011 min
Why designing for the average person leaves people out
In 2014 an engineer at Microsoft noticed, on Twitter, a custom controller built by Warfighter Engaged, a nonprofit that supplies gaming devices to injured veterans. The players it served had trouble with the standard Xbox controller. In 2015 a team in the Xbox division began its own prototype, and that work became the Xbox Adaptive Controller, which went on sale in September 2018.
The veterans were not the problem. A standard controller is built around one kind of hand, grip and reach, so a player whose body works differently meets a poor fit the moment they pick it up. Designers create poor fits like this without meaning to. Microsoft's inclusive design guide explains why. People judge their ideas by what they can do themselves, so they make things that are easy for some users and hard for everyone else.
The guide also says there is no such thing as a normal person to design for. Each person's sight, hearing, speech and touch change with age, health and place. A design that assumes all of them work at full strength all the time will fail someone, and it will fail each of us at some point.
021 min
What a mismatch means in inclusive design
Inclusive design defines disability as a mismatch between what a person needs and how a product, system or service is designed. The Inclusive Design Research Centre (IDRC), a research group at OCAD University in Toronto, uses this definition. It moves disability out of the person and into the design. A design can disable anyone it does not fit, and the same person can be a good fit for one design and a poor fit for another.
Microsoft's guide takes the same position. It notes that the World Health Organization revised its definition of disability in 2001 to describe an interaction between a person and the society around them. The guide counts temporary and situational limits as part of the picture, not only lasting ones.
A lecture that mismatches four students
The Centre argues that anyone a design excludes can experience disability, and it gives an audio-only lecture as its example. In that lecture a blind student is less disabled than a classmate who has not done the reading, a classmate who is less fluent in the language, or a classmate who has been awake all night. The lecture is built for a student who knows the background, understands the language well and can give it full attention. Every student who differs from that one is mismatched, whether or not anyone calls them disabled.
031 min
Where inclusive design came from
Inclusive design grew out of earlier work on removing barriers. From the 1950s, Europe, Japan and the United States moved toward barrier-free design, which took obstacles out of buildings for people with physical disabilities. In 1974 a magazine article called The Handicapped Majority argued that poorly designed products and spaces hurt ordinary users too, so disability was not a niche concern.
The term came later. Roger Coleman's 1994 presentation The Case for Inclusive Design, given at a congress of the International Ergonomics Association, is the first published use of it. Much of the early interest in the United Kingdom came from people living longer. A Cambridge team that included Coleman and John Clarkson built the free Inclusive Design Toolkit, and in 2005 the British Standards Institution published a guide to managing inclusive design.
In Toronto, Jutta Treviranus had been running a research group on technology for people with disabilities since the early 1990s. It became the Inclusive Design Research Centre. Digital products suited its approach, because software can change for each person in a way a building cannot. Design justice, a related movement, asks designers to begin with the people that design has historically left out.
Microsoft then brought the approach to product teams. It published an Inclusive Design toolkit and reduced the method to three short principles, which the next section sets beside the Toronto version.
042 min
The principles of inclusive design
Microsoft's toolkit states three principles. The first, recognize exclusion, asks teams to look for places where a design and a person do not fit. The second, learn from diversity, says excluded people have insight the team lacks, so they should take part from the start. The third, solve for one, extend to many, says a design made for a person with a permanent disability often helps many other people.
The IDRC names three dimensions instead. The first is to recognize, respect and design with human uniqueness and variability. The second is an open process that includes co-design with people who see the problem from different positions, including people who cannot use the current design. The third is to design with awareness that every design sits inside larger systems, so a change in one place affects people elsewhere.
Both sets lead to the same habits. Bring excluded people in early, expect more than one way to use a thing, and check what a change does downstream. Nielsen Norman Group (NN/g), a UX research firm, states the aim as meeting as many needs as possible, which is a different target from reaching as many users as possible.
Why the edge comes first
Treviranus explains the second dimension with a picture of a population. Plot what people need, and the result looks like a starburst, with about 80 percent of the dots in a tight centre and the rest spread far outward. The centre holds most of the people but little of the variety. Industry designs for the centre first because it is the cheapest place to start, and plans to serve the rest later, if at all.
She argues for the opposite order. The people who have the most trouble with a current design are the ones a team most needs to include, because they show what nobody has explored. She calls her process the virtuous tornado: a loop that starts with people who struggle with the present design and adds more people in each round. Her business argument is that a design shaped around many kinds of people can change more easily, so it needs fewer repairs when needs shift.
