If you know nothing about OKRs, please start here: Art of the OKR. Or buy my book, Radical Focus. I think it’s rather good.
.@cwodtke … I keep reading varying interpretations of "results" in OKR. Ranging from tasks/steps to measures of reaching objective. Tips?
— John Cutler (@johncutlefish) July 18, 2016
An Objective is the goal you wish to achieve in a given time frame (typically one quarter.)
“Customers love us so much, they are our best sales team.”
Key Results are the metrics that change if we succeed. When creating Key Results, I look at the objective and see if there are words that could be quantified. In the example above, love becomes NPS and sales becomes referrals.
KR: NPS >7
KR: Referrals +25%
KR: “How did you hear of us” survey results: Friends and Family up 20%
Each of these key results answers the question, “if our customers were our sales team, what numbers would move?”
Sometimes when setting your OKRs, you come up with a key result first. For example, your CEO demands that revenue grow to 500K/mo. You need to ask yourself (or your CEO), what does that number tell us? We’re ready to raise a series B? Visitors are becoming customers? Users are screaming “take my money, please!” Every number has a story to tell, and that story is your objective. Once you have your objective, you can ask yourself if there are other good metrics to watch. Let’s say it was “ready to raise the next round.” You might choose retention, conversion or engagement numbers for the other KRs.
Having three key results is not required, but it is a good way to triangulate success. Retention balances revenue, so your team isn’t just squeezing a few bucks out of customers for a short-term lift. Let the objective and key results inform each other.
Next, you come up with hypotheses around how to move these numbers. This becomes your strategy for achieving the objective. It can be anything from revamping the customer service process to better self-service help.
Now you can come up with the projects/tasks you need to accomplish in order to make those key results move. I call them tactics. It’s critical to come up with as many tactics as possible. Quantity leads to quality. These become your pipeline. I prefer pipeline to roadmaps. It’s important to stay nimble in if you want to hit your key results.
Sequence your tactics.
Rate each one for:
- likelihood of working
The top of the list has low effort/high impact/likely. The bottom has shots in the dark. Then comb through the top for learning dependencies. i.e. “even though we’re not sure this will work, it is low effort and we’ll learn things that reduces the risk on this likely but high effort project.” Start with low effort/high learning. Treat each tactic they way you treat a lean startup experiment. Measure, measure, measure.
Goals are fixed, strategy evolves, tactics change often.
Now you have a stack ranked pipeline, and it replaces your roadmap. Each time you try a tactic, you measure your KRs. If they don’t move, reorder the pipeline.
You need this freedom so you can evolve your strategy for achieving your objective.
Why Can’t Project Completion be a KR?
If you make a Key Result a project, you are locked into it even if it doesn’t work. Let’s pretend your OKRs look like this.
“Customers love us so much, they are our sales team.”
KR: New Self-service help area
KR: Love-driven marketing place, with tv commercials
KR: Customer service completes sales training.
It is completely possible to achieve every single one of these and not have any number you care about move. Revenue could stay flat. Acquisition could go down. Retention could do a belly flop. Once your team is checking to do lists instead of watching metrics, you’ve institutionalized self-delusion. You are also micromanaging, which is the easiest way to drive away A-players.
OKRs Allow Companies to Scale
Marty Cagan used this wonderful quote in his forward to Radical Focus
“Don’t tell people how to do things, tell them what you need done and let them surprise you with their results.”
— General George Patton
Executives should set inspiring objectives with difficult key results, then trust their people to find a way to make those results a reality. As a leader, you can coach people, you can advise people, and you can review progress but you must not dictate tasks. Many new managers struggle to stop being product managers and become true leaders. As your team grows to 200 or 500, imagine deciding what every single person does with each day. It’s ridiculous.
You don’t scale. A goal-driven learning organization does.
Questions Turn Task into Results
When frontline teams set their OKRs, they will stick tactics in there. Engineers, designers and product managers are solutions people. If you spot a project in an OKR, ask a few questions.
- Why this project? Why is important?
- What will accomplish? What will change?
- How do you know if it’s successful?
- What numbers will move if it works?
- How does that tie into the company’s objective?
If you get an OKR from your reports that looks like this:
O: New Self-Serve Help Area
Kr: Better search
Kr: New FAQ
You can help it become this
O: The company helps our customers succeed when they are struggling.
KR: “Did this help” rating rises 15%
KR: Problem resolved rating on FAQ improves 30%
KR: Peer-to-peer help forum has DAU of 2K
If a result is a result, you can change your tactics until the numbers move. If a result is a project, you have locked yourself into achieving it, even if you find out it’s the wrong approach.
Success is not checking a box.
Success is having an impact.
If you complete all tasks and nothing ever gets better, that's not success.
— christina (@cwodtke) April 29, 2015