75% Reduction in Testing Time, and a 95% Reduction in Testing Effort
Twist removed the biggest pain-point, by drastically reducing the overhead of test maintenance.
Testing was growing unwieldy for AutoTrader.co.uk, the UK's #1 automotive website, with multiple tools, lack of automation, and increasing test maintenance. Twist removed the biggest pain-point, by drastically reducing the overhead of test maintenance. With Twist, the team could add real value, and tested new functionality at a rapid pace.Download PDF
“Twist is helping us add real value and functionality to the website, by helping us build robust regression tests. We tried it with FitNesse and Quality Center, but it never really worked.”
- Current testing was maintenance-heavy, lengthy, and expensive exercise that was complicated by difficult testing tools.
- Development and test environments were geared only towards manual testing.
- One tester would be refactoring a test while another was running the test pack; merge hell was all too frequent.
- Huge traffic volumes, complex 20,000-line Java application, and frequent application changes meant that the test suites had to be robust.
How did Twist help?
- Modular test suite now exhaustively tests the site in 12 minutes, and can easily keep pace with application changes.
- Subversion (SVN) built in so we didn't have to step on each other's toes.
- BAs can write test specifications in plain English.
- Twist is built on Eclipse Integrated Development Environment (IDE) developers and QAs can both use the same product.
- Refactoring is supported and it is driver-agnostic.
Numbers don't lie.
- AutoTrader.co.uk has released new functionality twice a month with zero defects in production, for the past 3 months.
- Automated regression tests would take 50+ minutes to run. With Twist they take 12 minutes.
- Manual QA tests would take 6 weeks to test the entire site. With Twist, it takes a maximum of 2 days.
- With Twist, test coverage has increased to 89%.
- Twist usage has almost doubled from 32 people on one team to 60 people in 4 different teams.
Read the full Auto Trader Case Study (PDF)