Is Your Key Result Secretly a Task? A 12-Question Quiz

A sheet full of verbs, each with a date

Open any team's OKR sheet in week two of the quarter and count the verbs. Launch. Hold. Build. Analyze. Consult. Each line has a date, so it looks measurable, and each will be ticked off by quarter end. Then someone asks what changed for customers, and the room goes quiet.

That is the most common fault in OKR sets: a task written in key-result clothing. Google's published guidance draws the line in one sentence: key results “must describe outcomes, not activities” ¹. The quiz below gives you twelve invented key results, mostly drafts of the kind teams write. For each, you decide whether it is a real result, a task, a KPI in disguise or a line nobody can grade. Every answer explains the rule, so you leave with four questions you can put to any OKR set.

A task proves the work happened. A key result proves the work changed something.

The four ways a key result goes wrong

The first three come straight from Google's guidance: activity words mark tasks ¹, a goal that requires no change to anyone's activities is business as usual ¹, and a key result must allow a grade to be assigned objectively at the end of the quarter ¹. The fourth is Google's insufficient-key-results trap ¹. Andy Grove's version, from a 1970s teaching clip shown in John Doerr's TED talk, is the shortest: at the end “you can look and without any argument say, ‘Did I do that, or did I not do that?’” ².

Why tasks keep winning

Teams do not write tasks because they misunderstand OKRs. They write them because tasks are safe. A task is fully in the team's control; a result depends on customers, other teams and luck. Under any pressure to score well, people drift toward the line they can guarantee. Google's guidance names the same drift for whole OKR sets, which tend to be written around what a team “believes it can achieve without changing anything they're currently doing” ¹.

The fix is not to ban tasks. It is to attach each one to the result it is for. Google's own example: “Launch Foo 4.1” is weak, “Launch Foo 4.1 to improve sign-ups by 25 percent” is better, and simpler still is “Improve sign-ups by 25 percent” ¹. The launch moves into the plan, where it belongs.

Reading your score

The bands are our rubric, not a validated test; a twelve-question quiz is a practice set, not an exam. What matters more than the score is the pattern of misses. If you missed tasks, slow down on verbs. If you missed the disguised KPI, ask for the baseline before you judge any target. If you missed the sufficiency question, you are grading lines one at a time; read the set against its objective instead.

When a task is the right key result

The honest objection: sometimes the deliverable is the result. A regulatory filing, a plant certification, a contract signed. Nobody needs a downstream metric to know those mattered. Google's guidance agrees in its own way: among its examples of committed OKRs is delivering a defined feature or improvement by a set date ¹.

So the test is not “is there a verb?” It is “if this scores 1.0, does anyone outside the team notice?” A certification that unlocks a market passes. A portal nobody uses does not. Keep task-shaped key results for deliverables whose value is not in doubt, and make everyone else say what the work is for.

Where to go next

Draft your own set in the OKR writer, which flags activity words and missing numbers as you type. For the line between key results and KPIs, read OKR vs. KPI. Then take one question into your next review, the one that turns most tasks into results on the spot: what is this work for?

Questions

What is the difference between a key result and a task?

A task describes work the team will do, such as launching, building or analyzing. A key result describes what that work should change, with a number and a date, so it can be graded without argument. Google's guidance says key results must describe outcomes, not activities.

Can a launch ever be a key result?

Yes, when the deliverable's value is not in doubt, such as a certification that unlocks a market. Otherwise attach the launch to its purpose: Google's guidance prefers “Launch Foo 4.1 to improve sign-ups by 25 percent,” or simply the sign-up target.

What are good key results examples?

Good key results name a measure of what customers or the business do, a baseline, a target and a date: the share of new customers who finish setup in week one, from 40% to 65% by September 30. Add a guardrail where the cheapest way to hit the number would do damage.

Where to go next

• OKR Writer and Quality Checker
• OKR vs. KPI: Which Numbers to Hold Steady and Which to Move
• KPI Examples That Track Strategy, Not Activity
• OKR (Objectives and Key Results)
• OKRs (Objectives & Key Results)