Founders

How to Set Startup OKRs: Steps, Cadence and Scoring

OKRs, short for objectives and key results, are a way to pick a few goals for a set period and track them with numbers. For a startup the useful version is small: one or two objectives, a few measurable key results under each, a weekly check-in and an honest score at the end of the cycle.

What OKRs are

An objective says where you want to end up, in plain words that people can remember. A key result is a number that shows whether you got there: a metric that moves from a starting value to a target by a date. The work you do to move that number is a separate thing, usually called an initiative.

The method is credited to Andrew Grove, a co-founder of Intel, and was brought to Google by the venture capitalist John Doerr, according to a U.S. Chamber of Commerce explainer on the model. Doerr’s book Measure What Matters remains the standard reference, and Heavybit’s 2020 interview with Heather Doshay of Webflow recommends it as a starting point.

OKRs are not the same as KPIs. A KPI is a health metric you keep an eye on all the time, such as monthly recurring revenue or churn. An OKR picks one thing you want to change this cycle. Plenty of KPIs sit in a dashboard (see how to build a startup financial dashboard) while one or two of them get promoted into key results for a few weeks.

When a startup needs OKRs

Not on day one for a team of two who talk all day. Some guides say you can start as a company of one, and the Heavybit interview agrees that the early value is building a rhythm: someone owns the process, goals are written down and people look at them every week. The same interview warns that rolling OKRs out to everyone at once, at 150 people in her experience, was too much, and that a single team can go first while others follow.

A reasonable trigger is the moment two people could give different answers to “what matters most this month?”

How to set OKRs, step by step

  1. Pick the priority. Choose one company objective, two at most. Startup Super School suggests one to three objectives per quarter, and other guides land at two or three. Start at the low end.
  2. Write the objective as an outcome. Use a short sentence with a verb and a reason, such as “Make onboarding the fastest path to value.” Avoid slogans nobody could check.
  3. Add two to five key results. Each needs a metric, a baseline, a target and a date. If you cannot state today’s number, the key result is not measurable yet.
  4. Give every key result one owner. One named person updates the number and raises blockers.
  5. Separate tasks from results. “Launch a new onboarding flow” is a task. “Raise 7-day activation from 34% to 55%” is a result. Startup Super School uses that exact pair in its SaaS example.
  6. Let teams respond. Leadership sets direction, and teams propose their own OKRs that connect to at least one company key result. Hold that until the second cycle if the first one is still shaky.
  7. Post them where everyone sees them. A shared document or spreadsheet is enough. Several guides, including Heavybit’s, say not to spend the first cycle choosing software.

Alignment and measurable goals

The point of the exercise is alignment. When everyone can read the same short list of measurable goals, it is easier to say no to work that does not help the team achieve them. A good key result is measurable, has an owner and is hard enough that nobody is sure it will be achieved on the first try. Company OKRs give teams a target to link their own goals to, and team OKRs show leadership where the work is going.

Startup OKR examples

The numbers below are illustrations built from patterns in the guides listed at the end. Replace them with your own baselines.

Before product-market fit:

  • Objective: Learn whether small teams will pay for the product.
  • Key results: complete 25 customer interviews with the target segment; get 10 paid pilots at a set minimum price; reach a weekly active use rate you chose in advance among pilot users.

If this stage is where you are, customer discovery is the work behind those numbers.

Activation and retention (adapted from Startup Super School’s SaaS example):

  • Objective: Make onboarding the fastest path to value.
  • Key results: 7-day activation from 34% to 55%; time to first value action from 4.2 days to under 2; 30-day retention from 61% to 72%.

The terms are explained in SaaS math on activation, retention and churn.

Engagement (modeled on Liz Wessel’s founder post):

  • Objective: Improve engagement with the product.
  • Key results: daily active users from 1,000 to 2,000; churn from 8% to 4%; longer average sessions.

Funding and runway:

  • Objective: Reach a cash position where the next raise is optional.
  • Key results: cut monthly net burn to a stated figure; extend runway to a chosen number of months; sign a term sheet or close a defined amount. See burn rate and runway for the arithmetic behind the first two.