052 min
The persona spectrum and the curb-cut effect
Microsoft's persona spectrum is a tool for finding the other people who share one limit. It sorts a limit into three kinds. A permanent limit is long term, such as the loss of an arm. A temporary limit lasts a short time, such as a wrist injury or looking into a bright light. A situational limit comes from the setting, such as holding an infant or driving a car.
| Limit | Permanent | Temporary | Situational |
|---|---|---|---|
| Using one hand | Loss of an arm | A wrist injury | Holding an infant |
| Seeing | Blindness | Looking into a bright light | Driving a car |
The guide's own example makes the point. A device built for one person with one arm can also serve a person with an injured wrist and a parent who is carrying a baby. Microsoft's guide counts 26,000 people a year in the United States who lose an arm or a hand, and says the number passes 20 million once temporary and situational cases are included.
The same pattern holds for captions, which were made for people who are hard of hearing and now help people in a crowded airport and children learning to read. High-contrast screen settings began as help for low vision and now help people who use a device in bright sunlight. Angela Glover Blackwell named the general pattern in 2017: the curb-cut effect. She meant that laws and programmes designed for people who are left out often end up helping everyone.
The first curb cut in Berkeley
In the early 1970s, disabled advocates in Berkeley, California, poured cement against a curb to make a rough ramp for wheelchairs. Pressed by disabled activists, Berkeley installed its first official curb cut in 1972, on Telegraph Avenue. After curb cuts spread, parents with strollers, workers with carts and travellers with luggage used them too. Blackwell cites a study at a shopping mall in Sarasota, Florida, where nine out of ten pedestrians with nothing to carry still went out of their way to use a curb cut.
062 min
Inclusive design in real products
The Xbox Adaptive Controller shows the method in use. From 2017 Microsoft built the product with accessory makers and nonprofits that included SpecialEffect, Warfighter Engaged and the AbleGamers Foundation, and it also worked with a disabled gamer, Randy Fitzgerald. The controller does not ask a player to use one particular input. It is a hub with many small jacks on the back, so a player can plug in the switches, buttons or joysticks they already use. One product fits many bodies because each person chooses the inputs.
When an open design still costs money
The hub itself went on sale at US$99.99, and one of its creators stressed that a low price mattered. A critic, Grant Stoner, wrote in 2020 that the controller on its own was not too expensive, but that the switches and joysticks a player must add made a working setup costly. The case shows an edge of the method. A product can lower one barrier and leave another standing, so a team that checks its work should look at the full cost of use and not only the part it sells.
Smaller moves on the web
Everyday forms show the same thinking. NN/g points to forms that ask for a person's race with single-choice buttons. Checkboxes, or an option for two or more races, fit people who do not belong in one box. NN/g also describes last-name fields that reject hyphens, apostrophes or accented letters, which shuts out people whose real surnames use them. Facebook built a support page for people whose authentic names its sign-up refused.
For older users, NN/g reports that across three rounds of research the most common complaint from older users was that sites and apps are hard to read. It recommends large type, strong contrast, a control that enlarges text, and an optional dark mode for people with cataracts. Pictures and wording are design choices as well. Whom they show, and whom they leave out, tells people whether a product was made for them.
Machine learning adds a new risk. Treviranus notes that recommendation systems suggest what people similar to the user have already chosen, so the range of choices narrows around what is popular. That repeats the average-user mistake at a very large scale.
071 min
How inclusive design estimates who a design excludes
Inclusive design can estimate how many people a product excludes. The method comes from the Inclusive Design Toolkit at the University of Cambridge, and it is called the exclusion calculator.
A product places demands on the people who use it. A product excludes a person when any one of the demands it places on them is higher than what they can do. A product with very small text demands a lot of vision, so people with age-related long sight are shut out of it.
The toolkit sorts capability into five groups: vision, hearing, thinking, reach and dexterity, and mobility. A designer rates the demand a product makes on each group from low to high. The ratings are then combined with survey data on how many people have each level of capability, which gives an estimate of the share of the population left out.
Two limits matter. The survey data in the toolkit describes adults in Great Britain, drawn from a government survey in 1996 and 1997, and the toolkit itself calls it the best coherent source until newer data exists. The calculator also counts capability only. It does not show whether people want the product, can afford it or trust it.
082 min
Applying inclusive design
Picture a team that runs a ticket-booking site. Support staff keep hearing from people who give up at the payment step. The team starts by asking who the payment step excludes, and why. These moves turn that question into work.
Name the mismatch.
Write down who cannot finish the task and what the design assumes about them, such as steady hands, fast reading or one fixed name format.
Bring in the edge.
Recruit people who struggle with the current version, and keep them in the work from research to testing, not only at the final review.
Use the persona spectrum.
For each limit you find, list the temporary and situational cases that share it, to see how many people one fix will reach. Then check the fix with the people who live with the permanent case.
Offer choices, not one fixed setting.
Text size, an optional dark mode, captions with a transcript, and more than one way to enter a name or a payment.
Mix the team and the test group.
