Software Test Automation Set Up for Verifying Website Changes

Summary

A global advertising agency wanted to set up test automation to speed up testing and reduce the effort of verifying changes introduced to its website. INNERLUXES built a visual UI test automation framework on Pytest and Selenium that screenshots specified pages and overlays them against references to catch unplanned changes, integrated it into the agency's Azure DevOps CI/CD pipeline with scheduled runs and auto-distributed reports — cutting user-reported UI issues by 30% and reducing defect recognition to under a minute.

About the Client

The Client is a global advertising agency with a large international presence across many countries.

The Challenge

The Client wanted to set up test automation to speed up the testing process and reduce the effort spent on verifying changes introduced to its website.

The Solution

INNERLUXES's software test automation team took on the project. They set up the test automation process to reduce the effort spent on verifying the website's UI, performing testing activities with an open-source framework (Pytest) that executed automated tests based on Selenium to automatically check website pages for changes and reveal unplanned ones.

The test team configured the framework to screenshot the specified website pages, then overlay and compare those screenshots against a reference to locate changes in the UI.

For each page, test engineers configured a preferred screen resolution and, where a page featured elements that required considerable time to load, specified an appropriate load time.

To reduce testing time, INNERLUXES's test team integrated the tool into the Client's CI/CD pipeline running on Azure DevOps, which enabled automatically triggering test execution on a set schedule.

The test results were presented in reports showing the total number of executed tests and the number of passed and failed tests in graphs and diagrams. The reports were automatically generated and distributed to the website support team on a set schedule.

The Results

• The number of UI problems reported by website users decreased by 30%.
• Recognizing UI defects took no longer than 1 minute thanks to convenient visualization.
• The automatic distribution of daily test reports helped the website support team reduce the time and effort needed to fix reported UI defects.

Technologies and Tools

Selenium, Pytest, Allure Framework, Azure DevOps.