OKR below 1.0 requires an explanation for the miss (What Matters, Google's OKR Playbook).

One quarter is not a pattern. A single aspirational score of 0.78 is a good quarter; four quarters at 0.95 are a calibration problem. Put the last four quarters' scores for each recurring objective on a backup slide, and the pattern becomes visible without an argument.

The 6-slide quarterly OKR review deck

A quarterly OKR review deck needs six slides: a scoreboard, the committed results, the aspirational results, the lessons, what changes, and next quarter's draft. Committed and aspirational OKRs get separate slides so that a missed commitment cannot hide inside a respectable average.

Worked example: Brightmoor's full Q1 2027 scoreboard

Here are all nine key results, scored with the same formula. Objective 1 is the scoring table above; objectives 2 and 3 follow.

• Objective 1 (aspirational): (0.78 + 0.76 + 0.80) ÷ 3 = 0.78. Green.
• Objective 2 (aspirational): (0.38 + 0.69 + 0.67) ÷ 3 = 1.74 ÷ 3 = 0.58. Amber: real progress on adoption and pipeline, little on retention.
• Objective 3 (committed): (1.0 + 0.0 + 1.0) ÷ 3 = 0.67. Red: the expected score was 1.0.
• Aspirational average: (0.78 + 0.58) ÷ 2 = 0.68, inside the 0.6–0.7 range.

Now the trap. The blended average of all three objectives is (0.78 + 0.58 + 0.67) ÷ 3 = 2.03 ÷ 3 = 0.68, the same number. A deck that shows only "Q1 OKR score: 0.68" tells the board it was a textbook quarter, while the company broke one of its three commitments. That is why the scoreboard has no total.

The committed miss gets its own arithmetic on slide 2. February 2027 has 28 days, or 40,320 minutes. A 99.9% commitment allows 0.1% × 40,320 = 40.3 minutes of downtime; 99.82% uptime means 0.18% × 40,320 = 72.6 minutes, rounded to 73. The slide states the overrun (72.6 − 40.3 = 32.3, about 32 minutes), the cause (in the fictional example, a database failover during the third migration wave) and the fix with its owner and date. Our incident postmortem presentation guide gives the full format when a miss deserves its own review.

The lessons and changes follow from the numbers. Activation reached 51% against a 55% target, a gap of 4 points (55 − 51), so Q2 keeps objective 1 and adds an in-app setup checklist test. Net revenue retention moved 3 of a planned 8 points, and retention moves slowly inside one quarter, so Q2 changes key result 2.1 to a leading indicator the team can move in 13 weeks, such as the share of mid-market accounts with a renewal plan agreed 90 days before renewal.

Here are two slides from this article's worked example, as NOXI built them, unedited.

The alignment slide: show contribution, don't cascade every OKR

The alignment slide shows which teams contribute to which company key result, in one grid, instead of copying every team's OKRs into the leadership deck. re:Work advises that not every organisational OKR needs to appear in every team's OKRs, but each team OKR should connect to at least one organisational OKR.

Read the grid by row and by column. A row with no entries is a team with no stake in the company's priorities, which needs a conversation. A column with only "supports" and no "leads" is an objective nobody owns. The empty cells are the point of the slide: they show the room that focus was chosen, not forgotten.

OKRs vs KPIs on the same slide

KPIs measure the health of what the business already does; OKRs measure the change it wants to make this quarter. What Matters puts it as OKRs being measures for change and KPIs measures of health (What Matters). A good OKR deck shows both, on separate parts of the slide, so that a KPI that slips while everyone chases OKRs is noticed.

A metric can move between the two. At Brightmoor, uptime is a KPI in a normal quarter; during the migration it became committed key result 3.2, because the company had decided to change its platform without letting that number slip. When the migration closes, uptime goes back to the health strip. Saying this out loud in the setting deck avoids the question "why is uptime an OKR now?"

How many OKRs: company, team and individual

Set three to five objectives with about three key results each, which is re:Work's guidance; that is 9 to 15 key results for a company and fewer for a small team. A leadership deck should show company OKRs and the alignment grid, not every team's or individual's OKRs.

Individual OKRs rarely belong in a leadership deck, and they should not be used as a scorecard of people. re:Work states that OKRs are not synonymous with employee evaluations, and What Matters describes them as divorced from compensation. If you prepare a review of your own year, our performance review presentation guide is the right format; it can mention OKR results as context without turning a 0.6 stretch score into a poor rating. For the board, put the company scoreboard into the operating section of the board meeting presentation.

The OKR calendar for Q1 2027, with dates

For a quarter that starts on Friday 1 January 2027, brainstorm company objectives in November 2026, present the setting deck to leadership in early December, share company OKRs before the holidays, present team OKRs in the first January meeting, check in mid-quarter and review in the first week of April. re:Work's own sample timeline follows the same November-to-February rhythm.

re:Work describes mid-quarter and end-of-quarter check-ins and notes that some teams revisit OKRs a few times a quarter, to abandon objectives that clearly will not happen and to move resources to borderline ones. Whatever you adjust at the mid-quarter check, record it on the slide with the date; a target changed silently in February makes April's score meaningless.

