OKRs (Objectives & Key Results) are a framework for validating alignment through the organization. As a company, as a department, as a team: what are we focused on this quarter? What are we trying to make true?
Alignment looks nice when the team OKRs roll up to the department OKRs, and those add up to the company OKRs. This is logical. And terrible.
Occasionally Summation Makes Sense
In the Sales department, you have some revenue objective. You might divide that up between regions.
From there, a quota breaks down to individual reps.
Sales is the easiest department to measure this way; it is closest to the money. As the quarter shakes out, some teams will be under target, and some will hopefully be over.
Summation is the Worst
Consider pie.
This is ridiculous. The company is a complex system. Each department supplies capabilities to others, or creates conditions that let other departments succeed. We are not disconnected individuals, and our actions do not add to outcomes. Our actions create outcomes in interdependent ways.
Sales reps bring in revenue, yes. For that to happen, Marketing has to make a world where people know about our product and gather email addresses of interested people. Engineering has to create and operate a product that people want. Finance and People Ops have to make it legal and pleasant for any of us to work here at all.
Enabling is Better than Adding
When making OKRs for your department or team, look at the higher level ones. Ask: “do we have a role in making this happen?” Sometimes it’s clear enough.
Cascading Breaks Down for Enabling Teams
How is People Ops supposed to help with a new product? They don’t touch it at all. They enable every other department.
Sometimes it works to ask, “For this to happen, what must be true?”
Inside People Ops, Recruiting and Internal Education can make sure the company has the knowledge we need.
The infrastructure team can state its intention to prioritize reliability of the new feature. SREs can improve the feature’s observability, because for engineering to support and improve AI search, they need to see what is happening.
It isn’t always this clear. Maybe the infrastructure and finance projections are in place, and any issues that come up fall under normal work for this team. Does every team have a role in every company objective every quarter? No. Don’t pretend we do.
Catch-all Objectives have a Catch
Some teams are stuck creating OKRs in a rigid system that requires selecting a parent objective. This leads to department and company OKRs that are basically “We do our business.”
This solves the mechanical problem of ‘How do I pretend that my objective cascades upward?’
One point of OKRs is to say “not yet” to the other improvements we could be working on. The catch-all OKR ruins it. Focus is gone.
OKRs Cascade Upward to Strategy, not OKRs.
Only one team in engineering is focused on the new AI Search feature. Everyone else is developing and iterating other features, or increasing scaling capacity, or whatever they do. These teams can express their objective for the quarter in OKRs
OKRs are only this quarter, after all. The company has a strategy that’s wider than this. Next quarter, we might need to feed the most important fields to the AI, so a team is making that possible now. Overall our strategy of growing into a bigger market requires vigilance on costs, and some teams working on that, any given quarter.
The questions of “What role does my team have?” and “What has to be true for us to make this happen?” apply to company strategy, not just this quarter’s focus.
For each of department and team objective, can you tell a story about how this feeds our ability to realize our strategy? Yes: that’s alignment.