OKRs done right: Christina Wodtke's framework for focus and execution

Executive overview

Most OKR rollouts fail not because the system is flawed, but because companies implement it too fast, without the underlying conditions for it to work. OKRs are a vitamin, not a medicine — they amplify a healthy organisation but expose a broken one.

The core cadence is simple: commit on Mondays, celebrate on Fridays, grade and learn at quarter-end. The real value isn't the document — it's the weekly habit of asking "what am I doing this week to move toward the outcome I actually want?"

OKRs create focus, alignment, and a learning cycle — but only if the organisation already has strategy, empowered teams, and psychological safety.

What OKRs actually do

• Main benefit: concrete quarterly action where strategy stays abstract
• Creates a cadence of progress — forces regular check-ins against what matters
• Eliminates ambiguity about the single most important priority
• Generates a learning cycle: grade at quarter-end, carry insights forward
• Scales founder decision-making — teams own execution, leaders own direction
• Knowledge compounds: retention focus this quarter informs acquisition work next quarter

The atomic unit

• The real question isn't "do we have OKRs?" — it's "what are we doing this week to get closer to our goals?"
• Temporal landmarks (Mondays, quarters) create natural forcing functions to raise your head above the noise
• Ambiguous outcomes beat precise tasks: focus on why, not what

Mission, vision, strategy, OKRs — how they connect

• Mission: what you want to see happen over ~5 years; keep it specific enough to guide decisions
• Strategy: a strongly held hypothesis about how to win and fulfil the mission; answers questions like subscription vs. one-off, physical vs. digital
• OKRs: quarterly manifestation of strategy — what does winning look like this quarter?
• Roadmap: leave Q2+ OKRs loosely defined until Q1 results are in; too much upfront planning ties you down
• "Half-built strategy" — enough direction to act, enough slack to react to new information

Writing good objectives and key results

• Objective: one quarter's inspiring direction; should make you want to get out of bed, not hit snooze
• Bad objectives are either too fluffy ("delight customers") or too boring ("ship this feature")
• Key results answer "how do we know we succeeded?" — outcomes, not tasks
• Aim for three key results: one hard number, one quality/delight signal, one revenue signal
• Common mistake: key results are just task lists; a task has no outcome, only completion
• If a key result is binary (approved/rejected), ask whether the outcome is genuinely hard to achieve — it may still be valid
• Brainstorm every possible way to measure an outcome for 10 minutes; the weird ideas often surface the best signals
• Target ~70% achievement; a good goal should feel uncomfortable but not doomed

The weekly cadence

• Monday: commit — what will you do this week to move the ball forward? Send to team or accountability group
• Friday: celebrate — what was the most awesome thing that happened this week? This alone shifts culture
• Weekly status format: confidence level on each key result, last week's progress, next week's plan, blockers
• Blockers are where the learning lives — "I tried X but Y got in the way" builds organisational knowledge fast
• Weekly OKR check-in in meetings: 10 minutes max after the first few weeks; flag what's off, move on
• Repetition is retrieval practice — OKRs move from working memory to long-term memory through weekly review

Grading and retrospectives

• Grade at end of quarter; precision is less important than honesty — a rough "~80%" is fine
• Retrospective matters more than the grade: what went wrong, what got in the way, what would you do differently?
• Don't build elaborate measurement systems; speed and learning outweigh numerical precision

Rolling out OKRs

• Start with your best multidisciplinary team, not a struggling one
• OKRs will not fix a broken team; they will make the problems visible faster
• Give the pilot team 3 months, then debrief: what worked, what didn't, what should we adapt?
• Spread to two more teams, then two more; adopt with the management team in parallel
• Approval process: instead of boss sign-off, get three peers who know your work to review with a 24-hour turnaround
• Planning time: aim for one week per quarter to grade old OKRs and set new ones
• The most common failure mode: someone read Measure What Matters, got excited, implemented it too fast, and it collapsed

Signs your OKR process is broken

• Meetings are boring — you're reviewing task lists, not discussing whether strategy is working
• Teams can't make decisions without escalating — strategy is unclear or leaders aren't delegating
• The whole company is confused — if it's everyone, the problem is probably at the top
• People bring small problems to leadership instead of solving them — fear, not incompetence
• Use the five whys: trace any OKR failure back to its root cause (hiring, trust, clarity, psychological safety)

Measuring the hard things

• Everything can be measured to a useful degree, even if not precisely
• Start with off-the-shelf tools (NPS, DAU, retention) and iterate toward what's meaningful for your company
• Qualitative research is irreplaceable for understanding why users behave as they do
• Retention signals are always worth chasing — happy users stay, sell, and tell you when you're wrong
• Avoid the temptation to skip hard measurement; that's where the strategic differentiation lives

Product management fundamentals

• PMs serve the business first — user focus matters, but someone has to watch the oxygen (revenue)
• Too many PMs are "smart people who care about users and just showed up" without business model literacy
• Learn subscription vs. one-off, target market sizing, competitive dynamics — these are real skills
• "Product sense" is compressed experience; if you're early-career, focus on learning models, not intuition
• Start as an engineer or designer; get into small companies where you can poke into every corner
• If you can't navigate conflict, give hard feedback, or talk to strangers — reconsider the PM role