Published
Alpha vs Beta Testing: Differences, Examples and When to Use Each
Short answer
Alpha testing is acceptance testing done in the developer's own test environment by people outside the development team, such as colleagues or a QA group. Beta testing is acceptance testing done outside the developer's environment by real or potential users. Alpha comes first and finds major defects cheaply; beta comes second and shows how the product behaves on real devices, networks and workflows.
Last updated: 2026-11-12
What is the difference between alpha and beta testing?
The ISTQB glossary defines both as types of acceptance testing performed by roles outside the development organization. Alpha testing happens in the developer's test environment. Beta testing happens at an external site, away from the developer's test environment. The ISTQB glossary also notes that beta testing is often used for commercial off-the-shelf software to get feedback from the market.
Google's Workspace help page on testing phases gives a practical version of the same split. Alpha is a limited-availability test, usually by invitation, to verify functionality with a small set of customers. Beta is for broader customer testing, is often publicly announced, and Google says the average beta phase lasts about six months.
| Alpha testing | Beta testing | |
|---|---|---|
| Where | Developer's test or staging environment | Testers' own devices, networks and accounts |
| Who | Employees outside the dev team, QA, a few friendly customers | Real or potential users |
| Group size | Small | Larger, from dozens to thousands |
| Data | Often synthetic or seeded | Real usage and real data |
| Main question | Does it work as intended? | Do users find it usable and reliable in their own setting? |
| Typical defects | Crashes, broken flows, missing features | Device quirks, slow networks, confusing UX, edge cases nobody predicted |
| Support promised | None, or informal | Limited, set by the beta program |
| Comes | First, before beta | After alpha, before general release |
Qase's guides place alpha testing when a product is roughly 70 to 90 percent complete and beta testing when it is at least 90 to 95 percent complete. Treat those figures as rules of thumb from one vendor, not a standard.
When should you do alpha testing?
Run alpha testing once the main flows work end to end but you still expect defects you would not want a customer to see. The aim is a baseline: does the product do what the spec says?
Good alpha testers know enough to give technical feedback and are willing to report in detail. TechTarget notes that alpha data sets are often synthetic and kept small, which makes debugging and root cause analysis easier.
Typical alpha activities:
- Walk through each user story's acceptance criteria on a staging build.
- Try destructive inputs: empty fields, huge files, double clicks, expired sessions.
- Check seeded accounts for each role (admin, member, read-only).
- Record every defect with steps, build number and environment so developers can reproduce it.
If you already write acceptance tests as scenarios, alpha testing is the stage where people run them by hand while the automated versions are still being built. See acceptance test examples for scenario and checklist formats.
When should you do beta testing?
Run beta testing when the product is stable enough that a failure will not drive testers away, and you need evidence from conditions you cannot reproduce in house. Qase describes beta testers as real users who give feedback on design, functionality and user experience.
Beta testing answers questions alpha cannot:
- Does it work on the devices, browsers and operating system versions your users actually have?
- Does it hold up on slow or unstable networks?
- Do people understand the onboarding without being told what to click?
- Which features do they use, and which do they ignore?
Closed beta and open beta
A closed beta limits access to invited testers. An open beta lets anyone who finds the link join. Closed beta suits early feedback and sensitive features; open beta suits load, device variety and market feedback. Many teams run a closed beta first and open it once the obvious problems are fixed.
How do app stores support alpha and beta tests?
Both major mobile stores have built-in tracks, which map neatly onto the alpha and beta phases. These limits come from Apple and Google documentation checked on 12 November 2026, and they can change.
Apple TestFlight. You can invite up to 100 internal testers, who must be members of your App Store Connect team with certain roles, and up to 10,000 external testers by email or public link. If you invite external testers, the first build goes through TestFlight App Review. Internal testers fit alpha testing; external testers fit beta testing.
Google Play Console. Internal testing distributes a build to up to 100 testers for initial quality checks. Closed testing is for a wider invited group, managed by email lists or Google Groups. Open testing surfaces the app on Google Play so any user can join, and you can cap the number of testers (the minimum cap is 1,000). Testers cannot leave public Google Play reviews on test versions, so Google advises setting up a direct feedback channel such as an email address or form.
Web apps have no store, so teams use feature flags, a separate beta subdomain, or an allowlist of accounts to control who sees the pre-release build.
Examples of alpha and beta testing
Example 1: a SaaS dashboard. The team deploys a build to staging. Support and sales staff, none of whom wrote the code, spend a week working through the billing and reporting flows with seeded data (alpha). Two weeks later, 50 customers who asked for early access get the feature behind a flag and use it on their own data, reporting slow exports for accounts with large histories (beta).
Example 2: a mobile app. Developers share each build with 20 colleagues on TestFlight internal testing and fix crashes. They then invite 800 users from a waiting list as external testers and watch crash reports and in-app feedback by device model.
Example 3: a developer tool. An alpha goes to a handful of friendly teams under an invitation and with no support promises. A public beta follows, with documented known issues and a feedback channel.
Do continuous delivery teams still need alpha and beta?
TechTarget points out that in continuous integration and continuous delivery, formal alpha and beta phases do not map cleanly: every CI/CD iteration is in effect an alpha test, and alpha testing never ends. Most fast-moving teams replace the phases with progressive exposure: a build is tested automatically on every pull request, then released to staff, then to a small percentage of users, then to everyone.
Someone outside the author still needs to check the work against what was agreed, and real users still need to confirm it holds up outside your environment. More of the first check can now be automated.
Go/no-go checklist before opening a beta
- Every must-have acceptance criterion has a test, manual or automated, and it passes.
- There are no known crashes or data loss defects on the main flows.
- You have a feedback channel testers can find in under a minute.
- You can see errors and usage from beta builds (crash reports, logs, analytics).
- Testers know what is in scope, what is not finished, and how to report a problem.
- There is a clear exit rule, such as "no open severity 1 issues for seven days".
How do you know the tests behind alpha and beta cover the right things?
Both phases tend to miss coverage nobody can see. Testers explore what they notice, and so do coding agents. If an agent writes the tests for a feature, you still need an answer to "which acceptance criteria have no test, and which user paths do real users take that no test touches?"
TestChimp keeps scenarios as Markdown in your repo, links Playwright tests to them, and shows scenarios with no test so alpha testers can concentrate on the gaps instead of repeating what automation already covers. TrueCoverage, available on the Growth plan or during the trial, compares what real users do in production or beta against what your tests exercise. See spec-driven testing for the workflow.
Frequently asked questions
What is the main difference between alpha and beta testing?
Alpha testing is done in the developer's own test environment by people outside the development team. Beta testing is done outside that environment by real or potential users, on their own devices and networks.
Which comes first, alpha or beta testing?
Alpha comes first. Beta follows once the product is stable enough for external users and before general release.
Are alpha and beta testing types of acceptance testing?
Yes. The ISTQB glossary defines both as types of acceptance testing performed by roles outside the development organization.
What is the difference between a closed beta and an open beta?
A closed beta is limited to invited testers. An open beta can be joined by anyone with the link or by finding the app in a store.
How many testers can I have on TestFlight and Google Play?
TestFlight allows up to 100 internal testers and 10,000 external testers per app. Google Play internal testing allows up to 100 testers, with larger closed and open tracks available. Limits can change, so check the current docs.
Do I still need alpha and beta testing with CI/CD?
The labels matter less, but the purposes remain: independent checking against the spec, then validation by real users. Many teams use staged rollouts and feature flags instead of fixed phases.
See which scenarios have no test before testers do
Link Playwright tests to scenarios in your repo and spot the gaps. 14-day free trial, no credit card.