A team with different backgrounds, and a test group that matches your real users, notice different problems. Keep collecting feedback after launch.
Treat standards as the minimum.
Meeting the Web Content Accessibility Guidelines (WCAG) is the minimum, as the accessibility article explains. Add plain language and consistent layouts on top. Run usability testing with real people, because an automated scan cannot tell you whether someone finished the task.
Organisations also have outside reasons to act. Laws such as the Americans with Disabilities Act of 1990 and the UK Equality Act 2010 prohibit discrimination, and a design can create it without anyone intending it. Ask a qualified lawyer about a specific product. The business case is usually a larger set of customers, and Treviranus adds that a flexible design costs less to change later.
092 min
Inclusive design vs. accessibility and universal design
Inclusive design is often mixed up with two neighbours, and the three differ in what each one names.
| Inclusive design | Accessibility | Universal design | |
|---|---|---|---|
| What it is | A method that starts from human variety and involves excluded people | A quality: how well people with disabilities can use something | A single design that nearly everyone can use without adaptation |
| Scope | Ability, language, culture, gender, age and more | Disability | Everyone, in one form |
| Usual setting | Mostly digital, where an interface can change cheaply | Digital and physical | Mostly buildings and objects |
Microsoft's guide puts the difference with accessibility in one line: accessibility is an attribute of a product, and inclusive design is a method for getting there. A team can follow the method and still miss a standard, so the two work together.
NN/g separates inclusive design from universal design by number. Universal design aims for one experience that nearly everyone can use, while inclusive design accepts several versions as long as each reaches the goal. NN/g adds that this is why universal design appears more in buildings and objects, where making several versions is costly. The IDRC describes its own approach as universal design with provisos.
Two definitions that disagree
The British Standard of 2005 defines inclusive design as the design of mainstream products and services that as many people as reasonably possible can use, without special adaptation. The IDRC takes the other side on this point. It wants a system that can adapt to each person, which it calls one-size-fits-one. Cambridge's own toolkit adds that one product cannot always fit a whole population, and it suggests a family of related products. So "without adaptation" does not mean "one product". When someone uses the term, ask who is expected to do the adapting, the designer in advance or the system for each user.
102 min
Where inclusive design is contested
Inclusive design has critics inside the field, and their points change how a team should use it.
Temporary limits are not permanent ones
Frank Spillers, a UX consultant, argues that leaning on temporary and situational examples can hide what life with a permanent disability involves. A person who is distracted while driving does not know blindness, he says, and a person with an injury usually expects to recover. His remedy is lived experience: work with disabled people through the whole project and test with them regularly. The spectrum is useful for finding who else benefits, but it cannot replace talking to the person at its permanent end.
Simulating a disability
Some teams try to build empathy by wearing blindfolds or earplugs. Microsoft's own guide calls it misleading to rely only on that kind of simulation, and it says real understanding comes from learning how people adapt and from seeing the world from their view. Spillers goes further and says research finds that disability simulations do not work.
When access needs conflict
Sometimes a feature that helps one group makes things harder for another. A 2026 preprint by Xinru Tang and a co-author followed a nonprofit that grew from a small group for blind athletes into an organisation for people with many kinds of disability. Making its communication accessible meant balancing needs that sometimes conflicted. The authors treat such conflicts as more than a technical problem, because they show who holds power and who is accountable. Inclusive design has no rule that settles a conflict in advance. A practical working rule is to list who a choice helps and who it blocks, to ask the blocked people what they would change, and to offer a setting where one fixed answer cannot work.
?5 questions
Questions people ask
Is inclusive design only about disability?
Does an inclusive product have to work for everyone?
Is inclusive design required by law?
Who should take part in the research?
Where can a small team start?
§12 sources
Sources
Wikipedia contributors, "Inclusive design".
Inclusive Design Research Centre, OCAD University, "What is inclusive design".
Microsoft Design, "Inclusive 101 Guidebook".
Alita Kendrick, "Inclusive Design", Nielsen Norman Group, 30 January 2022.
Show all 12 sourcesShow fewer sources
University of Cambridge Engineering Design Centre, Inclusive Design Toolkit, "What is inclusive design?".
University of Cambridge Engineering Design Centre, Inclusive Design Toolkit, "Assessing demand and exclusion".
dscout, "Design in Three Dimensions", an interview with Jutta Treviranus.
Angela Glover Blackwell, "The Curb-Cut Effect", Stanford Social Innovation Review, Winter 2017.
A summary by Wikipedia contributors, "Xbox Adaptive Controller".
Frank Spillers, "The problem with Inclusive Design and Temporary Disabilities".
Xinru Tang and a co-author, "Designing for Collective Access: In Search of a Solution to Accessible Communication in a Mixed-Ability Non-Profit", arXiv:2605.10085.
Sensory Trust, "What is inclusive design".
