Scaling QA for a Global Microstock Marketplace

Challenge

One of the world’s largest microstock platforms — selling stock video, music, 3D models, and animation templates with a built-in AI content engine — had QA but it wasn’t scaling. When we joined the project, the state was:

• ~80 E2E tests, taking over an hour to run
• 170 smoke API tests in Postman + Newman
• No structured test documentation — no way to assess real coverage or find gaps

The platform’s complexity made this particularly high-stakes: licensing compliance, content integrity, and financial transaction accuracy all had to work correctly for both buyers and sellers. A specialized testing strategy was needed, not just more tests.

Solution

We provided Manual QA and Automation QA engineers, starting with an audit of existing coverage before building out from there.

Audit and process optimisation

• Reviewed all existing tests, removed duplicates and outdated scenarios
• Mapped current coverage against platform functionality to identify gaps
• Introduced Code Freeze policy and standardised bug reporting process
• Established QA best practices and structured the full QA workflow

Automation infrastructure

• Built a scalable automated testing framework using Webdriver.io
• Grew E2E test coverage from ~80 tests to 2700, achieving 90% coverage
• Configured parallel execution to bring full regression time down to 2 hours
• Set up Allure reporting for real-time test monitoring and result analysis
• Integrated automated reporting into Jira and Zephyr

Backend and API testing

• End-to-end API automation using Postman Flows — eliminating complex custom scripts
• Contract testing via Postman and Runscope (BlazeMeter)
• API response validation to catch data inconsistency issues at the source
• Dedicated test plans and checklists for all API integrations

Frontend and cross-platform testing

• UI test coverage with checklists and test plans for all key user flows
• Cross-browser and responsive testing via BrowserStack
• Mobile coverage using Appium
• Licensing, transaction, and content integrity checks specific to marketplace logic

Documentation and reporting

• 3153 test cases developed — full platform logic documented and maintained
• Automatic defect logging with status tracking
• Detailed PDF reports on testing activities for stakeholders

Tools used: Webdriver.io, Appium, Postman, Swagger, Runscope, BrowserStack, Allure, Zephyr, Jira, GitHub, VS Code

Results

The platform moved from slow, undocumented QA to a fully automated, continuously monitored test operation:

• 2700 E2E tests automated — up from 80 when we started
• 90% test coverage achieved across all platform functionality
• Full regression in 2 hours — enabling twice-weekly releases without quality trade-offs
• 835 defects identified and fixed before production — significantly fewer critical issues reaching users
• 3153 test cases documented — all platform logic covered and maintainable
• Real-time test result reporting for development and management teams
• Licensing, content, and financial transaction testing fully integrated into the QA process