Module 8 · Automation foundations

What to automate, and how to keep it reliable

Choosing checks with a good return, and avoiding flaky tests.

About 10 min reading and 5 min practice (estimates)

Loading progress…

Read

Learning objectives

Pick good candidates for automation.
Explain the test pyramid and why most checks shouldn't go through the UI.
Recognise common causes of flaky tests.

Explanation

Automation is software that checks software. It pays off for checks that are repeated often, are stable, and matter: critical journeys (sign-in, checkout), calculations with many cases, and regressions you never want to see again. It is a poor fit for one-off checks, rapidly changing screens, and judgement calls like "is this confusing?" — that's where exploratory testing shines.

The test pyramid suggests many fast unit tests, fewer API/integration tests, and a small number of end-to-end UI tests. UI tests are the slowest and most fragile, so use them for a few key journeys and test detailed rules at lower levels.

Reliability

Isolation: each test creates its own data and doesn't depend on another test's leftovers or order.
Waiting: wait for a condition (an element is visible, a request finished), never for a fixed number of seconds.
Stable selectors: find elements the way users do — by role and accessible name — not by fragile CSS paths.
Meaningful assertions: check the outcome that matters, with a clear failure message.
A flaky test (passes and fails without code changes) erodes trust; fix or quarantine it quickly.

Example: deciding for Dar Al Souq

Check	Automate?	Level
Discount rounding for 30 price combinations	Yes	Unit
Order is created with the right totals	Yes	API
A shopper can add an item and reach checkout	Yes, one journey	UI end-to-end
New banner design looks balanced	No	Human review
Arabic copy reads naturally	No	Human review

Common mistakes

Automating everything through the UI.
Using fixed sleeps ("wait 5 seconds") instead of waiting for conditions.
Letting tests share data, so one failure causes many.
Ignoring flaky tests until nobody trusts the suite.

Try

Practise what you just read. Answers are checked here; nothing is graded publicly or counts toward Bug Arena.

Exercise

A UI test sometimes fails because it clicks "Pay" before the total has finished updating. What is the best fix?

Choose one answer.

Add a 5-second sleep before clicking.
Retry the whole test up to three times.
Wait until the total shows the expected value, then click.
Remove the assertion on the total.

Review

Check yourself — you should now be able to:

Pick good candidates for automation.
Explain the test pyramid and why most checks shouldn't go through the UI.
Recognise common causes of flaky tests.

References

Martin Fowler — The Practical Test Pyramid (opens in a new tab)

Continue

When you’re ready, mark the lesson complete in the Try step and move on.

Marked as read when you reach the end of the lesson.