I wanted to follow up on the article about accessibility being a system-level problem.
There is this tension at the heart of accessibility work. Once you've worked in accessibility long enough, you'll see it.
You need system-level support to make lasting change, but you can't wait for permission to start. So you start building stuff from the ground up until you've got something real to show. Then you carefully, very carefully, translate that into organisational backing without losing what made it work in the first place.
And that's the balancing act.
If you push too hard for top-down strategy before you've proven the value, you'll get empty policies. If you stay too firmly at the unit level, you'll burn out fixing the same problems over and over. What you want is to build enough momentum to pull the organisation along with you, without letting them turn your work into PowerPoint.
This balancing act usually start with the most common silly things.
You've fixed that wonky button, to continue with the example from the original piece. You've built the component. And that component has helped three other teams. Now leadership wants to "take accessibility seriously."
Brilliant. Also, dangerous.
Because here's what happens next. They want a strategy. A roadmap. Quarterly objectives. A steering committee. And suddenly you're in meetings about meetings while the actual work stops.
Don't let them turn your momentum into bureaucracy.
When you pitch leadership, lead with what you've already shipped. Show the accessible components teams are using. Share the support tickets that dropped. Point to the features that now work for everyone. Make it tangible, not theoretical.
Ask for what amplifies the work. That's usually dedicated time for the people to do it. That's budget for training. That's authority to make accessibility a release requirement.
What you don't need is a new governance framework.
If they want metrics, give them the ones that matter. How many teams are trained in accessibility. How many releases include accessibility testing. Not how many documents exist or committees meet.
Your goal isn't to build an accessibility empire. Take it down a notch or two. Make it as simple as "stop considering accessibility as a separate thing."
Stay close to the implementation. The moment you're too far from the unit-level, you've lost.