Making expert accessibility knowledge usable by non-experts
A report alone does not make a space more accessible. I designed and built a product that helps people inside an organisation take part in accessibility work and act on findings without becoming accessibility experts.
- Role
- Sole designer and developer, with two client-side accessibility experts as domain partners
- Context
- Built for a two-person accessibility consultancy after helping secure an Access Activator grant
- Project Type
- Research, information architecture, product design, production code
- Timeframe
- 8 weeks, 2026
- Status
- Launched July 2026. Now in pilot.
- User Research
- Information Architecture
- Content Design
- Product Design
- Development
The report was a deliverable, not a capability
An accessibility audit combines standards with expert judgement about how people actually use a space. The work is difficult to scale across an organisation with many sites, but producing more reports was only part of the problem: once a dense, bespoke report arrived, momentum often stopped.
The product needed to help people inside the organisation take part in accessibility work and act on findings without requiring site managers to become accessibility experts. I worked with two accessibility specialists to turn their knowledge into questions people could answer on site and findings programme leaders could prioritise across a portfolio.
Testing changed the organising model
I developed two concepts through three rounds of moderated testing with seven participants: accessibility assessors, site managers, programme managers and people without domain expertise. A later session tested the questions in a real building rather than over a screen.
The first audit followed the way accessibility standards are filed, grouping questions by category. An experienced assessor exposed the practical failure: answering every lighting question meant walking the whole building, then repeating the route for doors and every other category.
Before · organised by category
After · organised by physical journey
I rebuilt capture around physical journeys: car park to front door, entrance to service desk. A site manager could stand in one place and answer everything relevant there. Reporting retained the category view because the two users needed different ways into the same knowledge.
The structure that is right for maintaining expert content is rarely the structure that is right for using it.
Research changed the rules, not just the screens
The domain knowledge began as thousands of lines across reports and reference material. I built a custom pipeline to structure it into several hundred tasks and generate the first question set.
I then wrote evaluation questions whose answers became higher-level generation rules. I tested the generated content with users and fed their findings back into those rules before regenerating the dataset. Because routes referenced a shared knowledge base, changing the audit model was a reorganisation rather than a complete rewrite.
Working software in eight weeks
I delivered a multi-tenant product with mobile capture and programme-level reporting, then handed the production code and infrastructure to the client. It launched at the Access Activator showcase in July 2026 and is now in pilot.
The project was selected as one of six from 160 grant applications. That is a selection outcome, not proof of behaviour change. The product is instrumented against a 40% audit-time reduction target; that result is still pending.