What is User Acceptance Testing?
User acceptance testing (UAT) is a testing phase where actual end users or their representatives validate that a product or feature meets their requirements before it goes to production. It is the final quality gate before release.
UAT answers a simple question: "Does this do what we agreed it would do?" It compares the built product against the acceptance criteria defined during planning. If the product passes UAT, it is approved for release. If not, it goes back for fixes.
Why UAT Matters
Internal testing catches bugs. UAT catches misunderstandings. A feature can pass all automated tests and still fail UAT because the team interpreted a requirement differently than the user expected.
UAT also builds stakeholder confidence. When a key customer or internal champion has personally verified a feature before launch, they become an advocate rather than a critic.
How to Run UAT
Define test scenarios based on real workflows. Do not give testers a list of buttons to click. Give them tasks: "Import your Q3 sales data and generate a performance report." This mimics real usage and reveals issues that scripted testing misses.
Select the right testers. Ideal UAT testers are actual users or stakeholders who understand the domain. Internal team members who are too familiar with the product will not catch the same issues as users approaching it fresh.
Provide a structured feedback mechanism. Use a form or tracking system where testers log issues with severity, steps to reproduce, and expected vs. actual behavior. Unstructured "it does not work" feedback is hard to act on.
Set clear pass/fail criteria. Define how many critical issues block release and how many minor issues are acceptable. Without criteria, UAT becomes an endless cycle of feedback and fixes.
UAT in Practice
Enterprise software companies like SAP and Oracle formalize UAT as a contractual milestone. Customers sign off on UAT before the implementation is considered complete. This process ensures that the delivered product matches the agreed requirements.
At Atlassian, internal employees participate in "dogfooding" programs that function as extended UAT. Teams use pre-release versions of their own products for daily work, catching issues that external testers would eventually find.
Common Pitfalls
- Skipping UAT for speed. "We tested it internally, it is fine" leads to post-launch surprises. Always validate with real users.
- UAT as the only testing. UAT should complement automated testing and QA, not replace them. Do not waste UAT testers' time on basic bugs.
- Wrong testers. Developers testing their own code is not UAT. The whole point is fresh eyes from the user's perspective.
- No time to fix issues. Schedule UAT early enough that there is time to address findings before the release date.
Related Concepts
UAT validates against acceptance criteria and is a component of the Definition of Done. It feeds into release management as a go/no-go input. Beta testing is a broader form of UAT with a larger group. Usability testing evaluates design quality, while UAT evaluates requirements satisfaction.