In professional conversations, the first signals that emerge about digital accessibility are often those that the eye picks up on: low contrast, a button that's too discreet, text that's too dense. These elements are important, of course, but they represent only a tiny fraction of the subject. So why this single scope?
Many users surf the Web by vision and spontaneously project this experience as their only reference. This prism - or bias - directs our attention towards what we can see and makes us forget everything that we cannot see. However, a large part of accessibility lies elsewhere: beneath the graphic surface lies a whole decisive territory.
This bias is not insignificant. It directly influences design decisions, project priorities and team trade-offs. What is not perceived is neither corrected, nor even identified as a problem.
The same site, radically different experiences depending on how it's used
The spontaneous assessment of a site is therefore often based on a visual impression: harmony of colours, rhythm of spacing, overall balance. In this context, a site can seem accomplished, coherent and sometimes even exemplary. But this feeling is just one of many.
These disparities are not anecdotal. They create invisible situations of exclusion for the teams that design and run the systems. A service may be perceived as functional internally while being unusable for a portion of users.
A design bias, not a lack of skill
It would be easy to see this as a lack of skills, but it's more of a prism. Historically and culturally, the Web has been shaped by vision-centred approaches. (and the mouse).
Training, tools and professional reflexes still encourage this perspective: scan, click, move fast. Nothing really prepares you to read a page entirely from the keyboard, to interpret a structure with a screen reader or to understand what an interface becomes when interactions are based on other capabilities.
This perceptual bias limits the way we imagine the experience, not out of ill will, but because another reality remains unexplored. It explains why many devices pass internal validations while remaining partially inaccessible in real-world situations.
Why does this bias tend to persist in organisations
Several structural factors explain the persistence of this bias in digital projects.
- The absence of confrontation with real-world usage is a first determining factor. Interfaces are designed, tested, and validated without direct exposure to varied usage situations. Teams project their own browsing habits without perceiving the experiential differences that other users may encounter.
- Internal homogeneous validation reinforces this phenomenon. Tests are often carried out by similar profiles, sharing the same frames of reference and interaction capabilities. This homogeneous framework reproduces a unique prism and limits the detection of obstacles.
- Positioning accessibility as a technical constraint is another key factor. When addressed downstream, as a compliance or remediation step, it doesn't challenge initial design choices. Interfaces are then adjusted, but rarely rethought.
These mechanisms are not due to a lack of expertise. They reflect an organisation of work which does not permit a full understanding of the diversity of uses.
Digital accessibility It doesn't start with technical fixes. It starts with an awareness: that of the limits of one's own perspective. As long as organisations design solely from their own usage, some obstacles will remain invisible. And what remains invisible can neither be corrected nor managed.