Cultural transformation in any organization begins and ends with the numbers. That’s part of why it can be so difficult to build and maintain a useful design system.

If you’ve already convinced your boss that a design system can help hit her budget and metrics (more on how to do that here), you’re ready to hang some numbers on your project.

But you’re not looking for just any numbers. They need to tell the story of how your design system is having an impact on company objectives. In other words, you need to make it crystal clear how people who use the design system are succeeding.

But how on earth do you start to measure the impact of a design system?

Success = Measuring progress. Not results.

The first thing to understand about collecting data is that we’re measuring progress. Not targets. Not results. Not performance.

We’re measuring progress.

We like to use OKRs, which are Objectives and Key Results. The concept of OKRs was originally implemented at Intel in the 1970s and is now widely used at companies like Google, LinkedIn, and Twitter.

OKRs should answer the question: How do we know we’re making progress?

Which means you get to unlearn everything the education system ever taught you about success and failure about your final grades.

Embracing 60%: A new definition of success

Think for a minute about your worst high school subject. Maybe you got back a test with “60%” in big, red letters. And an “F” beside the score. Circled multiple times. Just to add a little extra pain.

But when you’re setting and measuring effective OKRs on a scale of 0 to 100, 60% gets you inside the success zone!

Remember, we’re measuring progress. So if the cycle you’re scoring — usually a quarter at a time — gets you more than halfway to the number you want to hit, you get to celebrate.

Here’s the thing. If you score anything above an 80, you’ve probably sandbagged — and set an OKR that wasn’t ambitious enough.

If your score is something like zero — or anything under 60, you probably have been way too ambitious.

The tricky part about OKRs: Often the measurements are less about tracking what got done, and more about forecasting what is possible in any given timeframe.

So brace yourself. The first couple of times you do OKRs you’re going to shoot for “bad.”

Getting started with OKR

The cornerstone of any OKR is storytelling around progress, not results — and you want to share your progress stories. As you go.

So just as important as setting benchmarks is the process by which you review and update.

It’s not enough to set good OKRs. You make them matter when you build a practice of tracking, so add the following one-hour events to your calendar:

Every Friday:

Look at the data for all of your key results to record what’s changed.

Every Monday:

Write an update to your fellow design team members. Progress — or the lack of it — must be visible. Remember to spend more time on key results, not objectives. Especially on what each team member can work on during the week to influence the next week’s key results in the right direction.

The last Friday of each month:

Draft a high-level message to all the people you initially interviewed about the first things to focus on as you build a design system and report what you’ve learned in the previous four weeks. Focus more on objectives than key results—the opposite weighting of the message you send to the team each week.

You can automate your review. Don’t. At least not yet.

For at least the first couple of cycles, do the review work manually so that you are intimately familiar with the data, how it’s being collected, and how much it can be influenced.

You will be really bad when you build your first OKR. The second time you build an OKR, shoot for “less bad.” So the sooner you get started, the sooner you will improve. Be kind to yourself.