Testing With Somebody Who Needs It

What a real session finds

Checklists find the things somebody thought to put on a list.

Somebody who uses assistive technology every day finds the things nobody thought of, in the first ten minutes, usually in the order of how much they matter.

They also tell you which of your known problems actually matter and which are theoretical, which is information no audit provides.

For a small business one session, done once, is enough to reset how the site gets built from then on.

How to find somebody

• A local disability organisation.
• A remote testing service.
• A vision or hearing charity with a tech programme.
• A local college's accessibility office.
• Somebody you already know.

The first is the most reliable route for a small business and the least used. Many organisations maintain contact with people who do this work and will pass on a request.

Pay properly

Which is the part people get wrong and it matters.

This is expert consultation, not a favour, and the going rate for user testing applies.

Somebody who uses a screen reader daily has expertise most developers do not, acquired the hard way.

Asking for it unpaid, or offering a voucher, is the mistake that makes people decline and tell others.

An hour at a professional rate is a small cost against what it finds, and against what a written audit would have cost.

Say the rate when you make the request rather than afterwards, which removes the awkwardness from it entirely.

A worked example

A business paid for one ninety-minute session with a screen reader user found through a local organisation.

Within twelve minutes she had identified three things that made the site unusable and dismissed two of the five items on their existing audit as irrelevant.

One was a form field label they had considered fine, which was being announced in a way that made the field meaningless.

The three fixes took their developer half a day.

They have since had her look at two other things and treat it as a normal cost of a project.

The session cost less than the audit had, and produced actions rather than a list.

Give them a task, not a tour

Which is how to run the session.

Do not walk them through the site explaining it.

Give a real task: find out what this service costs, and send us an enquiry about it.

Then be quiet and watch, which is genuinely difficult and is the whole method.

Every time you explain something, you have removed the finding that was about to happen.

Ask questions at the end rather than during, and ask what was most annoying rather than what was wrong.

Record it if they agree

Since the recording does work afterwards.

A few minutes of somebody struggling with your form persuades a developer, a partner, or a client instantly.

Ask permission explicitly, explain who will see it, and keep it internal.

A short clip is far more useful than the whole session, so note the timestamps of the two or three worst moments as they happen.

That is also what you will show in a year when somebody proposes adding a carousel.

One person is not everybody

Which needs saying to keep this honest.

Somebody using a screen reader finds different problems from somebody using voice control, keyboard only, or high magnification.

One session tells you a great deal and does not tell you the site works for everybody.

If you can afford two, a screen reader user and somebody who navigates by keyboard without one are the pair with the least overlap.

For most small businesses one session is the realistic number, and it is enormously better than none.

Say in your record which kind of testing you did, so nobody later assumes it covered more than it did.

Do the obvious fixes first

Which respects their time and yours.

Run the free checks, fix the form labels and the contrast, and do the keyboard walkthrough before the session.

Otherwise you pay somebody expert to find things a free tool would have flagged in a minute.

Arrive at the session with the obvious things done, so the hour goes on what only a person can find.

Tell them what you have already addressed, which lets them skip past it and spend the time on what you could not find yourself.

It also demonstrates that you took the free work seriously before paying anybody, which sets the tone of the session.

Ask what they use, and take the answer seriously

Since assumptions about the technology are frequently wrong.

People use different screen readers with different browsers, and behaviour varies enough that something working in one combination fails in another.

Ask what they use and note it, so that when a developer says it works, you know which combination was tested and which was not.

Many people also use several tools together, or magnification alongside speech, which is a common arrangement nobody designs for.

That detail belongs in your record alongside the findings, since it defines what the session actually established.

The counter-case

It costs money and arranging it takes effort.

A very small business with a five-page site may get most of the value from the free checks and the keyboard walkthrough.

Finding somebody can also take a few weeks, which does not suit an urgent fix.

And a single session on a site about to be rebuilt is timed badly, so do it during the rebuild instead where the findings can change the design.

That is also the cheapest point at which any of this can be acted on.

Fix the obvious things first, find one person through a local organisation, pay a professional rate, and give them a real task.

Arranging it

1. Do the free checks first.
2. Contact a local organisation.
3. State the rate in your request.
4. Book ninety minutes.
5. Give a real task.
6. Be quiet and watch.
7. Ask to record it.

Step six is the hardest and the most important, since every explanation you offer removes the finding that was about to happen.

Persuading whoever pays for it is covered in making the case for accessibility internally.

Frequently asked questions

West Coast Media Solutions Inc. provides web design, web development, hosting, digital marketing, and business consulting to organisations across Canada, drawing on more than twenty-five years in the field.