Regression Testing Example: A Step-by-Step Walkthrough

Regression testing identifies when a bugfix or new feature breaks existing functionality. It prevents the “one step forward, two steps back” cycle common in high-velocity development. Without it, you ship fixes for reported bugs while silently breaking the core features your users rely on.

This tutorial walks through a concrete regression testing example. We examine a real-world scenario, write a Playwright test, catch a regression, and automate the process.

In this guide

• The scenario
• Step 1: Define the test case
• Step 2: Write the automated test
• Step 3: Run it and catch the regression
• Step 4: Fix and verify
• Step 5: Automate in CI/CD
• Scaling coverage with AI
• High-value regression test examples
• FAQ

The scenario: a fix that breaks everything else

Imagine a standard e-commerce checkout page. It manages two primary business rules:

1. Discount Codes: Users enter a code to receive 10% off their order.
2. Free Shipping: Orders over $50 qualify for free shipping.

A user reports a bug: the discount field accepts expired codes. A developer corrects the validation logic in the calculateTotal() function. The discount bug is fixed.

However, the fix inadvertently alters how the “Order Total” variable is passed to the shipping calculator. Now, even orders exceeding $50 are charged for shipping. The developer only re-tested the discount field. They did not verify the shipping threshold. A regression is born.

Step 1: Define the regression test case

A successful regression test defines the flow, actions, and expected outcomes precisely. Documentation prevents ambiguity during manual or automated execution.

Step 2: Write the automated test

Manual testing is insufficient for frequent releases. We use Playwright with TypeScript to automate this regression check. This ensures the threshold logic remains intact after every code change.

import { test, expect } from '@playwright/test';

test('verify free shipping threshold remains active over $50', async ({ page }) => {
  // Navigate to the store
  await page.goto('https://shop.example.com');

  // Add a high-value item to trigger the threshold
  await page.getByRole('button', { name: 'Add Enterprise Plan' }).click();

  // Navigate to checkout
  await page.getByRole('link', { name: 'Proceed to Checkout' }).click();

  // Assertion: the shipping line must reflect the free threshold.
  // We use a stable selector for maximum resilience.
  const shippingCharge = page.locator('#shipping-total');

  // Verify the baseline behavior
  await expect(shippingCharge).toHaveText('FREE');
});

We prioritize text-based selectors and ARIA roles. This approach creates resilient tests that survive UI refactors. Brittle CSS selectors like .btn-blue-rounded cause false positives when styles change.

Step 3: Run the test and catch the regression

Run the suite after implementing the discount code fix:

npx playwright test

The result confirms the failure:

✘  verify free shipping threshold remains active over $50 (1.4s)
   Error: expect(locator).toHaveText(expected)

   Expected: "FREE"
   Received: "$12.00"

   Locator: locator('#shipping-total')

The test caught the regression. The discount fix broke the shipping logic. Because we ran this suite, the bug never reaches the production environment.

Step 4: Fix the regression and verify

The developer identifies the error: the calculateTotal() function was returning a string instead of a float after the discount update. The shipping logic failed the total > 50 check because of the type mismatch.

After correcting the data type, we re-run the test:

✓  verify free shipping threshold remains active over $50 (1.2s)

The suite is green. Both the original discount bug and the new shipping regression are resolved.

Step 5: Automate in CI/CD

Regression tests only provide value if they run automatically on every change. Integrate the suite into your GitHub Actions or GitLab CI pipeline so no Pull Request is merged unless all regression checks pass. See how to set up CI/CD pipelines for a full walkthrough.

# .github/workflows/regression.yml
name: Regression Suite
on: [pull_request]

jobs:
  test:
    timeout-minutes: 10
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - uses: actions/setup-node@v4
        with:
          node-version: 20
      - name: Install dependencies
        run: npm ci
      - name: Install Playwright Browsers
        run: npx playwright install --with-deps
      - name: Run Regression Tests
        run: npx playwright test

Scaling coverage with AI

The example above covers a known risk. However, the regressions that damage companies most are the “unknown unknowns”: features no one thought to test. Manually writing scripts for every edge case is impossible for modern dev teams.

AegisRunner solves this by automating the discovery process. Our AI crawler explores your application, identifies every interactive element, and generates comprehensive test cases automatically.

• Zero Manual Scripting: Stop writing code for every button and form.
• Auto-Healing: If you rename a class or move a component, our resilient selectors adapt without breaking the build.
• Deep Analysis: Every run includes AI-powered audits for SEO, accessibility (WCAG), and security.
• No Lock-In: Export any generated suite to standard Playwright TypeScript — the same kind of code shown above.

Learn how to automate regression testing to eliminate the maintenance burden.

High-value regression test examples

If you are starting a manual suite, prioritize these high-impact flows first. These are the areas where regressions most directly affect revenue and user retention:

• Authentication: Login, logout, and password reset flows.
• User Onboarding: Signup forms and email verification.
• Search Functionality: Result accuracy and empty-state handling.
• Data Integrity: CRUD operations (Create, Read, Update, Delete) on user profiles or dashboard settings.
• Permissions: Ensuring restricted content remains inaccessible to unauthorized roles.
• Integrations: Verifying third-party APIs (Stripe, HubSpot, Twilio) continue to communicate correctly.

For a deeper look at how to select the right platform for these tests, consult our Regression Testing Tools: The 2026 Buyer’s Guide and the 10 best regression testing software comparison.

Start catching regressions today

Regression testing is the only way to ship code with confidence. Manual walkthroughs are too slow. Hand-coded scripts are too brittle.

AegisRunner offers autonomous regression testing on autopilot. We discover your pages, generate your tests, and catch regressions before your users do.

Start your first crawl in minutes — no credit card required — or compare plans.

Frequently asked questions