The whole notion of measuring adoption as a percentage of our codebase or similar has never sat right for me. Because of that it’s something I come back to time and again to work out why. Part of it has always been that if we’re measuring in percentages, typically that implies 100% is what we’re aiming for…but that can be a bad signal too.

If our products use nothing but the design system (so 100% composed of the DS), that says to me that we’re not solving any new problems or experimenting. So percentage wise what does good look like? How much is the right amount of the product to be using the DS? It varies all the time, so it’s not all that useful. If a team uses 100% of the available components at least once, that might be a positive signal but doesn’t tell us whether the other components they use should be in the design system or not, but it’s a useful metric to initiate a conversation.

Why do we care about adoption at all - that’s the more interesting question and one that can help much more. We might want to know how well used the system is by all products or teams. We might want to know where the issues are. These reasons behind wanting to know adoption can lead us to reframe the conversation. It’s not really about adoption but opportunity.

One side effect of a low percentage of adoption is that it can immediately look bad for the DS team because ‘numbers should be higher’. As opportunities it might be more strategic “team A could get real value from using x from the DS”. This goes to providing value over something that might feel like “you need to use the DS for…reasons…”.

Does it matter if not every component in a product is from your design system? No, typically it doesn’t. Built from smaller components and with tokens, it might be built of the system rather than from it. Solving one off, very specific problems should be part of how we work but with follow up conversations to work out when these tests or exploration might be brought into the system.

Adoption can feel like an uphill battle that we’ll not win and always be accountable for. Spotting and helping support opportunities uses much the same data but in a far more actionable and positive picture. It’s not about percentages, it’s about signals and how we interpret and use them - then it becomes a success story.