Leading software company eliminates visual regressions in CI/CD with BrowserStack Percy

A globally recognised leader in enterprise software and open source technology, this company builds and maintains a wide portfolio of platforms and products used by organisations across industries worldwide. With large-scale platform migrations, complex multi-environment deployments, and cross-functional teams including developers, QA engineers, and designers all working in parallel, ensuring the visual integrity of its web applications at every stage of the development lifecycle is a significant and ongoing challenge.

The Senior Software Quality Engineer at the centre of this story has worked across multiple domains within the organisation, including customer portals, case management systems, chatbot AI, and Salesforce instances. With a focus on bringing efficiency and reliability to every product, the need for a scalable, automated approach to visual regression testing became clear during a full platform migration from Drupal D7 to Drupal D10.

By adopting BrowserStack Percy and integrating it directly into their GitLab CI/CD pipeline, the team eliminated manual visual regression checks, enabled cross-browser visual validation in a single execution run, and built a collaborative review workflow that keeps developers, QA engineers, and designers aligned on visual quality before every release. The team has relied on Percy for three and a half years and continues to use it today.

After exploring visual testing options during the D10 migration, the team selected BrowserStack Percy for its integration capabilities, AI-powered visual engine, and collaborative review features. Percy was integrated with their existing Cypress automation framework and embedded directly into the GitLab developer pipeline, making visual regression testing an automatic part of every code push.

1. Seamless integration with Cypress in under an hour. Percy was integrated with the team’s existing Cypress automation framework in approximately 50 minutes. The documentation was clear and straightforward, and no significant rework of the existing automation setup was required. This low barrier to entry meant the team could get value from Percy almost immediately without a prolonged onboarding process.

2. Integration into the GitLab developer pipeline. Rather than running visual tests as a separate, manual step, Percy was embedded directly into the GitLab CI/CD pipeline alongside the developer workflow. Every time a developer pushed changes to the repository, the automation tool ran automatically, and an email report was sent to the team showing what visual changes had been introduced. This made visual feedback immediate and removed the need for QAs to manually initiate visual checks.

3. Cross-browser visual validation in a single execution run. With Percy, the team could validate visual consistency across Chrome, Firefox, Edge, and Safari in a single pipeline run, rather than repeating manual checks across each browser separately. Screenshots were rendered across multiple browsers and resolutions automatically, and any differences were surfaced in a centralised review dashboard.

4. Percy Visual AI Engine to eliminate false positives. Percy’s Visual AI Engine intelligently filtered out noise caused by minor rendering differences, anti-aliasing, and hardware-level GPU variations, ensuring that only meaningful visual changes were flagged for review. This meant the team spent their time on real visual regressions rather than false alarms generated by environmental inconsistencies.

5. A centralised Percy cloud as the shared visual review layer. The Percy cloud dashboard gave every stakeholder, including developers, QA engineers, and designers, a single place to see the current visual state of the application, review flagged differences, and approve or reject changes. Designers could check whether the built output aligned with the intended design. Developers could see exactly which of their changes had introduced visual regressions. QA engineers could track the review and approval status of every snapshot across every build.

6. Root cause analysis for faster debugging. Percy’s root cause analysis capability allowed the team to deep-dive into specific visual diffs and see exactly what had changed at the HTML, CSS, or positional level. Rather than visually inspecting a diff and guessing at the cause, engineers could see the precise code changes responsible for a visual regression and act on them directly.

“It does not take a lot of time to set up. I integrated Percy with Cypress personally and it took me hardly 50 minutes. The seamless integration with any automation framework is one of the biggest things that drew me to it.” – Senior Software Quality Engineer

Three and a half years after adopting Percy, the team continues to rely on it as a core part of their quality engineering workflow. The impact has been felt across release confidence, team collaboration, and the efficiency of the QA process.

1. Visual regressions caught before they reach production. The primary goal of every tester is a successful release, and Percy delivers on that directly. By automating visual regression checks as part of every pipeline run, the team now catches visual bugs at the point of development rather than in QA or, worse, in production. Developers receive immediate feedback on the visual impact of their changes, making defects faster and cheaper to fix.

2. Cross-browser visual validation at scale. What previously required manual repetition across four browsers and multiple resolutions is now handled in a single automated run. The team can validate visual consistency across Chrome, Firefox, Edge, and Safari simultaneously, with results available in the Percy dashboard within minutes of a pipeline run completing.

3. Developer confidence in visual quality. Developers now have confidence that visual testing is running automatically alongside their functional tests. Rather than relying on QA to manually catch visual regressions after the fact, they receive automated feedback as part of their normal push workflow, allowing them to address visual issues while the context is still fresh.

4. Stronger collaboration between designers, developers, and QA. The Percy cloud dashboard has become a shared visual reference point for the whole team. Designers use it to verify that the built output matches the intended design. Developers use it to understand the visual impact of their code changes. QA engineers use it to track review and approval status across builds. This centralised view has replaced informal, fragmented feedback with a structured, transparent collaboration workflow.

5. A visual testing practice sustained and scaled over three and a half years. The team adopted Percy during a high-pressure platform migration and has continued to rely on it ever since. This longevity reflects not just the tool’s capability but its fit with the team’s workflow: easy to integrate, easy to maintain, and consistently valuable across different projects, domains, and team members.

6. Rapid onboarding for new team members and frameworks. The ease of integration means that new team members or new projects can adopt Percy quickly without significant ramp-up time. The 50-minute Cypress integration is a benchmark the team references when introducing Percy to others, and the straightforward documentation means that the setup experience is repeatable and predictable.

“Percy gives our developers confidence that visual testing is taken care of. Our designers review the Percy cloud to check what is there. Our developers go to the reports and see what is breaking. It increases team collaboration across the board.” – Senior Software Quality Engineer