finance
OKR (Objectives and Key Results)
Goal framework: one ambitious objective paired with 3-5 measurable key results.
Definition
OKRs (Objectives and Key Results) were created at Intel and popularized by Google. Structure: one qualitative Objective (the 'what'), 3-5 quantitative Key Results (how progress is measured). OKRs are aggressive - hitting 70% is considered success. They run quarterly, not annually. The discipline forces teams to define what success looks like in numbers, not vibes. Common mistake: writing tasks as KRs ('launch new website') instead of outcomes ('achieve 5% conversion rate on new website').
Writing OKRs that actually drive change
The hardest part of OKRs is writing them well. A good Objective is qualitative, ambitious, time-bound, and tied to a strategic priority: 'Become the default fractional CFO for US Shopify brands at 1 to 5M ARR'. A bad Objective is a task list: 'Improve marketing'. Key Results must be measurable outcomes, not activities. 'Publish 12 blog posts' is an output and belongs nowhere; 'Grow organic leads from 50 to 200 per month' is an outcome and belongs as a KR. A useful test: if hitting the KR would not move the Objective, it is the wrong KR. Most teams write 8 to 15 KRs per quarter and produce no real change; 3 to 5 sharp KRs per Objective with one clear owner each is the discipline.
The 70 percent rule and why it matters
Google's original OKR doctrine sets a 70 percent achievement target. Hit 100 percent on every KR, your OKRs were too easy. Hit below 50 percent, they were unrealistic. The 70 percent zone is where genuine stretching happens. US tech companies (Google, LinkedIn, Stripe, Asana) hold this line; many traditional businesses misuse OKRs as performance management and tie compensation to 100 percent achievement, which collapses the system. If commissions or bonuses depend on OKR achievement, employees write soft OKRs they can sandbag. OKRs are aspirational goal-setting; performance reviews and compensation should run on separate metrics like KPIs and individual contributions.
OKR cadence and tooling
Quarterly OKRs are the standard. Set them in week 0 of the quarter, score them in week 13. Weekly check-ins through the quarter ask three questions per KR: what progress was made, what is blocking, what is the next action. Tooling options for US teams: dedicated OKR platforms (Lattice, Workboard, Gtmhub, 15Five, Quantive), built-in OKR features inside HR platforms (Lattice, Culture Amp), or just a structured Notion or Asana template. The tooling choice matters less than the cadence. Many companies bought Workboard, never reviewed weekly, and abandoned OKRs in 6 months. A weekly 30-minute OKR review with the leadership team is non-negotiable; the tool is a logging surface, not the program.
When OKRs fail and why
OKRs fail for four reasons. First, too many OKRs (more than 3 per team, more than 5 company-wide): attention collapses. Second, KRs that are tasks not outcomes ('launch website' instead of 'increase site conversion to 4 percent'). Third, no weekly cadence: OKRs set in January, forgotten in February, scored in March. Fourth, tying compensation directly to OKR achievement, which incentivizes sandbagging. The fifth and most common failure is that OKRs become parallel to actual work instead of representing the most important work. The fix: every leadership meeting, the first agenda item is OKR progress. If you cannot connect this week's work to an OKR, either the work or the OKR is wrong.
FAQ
OKRs versus KPIs - when do I use which?
Use KPIs for ongoing health monitoring (MRR, churn, gross margin, NPS) and OKRs for time-bounded change initiatives (launch new tier, expand into new segment, cut payback period). A mature company runs both: KPIs report business health weekly, OKRs drive the strategic agenda quarterly. KPIs are 'how are we doing'; OKRs are 'what are we changing'. If a metric appears on both your KPI dashboard and your OKR list, it is probably mis-categorized in one place.
Should I cascade OKRs down through the org?
Loosely, not strictly. Top-down cascading (CEO sets OKRs, each VP sets sub-OKRs that ladder up, each manager sets sub-sub-OKRs) produces alignment but kills initiative. Modern practice (LinkedIn, Asana, Stripe) is hybrid: leadership sets 3 to 5 company OKRs, then each team writes its own OKRs that connect to one or more company OKRs but are not direct subdivisions. This preserves bottom-up creativity. Cascade no more than 2 levels: company and team. Individual OKRs almost always devolve into task lists.
How long do OKRs take to implement properly?
First quarter: clumsy. Most teams write task-list KRs, miss weekly check-ins, and score chaotically. Second quarter: rewriting habits with better discipline. By quarter four, the team writes outcome-driven KRs natively and the weekly cadence is muscle memory. Plan for 12 months of practice before declaring OKRs working. The mistake is cancelling OKRs after one bad quarter. The mistake is also expecting them to be the answer in week one.
Can solo founders use OKRs?
Yes, with one twist. For solopreneurs, OKRs become a personal strategic focus tool: 3 Objectives per quarter, 2 to 3 KRs each, reviewed weekly with yourself or a peer accountability partner. Without a team to align, the value shifts from coordination to focus. The discipline of writing measurable KRs forces a founder to confront whether they are working on the things that move the business or just things that feel productive. Use Notion, Things, or a Google Doc; tooling is irrelevant at one person.
Do OKRs work in service businesses, or only in tech?
They work in service businesses, with adjustments. The tech-style 'moonshot OKR' culture does not translate to consulting or accounting firms where deliverables are contracted with clients. But the discipline of writing 3 to 5 Objectives per quarter with measurable KRs absolutely applies. For US service firms, common OKR Objectives include landing new customer segments, building productized offerings, raising average deal size, and reducing partner dependency. Run them as quarterly strategic focus exercises, not as performance management.
In your business
- →Quarterly OKRs, not annual
- →Key Results must be measurable - outcome, not output
- →Don't cascade more than 2 levels - founder/leadership OKRs and team OKRs is enough