Seven mistakes leaders flag, as a pass/fail checklist

Before an OKR deck goes to leadership, check it against seven pass/fail tests. Each one corresponds to a mistake that Google's playbook, the re:Work guide or the review meeting itself will expose.

Two more mistakes show up only in the review deck: scoring people instead of goals, and reporting one blended number. Both are covered above; both are fixed by the layout, not by better wording.

3 AI prompts for an OKR presentation

The three prompts below work in NOXI or in any AI presentation tool. Replace the brackets, keep your OKR table as the single source of numbers, and check every score against it before the deck is shared.

How to build the OKR decks in NOXI

NOXI can turn your OKR spreadsheet into the setting deck and, at the end of the quarter, into the review deck, in the style of your company template. It is free, with no card and no watermark; uploading files needs a free account, and a full deck uses about 25 of the 1,000 monthly credits.

Frequently asked questions

How do you present OKRs?

Present OKRs in two decks. Before the quarter, a setting deck shows the baselines, why these objectives, one slide per objective with key results as baseline → target and an owner, an alignment grid and the review dates. After the quarter, a review deck shows the scores, the committed misses with explanations, the lessons and what changes next quarter.

How many OKRs should a team have?

Three to five objectives with about three key results each, according to Google's re:Work guide, which warns that more objectives over-extend teams and diffuse effort. A small team often needs only two or three objectives. If an important piece of work is not covered, add it to an existing objective's key results before adding a new objective.

What is a good OKR score?

For aspirational OKRs, about 0.6 to 0.7 on a 0.0–1.0 scale; Google's re:Work guide calls 60% to 70% the sweet spot. For committed OKRs, the expected score is 1.0, and anything lower needs an explanation. Scoring 1.0 on aspirational OKRs every quarter usually means the targets were set too low.

How do you calculate an OKR score?

Score each key result as (actual − baseline) ÷ (target − baseline), capped between 0.0 and 1.0, then average the key results to score the objective. In the fictional Brightmoor example, activation moved from 38% to 51% against a 55% target: 13 ÷ 17 = 0.76. Yes-or-no key results score 1.0 or 0.0.

What is the difference between OKRs and KPIs?

KPIs measure the ongoing health of the business, such as gross margin or churn, and stay in place every month. OKRs measure a change the company wants this quarter, from a baseline to a target, and close at the end of the period. A good OKR deck shows KPIs as a separate health strip, so a slipping KPI is noticed.

How do you make an OKR slide in PowerPoint?

Put the objective in the title, add a committed or aspirational tag, and insert a table with one row per key result and columns for metric, baseline, target and owner. Add a status column during the quarter and actual and score columns at the end. Keep it a table rather than a chart so baseline, target and actual sit side by side.

How often should OKRs be reviewed?

Update confidence weekly, check in formally at mid-quarter, and grade at the end of each quarter. Google's re:Work guide describes mid-quarter and end-of-quarter check-ins and notes that some teams revisit OKRs a few times a quarter. Record any target changed at the mid-quarter check, with its date, on the slide.

Should OKRs be tied to performance reviews?

No. Google's re:Work guide says OKRs are not synonymous with employee evaluations, and What Matters describes them as divorced from compensation. Tying scores to pay or ratings pushes people to set targets they know they can hit, which defeats aspirational OKRs. Discuss OKR results as context in a review, never as the rating itself.

What should an OKR review presentation include?

Six slides: a scoreboard with each objective's type and score and no blended total, the committed key results with an explanation for each miss, the aspirational key results with scores, two or three lessons tied to numbers, keep, change or drop for each objective, and next quarter's draft objectives with the decisions needed.

Sources & methodology

How this guide was made: the definitions, scoring scale, expected scores, number of objectives and the position on performance evaluation come from Google's re:Work guide, Google's OKR playbook as published on What Matters, and What Matters' own definitions. The two-deck structure, the scoring formula with baselines, the colour bands, the calculator, the checklist and the slide layouts are editorial recommendations, not a standard. Brightmoor Analytics and all its figures are fictional; every score was checked in a spreadsheet. All sources were checked on 14 October 2026.

• [1] OKR history (Andy Grove, John Doerr, Google in early 2000), 0.0–1.0 grading with a rough average, the 0.6–0.7 sweet spot, three to five objectives and about three key results, transparency, OKRs not employee evaluations, the six-features example, mid-quarter check-ins, team OKRs linked to organisational OKRs and the sample November–February timeline — Google re:Work, Guides: Set goals with OKRs (first published 2016).
• [2] Committed OKRs expected to score 1.0 with an explanation for any miss, aspirational OKRs expected to average 0.7 with high variance, key results as outcomes not activities, and the classic OKR traps — What Matters, Google's OKR Playbook.
• [3] "I will [objective] as measured by [key results]", three to five key results per objective, red, yellow and green grading, OKRs as measures of change and KPIs as measures of health, OKRs divorced from compensation — What Matters, What is an OKR? Definition and examples (Ryan Panchadsaram).
• [4] Book reference: Measure What Matters by John Doerr, Portfolio, 24 April 2018 — Penguin Random House, Measure What Matters.
• [5] John Doerr's talk introducing OKRs, TED2018 — TED, Why the secret to success is setting the right goals (April 2018).

Build your OKR setting and review decks in NOXI, free: attach your OKR table, match your company template, and export an editable PowerPoint or PDF.