Picture a small decision most people make without thinking twice: meeting a friend for coffee on a Saturday morning. You search a few places, look at photos, check the hours, maybe glance at a review. Twenty seconds later, you have a plan.
Now picture the same decision with even one access need in the mix: a wheelchair, a service dog, a hearing aid, a tendency to feel overwhelmed by loud rooms, a parent pushing a stroller, a temporary injury. Suddenly the twenty-second decision becomes thirty minutes of detective work.
You look for the word "accessible." If it's there at all, it tells you almost nothing. You scrub through review photos hoping to spot a step at the entrance. You scroll the comments for the phrase "wheelchair." You call the business, and discover their website says one thing, the host says another, and reality is a third.
This is the everyday experience of finding an accessible place today.
"Accessible" is a label, not information
The single most common piece of accessibility data on the internet is a boolean: accessible, yes or no. It is also one of the least useful pieces of information ever published.
"Accessible" can mean any of these things:
- The entrance has no steps
- The entrance has one step, but staff will help
- There is a ramp, but it's too steep to use safely
- There is a side door that is sometimes unlocked
- There is an elevator, but it leads to a different floor than the restaurant
- Someone, at some point, told someone else it was accessible
None of these are the same place. None of them are the same experience. And none of them are something a software system can reason about.
The information is fragmented across places it shouldn't be
If you sit with this problem for a while, you notice that accessibility information is everywhere and nowhere at the same time:
- Mapping platforms know the location, hours, and menu, but expose access as a single yes/no flag.
- Review sites contain rich first-hand accounts buried in comment threads written years ago.
- Business websites publish claims that may not have been updated since the furniture was rearranged.
- Specialized accessibility apps and communities maintain richer data, but for a fraction of places.
The result is that finding the truth about a single venue takes opening four or five different tools and stitching the answers together yourself. We do not accept that for restaurant hours. We do not accept that for transit schedules. We accept it for accessibility because no one has built the shared layer yet.
"Accessibility" is not one thing
Even if every business in the world filled in a perfect "accessible: yes" field, most people would still be left out.
Accessibility is mobility, but it is also vision, hearing, sensory sensitivity, communication, cognitive load, lighting, noise, crowding, restrooms, signage, scent, temperature, queueing, and more. Each of these is the deciding factor for someone. A space with a perfectly accessible entrance may be unusable because the restroom is up a flight of stairs, or because the music is loud enough to make a conversation impossible, or because there is no way to order without speaking.
A useful accessibility data layer has to recognize that access is a multi-dimensional thing. A single field cannot describe it. A pair of fields cannot describe it.
And it goes stale, fast
The other thing about physical places: they change. Furniture moves. A ramp gets blocked by a delivery cart. A bathroom is "temporarily" out of order for six months. Construction reroutes the entrance through an alley with no signage. A new owner changes the policy on service animals.
Accessibility data is not a document. It is closer to a weather report. Treating it as a static label is most of why people don't trust it.
What would actually help
The honest answer is that we need structured, machine-readable, refreshable accessibility data, and we need it in enough places that asking the question becomes ordinary.
That is what we're building at Open Accessibility. Not a label. Not a badge. A shared data layer designed for the real complexity of how people actually use spaces.
The fact that this is still hard in 2026 is not a permanent feature of the world. It is a problem with a shape, and the shape is one we can fix.