Skip to content
Tutorials Center
My learning

Topics

Lesson 30 of 60 ยท Test Planning and Documentation

Test Closure Activities and the Test Maturity Model

Learn what happens when testing ends: test closure activities, the test summary report, sign-off, lessons learned, and the TMMi maturity levels.

Testing does not end when the last test passes. Someone has to decide whether the work is truly done, record what was learned, hand the assets over and tidy up. This end phase is called test closure. And on a bigger scale, organizations ask a related question: how good is our testing process overall, and how can it improve? The Test Maturity Model family (TMM, and its successor TMMi) gives a framework for answering that.

When is testing "done"?

Testing is never provably complete (see the seven principles). Teams therefore agree on exit criteria in the test plan and treat testing as finished when they are met, or when management makes an informed decision to stop. Typical criteria:

  • All planned tests were executed and the target pass rate was reached.
  • Required coverage (requirements, risk, code) was achieved.
  • No open defects above an agreed severity; remaining known issues are documented.
  • Defect discovery rate has flattened.
  • Budget or time ran out, and the residual risk was accepted (a legitimate but risky reason).

Test closure activities

Closure usually happens at the end of a release, a sprint series or a project phase. The common activities are:

  1. Check completion. Confirm planned tests ran, defects are in a final state, and deliverables exist. Verify that open defects have a recorded decision (fix, defer, accept).
  2. Prepare the test summary report. Summarize what was tested, results, defects, deviations and residual risk (see below).
  3. Archive testware. Store test cases, scripts, data, environment configuration and results so the next release can reuse them (the testware is a valuable asset).
  4. Hand over. Give the maintenance or operations team the known issues, workarounds and the regression suite.
  5. Clean up. Release test environments, accounts and tooling; delete or mask test data containing personal information (see Test Data and Environments).
  6. Hold a retrospective. Discuss what worked, what failed and what to change, then record lessons learned.
  7. Collect metrics for the next estimate: effort, defect counts, escape rate, cycle time.
  8. Obtain sign-off. Stakeholders formally accept that testing met its goals.
flowchart LR
  A["Check completion"] --> B["Summary report"] --> C["Archive testware"] --> D["Hand over<br/>known issues"] --> E["Clean up"] --> F["Retrospective<br/>lessons learned"] --> G["Sign-off"]

Closure is a short checklist done in the same order each time.

The test summary report

The test summary report (also called test closure report) is the main closure document. It is addressed to stakeholders who want to know whether it is safe to release. Typical sections:

  • Scope and context: product, version, environments, period.
  • Summary of testing: test cases planned, executed, passed, failed, blocked, not run.
  • Defect summary: found, fixed, open, by severity and by module; defect leakage if known.
  • Coverage: requirements and risk areas covered.
  • Deviations: what changed from the plan and why (for example, areas not tested).
  • Residual risks and known issues, with workarounds.
  • Lessons learned and recommendations.
  • Conclusion and recommendation: ready, ready with conditions, not ready.

Write it so a manager can read the first page only and still make a decision. A detailed example is built in Project: Test Plan and Report.

Try it: build a summary report from raw results

This stand-in turns a list of results and defects into the numbers a summary report needs and prints a recommendation.

JS Generate a test summary

Press Run (or Ctrl + Enter) to see the console output here.

No output.

The recommendation logic is deliberately simple. Real organizations write their own rules in the test plan and agree them before testing, so that nobody is negotiating under deadline pressure.

Lessons learned

A retrospective is only useful if it produces actions. Good prompts:

  • What found the most important defects? What found none and could be dropped?
  • Which defects escaped to production, and why did tests miss them?
  • Where did we lose time (environments, data, unclear requirements)?
  • Which estimates were wrong and by how much?
  • What one change will we try next time?

Record each finding with an owner and a date, and check them at the start of the next cycle.

The Test Maturity Model (TMM and TMMi)

So far this lesson covers one project. A separate question concerns the whole organization: how mature is our testing process, and how do we improve it systematically? The Test Maturity Model (TMM) was an early model for this. Its successor, TMMi (Test Maturity Model integration), is maintained by the TMMi Foundation and is modeled after CMMI, a general process-maturity model. It describes five levels, each with process areas that must be satisfied.

Level Name Idea
1 Initial Testing is ad hoc and chaotic; success depends on heroes; testing is mixed with debugging.
2 Managed Testing is planned and monitored as a separate activity. Key areas: test policy and strategy, test planning, test monitoring and control, test design and execution, test environment.
3 Defined A standard organization-wide test process exists. Key areas: test organization, test training, test life cycle and integration, non-functional testing, peer reviews.
4 Measured Testing is quantitatively managed with measurements, quality evaluation and advanced reviews.
5 Optimization Process improvement is continuous: defect prevention, quality control, test process optimization.

Organizations are assessed against the model and can earn a formal rating, but many teams simply use the levels as a map: "we are somewhere between 1 and 2; to reach 2 we need a written test policy and a real test plan for each project." The model is a guide, not a goal in itself; chasing a certificate without real improvement is a common trap.

Do not confuse it with the test pyramid or with ISTQB, which certifies people, not organizations.

Check your understanding

Quick check ยท 4 questions

of 4 answered You scored / 4 ยท perfect!

  1. 1. Which is NOT a typical test closure activity?

    Closure wraps up testing; feature development is not part of it.
  2. 2. What is the purpose of the test summary report?

    The report condenses results, open defects, deviations and residual risk into a decision aid.
  3. 3. Why must test data with personal information be cleaned up during closure?

    Leftover personal data in non-production systems is a common source of leaks and compliance problems.
  4. 4. At which TMMi level does the organization have a standard defined test process across projects?

    Level 3 introduces an organization-wide defined process; level 2 manages testing at project level.

Summary

  • Testing is finished when exit criteria are met or management accepts the remaining risk, not when "all bugs are found".
  • Closure covers completion checks, the summary report, archiving testware, handover, clean-up, retrospective, metrics and sign-off.
  • The summary report should support a go/no-go decision on its first page.
  • TMMi describes five maturity levels from Initial to Optimization for improving the whole organization's testing process.

Next up: Risk-Based Testing.