I was in a planning session some time ago. I can't remember the exact thing we were building, but I do remember accessibility appeared there as one line in a ticket.
The form should be accessible.
That was it. Beautiful, really.
So much hope and optimism in those five words.
Of course everyone nodded, because what else could you do? Nobody wants to be the person who wants to spend the next 20 minutes discussing what that means.
So the work started.
Everyone was responsible for accessibility. Let's go!
A few weeks later, the ticket came back.
The errors were confusing. The required fields weren't clear. The order of information made sense visually, but not when you actually tried to complete the thing with a screen reader. Nobody was sure what should happen when someone skipped a section.
So we had to update the design. We had to rework most of the form elements and then of course QA had to retest it.
Weird. I thought the form was supposed to be accessible. It said so on the ticket and everyone was reponsible to make that happen.
No one really did anything outrageous. Everyone agreed accessibility mattered. Everyone played their part beautifully.
But nobody played the part of deciding what "accessible" meant for that specific form before the work moved forward. Not in principle, but as a decision.
What must be true before we call this done?
I think about those five words a lot.
The form should be accessible.
It sounded so responsible.
But sometimes vague responsible language is where rework starts.
Because if the requirement is too vague to guide a decision, it's probably too vague to protect the outcome.