Free template

Free lean test plan template (one page)

A test plan people actually read. Ten short sections, one page, and entry and exit criteria you can check with a yes or a no.

By the TestPlum teamChecked

Short answer

A lean test plan says what you will test, what you will not, which risks matter most, how you will test, where, who does what, and the measurable criteria for starting and finishing. This one-page Markdown template covers all of that with a filled example for a checkout release.

What a lean test plan is for

A test plan is an agreement made before testing starts. It tells product, engineering and QA what will be tested for this release, how, and what “done” means. Long plans go unread. A one-page plan gets read, questioned and agreed, which is the whole point.

Write one per release or per large feature. Keep it next to the work: in the repository, the wiki page for the release, or the ticket. Update it when the scope changes, not after the fact.

The plan sits above the other templates. It names the risks and areas to test; the test case template holds the tests themselves; the go/no-go checklist is where the exit criteria are checked at the end. If you only have time for one section, write the exit criteria: they decide the release.

Every section explained

The sections of the lean test plan template
SectionWhat to writeKeep it short by
ScopeThe changes in this release that need testing, plus the regression areas.Listing features, not test cases.
Out of scopeWhat you will not test, and why.One line per item. This is where surprises are prevented.
RisksWhat could go wrong, with likelihood and impact (low, medium, high) and how you test for it.Four to eight risks. Test the high ones first.
ApproachWhat is automated, what is manual, what is exploratory.One line per kind of testing.
Environments and dataWhere tests run and the test data they need.A small table.
Entry criteriaWhat must be true before testing starts.Yes or no checkboxes only.
Exit criteriaWhat must be true to recommend shipping.Numbers with a threshold, agreed in advance.
ScheduleThe few dates that matter.Milestones, not a task list.
RolesWho owns the plan, who tests, who decides.Names, not departments.
Evidence and sign-offWhere results live and who signs.A link and a signature table.

A filled example: risks for a checkout release

The risk table does the most work in a lean plan, because it decides what gets tested deeply. Here is the one from the download.

Example risk table from the test plan template
RiskLikelihoodImpactHow we test for it
Charged but no order createdLowHighCritical tests on every build, plus a manual check of the payment provider dashboard
A declined card empties the cartMediumMediumTwo tests for declined and expired cards
Discounts applied twiceMediumMediumTwo tests and a 60-minute exploratory session
Saved cards shown to the wrong customerLowHighA manual test with two accounts on two devices

Good and weak entry and exit criteria

Criteria are only useful if anyone can check them and get the same answer. Write them so the answer is yes or no.

  • The thresholds in the template (100%, 95%, 0, 80%) are examples. Pick numbers that fit your risk, agree them before testing starts, and do not change them in the go/no-go meeting.
  • Count skipped and untested tests against completion, never toward the pass rate. Otherwise skipping the hard tests makes the numbers look better.
Weak criteria next to measurable versions
WeakMeasurableWhy
Testing is mostly done100% of tests in scope have a final result“Mostly” means something different to everyone.
Quality is acceptablePass rate of executed tests is 95% or higherA number you can read from the results.
No major bugs0 open bugs of critical priority or blocker severity“Major” is argued about in the meeting; a severity is not.
Key features tested0 critical-priority tests failing or blockedBlocked counts: a critical test nobody could run is not verified.
Build is ready for QAThe release candidate is on staging and the automated suite passes on itTwo facts anyone can check.

Common mistakes

  • Copying last release's plan unchanged. The scope and risks are different every time; the template is a starting point, not a form to fill.
  • Leaving out the out-of-scope list. It is the section that prevents “I thought you tested that”.
  • Risks without likelihood and impact. Without them every risk looks equal, and testing effort is spread thin. See risk-based testing.
  • Exit criteria nobody can measure. If the numbers are not recorded anywhere, the criteria cannot be checked.
  • No named decision maker. Write who decides go or no-go before the meeting, not during it.

Keep it in Git or a shared document

The template is plain Markdown, so it reads well on GitHub and GitLab, in most wikis and in any editor. Keeping it in the repository has a real benefit: a change to scope or criteria goes through a pull request, so the people who agreed the plan see it change.

In a shared document, turn on version history and add a “last updated” date at the top, which the template already has. Either way, link the plan from the release ticket so it is found when it matters.

Turning the plan into evidence with TestPlum

The plan is a promise. The evidence is the test runs, results, bugs and requirement coverage behind it. Kept in spreadsheets, that evidence is hard to keep current, and the exit criteria end up checked by hand on the morning of the release.

In TestPlum a release has its own runs, requirements and thresholds, and its verdict checks the same exit criteria from recorded results: runs in scope, completion, pass rate, critical tests failing or blocked, open critical bugs and requirements verified. Each failing check links to what blocks it. See release readiness in TestPlum and test runs. The scope, risks and roles stay in your plan; TestPlum does not write the plan for you.

Download

Frequently asked questions

What should a test plan include?

Scope, what is out of scope, the main risks, the testing approach, environments and data, entry and exit criteria, a schedule, roles, and where the evidence and sign-off live. A lean plan fits all of that on one page.

How long should a test plan be?

As short as it can be while still being agreed. For most releases one page is enough. Longer plans are usually needed only where a regulator or contract asks for specific documents.

What are entry and exit criteria in testing?

Entry criteria are the conditions that must be true before testing starts, such as the build being on staging. Exit criteria are the conditions that must be true to recommend shipping, such as a pass rate threshold and no open critical bugs. Both work best as yes or no statements.

What is the difference between a test plan and a test strategy?

A test strategy describes how a team or organisation tests in general and changes rarely. A test plan applies that strategy to one release or feature, with its own scope, risks, dates and people.

Who writes and approves the test plan?

Usually the QA lead writes it, and product, engineering and QA agree it before testing starts. The person who makes the final go or no-go decision should be named in the plan.

Is this template based on a standard?

It is a lean, practical format, not a copy of a standard. Teams that must follow a formal documentation standard should check that standard's required contents and add any missing sections.

Find out whether your next release is ready

Create a workspace, import your tests or write a few, and record a run. Everything on this site is in the product today.