Automated Regression Testing Cut Effort by 60%

Feb 21, 2026 Bats, Quality Assurance

A large UK insurer modernising a heavily customised system faced a familiar bottleneck in the delivery process: manual regression testing.

With a steady release cadence, regression testing was squeezed into the final days before the go-live, resulting in delayed releases and increased uncertainty. This is where automated regression testing makes the biggest difference – by running repeatable checks more often, providing faster feedback, and reducing release risk through earlier defect detection.

The starting point: regression as a last-minute scramble

Before the new system was introduced, regression testing relied entirely on manual effort. One manual tester was responsible for verifying both new functionalities and regression coverage. With a 6-week release cycle, around 5 MD were needed for full regression testing. Regression tests were only executed right before a release and there were no automated regression tests in place.

The risk spike: a new system, even more testing pressure

Following the reimplementation of the system, the release cadence remained at 6 weeks, but the testing burden increased. New features were continuously added, expanding the scope of regression testing and increasing the time required for it to over 8 MD. This was still the responsibility of a single manual tester. With regression testing only taking place near release, the organisation was taking on a higher level of risk into production, simply because feedback was arriving too late.

The turning point: automated regression testing with BATS

To break the cycle, Sollers implemented automated regression testing using the Business Automation Testing Suite (BATS). BATS supports the creation, execution, and reporting of test scenarios to accelerate repeatable checks and reduce manual effort, which is especially valuable in UI-heavy processes.

The solution included:

Implementation of around 180 automated regression tests
A model in which one manual tester could focus on new features, while regression testing became a predictable, repeatable routine
Automated regression runs executed daily to enable earlier defect detection (rather than discovering issues only at release time)

Results: reduction from more than 8 MD to 3 MD and earlier defect detection

Following automation, the time required for full regression testing decreased to approximately 3 MD (2 MD manual + 1 MD automated), representing a reduction of around 60% from the initial 8 MD. Thanks to the earlier detection of defects, the quality of releases improved, and the manual tester gained time to validate new functionalities even as the scope continued to grow. These outcomes are consistent with the benefits that teams typically gain from shifting repeatable regression checks into automated, continuous execution.

Why it matters

This business case illustrates a common enterprise pattern: when regression testing is manual and carried out at last minute, it consumes capacity, compresses schedules and increases the likelihood of defects being released into production. Automated regression transforms the economics of quality. Daily, repeatable runs make issues visible earlier, when they are cheaper and less disruptive to fix. This also improves planning over time: regression effort becomes more predictable, release readiness is easier to assess, and QA teams can dedicate more time to higher-value tasks such as exploring new functionality, improving coverage in high risk areas, and supporting faster, safer releases.