Skip to content

Accessible Engineering scorecard

An Accessibility Scorecard is a way to get a high-level answer to “Where might we want to improve (in terms of accessibility)?” We do this by looking at a few areas, and indicative behaviours in each.

🧠 We think accessibly

Behaviours that demonstrate this include being familiar with:

⌨️ We develop accessibly

Behaviours that demonstrate this include:

🧪 We test accessibly

Behaviours that demonstrate this include:

  • Including accessibility in some unit tests (For example: jest-axe in unit tests);
  • Including accessibility in some end-to-end tests (For example: axe-playwright or cypress-axe for E2E tests);
  • Testing with assistive technology, in particular testing with screen readers;
  • Doing (or asking for) an accessibility audit.

📈 Scores

For each of the “We…” sections below, review the list of behaviours. Pick a score from 1 to 5.

Score Meaning
1 Not confident at all
2 Not confident
3 Somewhat confident
4 Confident
5 Very confident

The score follows something like a classic Likert scale.

When picking our current level of confidence, consider frequency and depth of doing, documenting, and discussing.

Examples of confidence levels
  • A score of “5 / Very confident” might look like: we do these things with the appropriate frequency; we have quite a lot of documentation about them; we discuss them at the right times, probably quite often.
  • A score of “3 / Somewhat confident” might look like a mix. We do these things quite often, and we discuss them sometimes, but we don’t have any documentation. Or: we don’t do these things often, but we do discuss them sometimes, and have scattered bits of documentation.
  • A score of “1 / Not confident at all” might look like: we do these things rarely or not at all; we don’t have any documentation about them; we don’t discuss them often.

Note: scores are always whole numbers. These measures are subjective and approximate. That means it’s a better fit to say our score is “about 4” instead of ”exactly 4.25”. Any average of scores should be rounded to the nearest whole number.

More detail on picking a score

If we’re struggling to settle on a score, there are two approaches we can take.

  1. Don’t overthink it! 😅 The numbers are vague on purpose anyway, and this isn’t an exact science. The point of scoring is to figure out where we want might to improve, not to figure out the “right” score. We give our best guess, then move on.
  2. Weight by Doing, then Discussing, then Documenting. A score of 4 or 5 means we’re doing good or really good in all three things.
    • The most important thing is Doing. If we’re Discussing and Documenting but not Doing, we aren’t providing value to our customers.
    • If we’re good at Discussing, we’re probably helping to Shift Left and work on accessibility when it’s quickest and easiest to do so.
    • Documenting helps us be more efficient in our Doing and more structured in our Discussing.

Deciding on the scores

We want to answer:

  • What’s your confidence level for the team as a whole?
  • Optional extra: what’s your confidence level for yourself?

Fill in the scores in a similar way to how some teams do planning poker: decide in secret before writing it down.

An example of a scoring process
  1. Read out the category we’re deciding on. For example: “We think accessibly.”
  2. Read through the list of behaviours for the category (at the top of this page).
  3. Briefly discuss the behaviours as they apply to your team. Use the “Examples of confidence levels” section above.
  4. Set a short timer for everyone to privately decide their score, then enter the numbers.
  5. Discuss any numbers at the edges. For example: One vote for 5 among four votes for 3.
  6. Calculate the rounded average. That’s your score!

🙋 Getting help

After you’ve scored the three “We”s, have a think about which area you’d like to make a change in first. What could you do to increase your score in the next few months?

If you’re not sure which area to start with, prefer developing to testing, and prefer thinking to developing. Each area builds on top of the previous one. More accessible thinking leads to more accessible developing. More accessible developing leads to more accessible testing.