Notice what is missing from all four: “ship feature X” and “run five campaigns.” Those are initiatives that might move the number.

OKR cadence for a small team

Most guides run quarterly cycles with weekly check-ins. A few argue that early-stage teams do better with six- or eight-week cycles because priorities change quickly. Either works. Pick one length and keep it for at least two cycles.

A simple rhythm:

  • Start of cycle. A planning session of an hour or so: objective, key results, owners, baselines.
  • Every week. A short check-in, 10 to 30 minutes depending on the guide. Each owner reports the current number, any blockers and a confidence score. Startup Super School uses a 0 to 10 confidence score and talks through anything under 5.
  • Middle of the cycle. A review of key results that are off track. Kill, change or push harder, and change a target only if the business context truly shifted.
  • End of cycle. Score each key result and hold a retrospective. Carry what you learned into the next set.

Your board may also want to see the results; startup board meetings covers what usually goes in the pack.

How to score OKRs

Score each key result from 0.0 to 1.0, where 1.0 means the target was fully reached. Several guides treat roughly 0.6 to 0.7 as a healthy score for ambitious goals, and Heavybit’s interview says a team that achieves 70% of its goal is doing well. Constant 1.0 scores usually mean the targets were too easy. Scores of 0.4 or lower deserve a look at whether the target was unrealistic or something blocked the work.

Not every goal should be a stretch: anything tied to a commitment, such as making payroll or meeting a covenant, belongs outside the 70% logic.

Common OKR mistakes

  • Too many objectives. The most frequent complaint across the guides: if the team cannot name the top priority, there are too many.
  • Tasks written as key results. An OKR vendor’s analysis of its own customers, published at okrstool.com, found that 52% of key results it examined were tasks or existing metrics. The data is vendor data, so read it as a hint.
  • No owner. The same vendor found about half of key results had no owner.
  • Never updating. Same source: teams that kept key results updated hit their goals far more often than those that did not (68% versus 35%). That is correlation from one platform, not proof.
  • Tying OKRs to pay or ratings. Several guides say it pushes people toward safe targets. Keep the OKR review apart from the performance review, as the Heavybit interview also advises.
  • Only watching lagging numbers. Revenue moves late. Add a leading indicator such as trial sign-ups or feature adoption.
  • Vanity metrics. Liz Wessel lists examples: leads with no quality data, traffic with no conversion rate, raw sign-ups instead of trial-to-paid conversion.
  • Perfectionism. Heavybit’s interview says the first round can be messy. A best guess at the baseline is fine for the first one or two cycles.
  • Quitting early. One cycle will probably disappoint. The vendor data above suggests completion improved over later cycles, though again from a single company’s customers.

Questions founders ask

How many OKRs should a startup have? Most sources point to one to three objectives with two to five key results each. Starting with one objective and three key results is easier to keep up.

Quarterly or shorter cycles? Quarters are the default. Six or eight weeks can suit a team whose priorities move fast. Consistency matters more than length.

Do OKRs replace a financial model or budget? No. They sit on top of them. The targets in your key results should be consistent with the plan; what makes a good financial model is a useful check.

Can a solo founder use OKRs? Yes, as a weekly habit. Write one objective, three numbers and look at them every Monday. For more on running the company day to day, see the founders section.

Sources and limits

Sources read for this page, all fetched in November 2026: Heavybit, interview with Heather Doshay (January 2020); Startup Super School, “How to Set Company OKRs”; U.S. Chamber of Commerce, “Why Business Owners Need OKRs”; Liz Wessel, “OKRs and KPIs for Startups”; okrstool.com, OKR best-practice analysis (vendor data); and several other practitioner guides on startup OKRs. Most are vendor or consultant blogs without peer review, so the 0.7 scoring target, the number of objectives and the cycle length should be treated as common practice, not as proven rules. Example numbers are illustrative. This is general information, not advice for your company.