If you hire a test manager, you naturally want to improve your test process. Companies often lack the professional preparation and monitoring of the tests. This is where external test managers help, who take over the planning and management and thus improve and constantly monitor the effectiveness of the processes. An external test manager also brings new perspectives with him – which pays off.
As an external professional, however, it is often not easy to have a direct overview of IT projects and ask the right questions. We have collected the five most common errors from new test managers here:
1. Wrong Prioritization of Requirements
When testing the software, the test process should check whether and how well the specified requirements are met. An important task with a high error potential is the prioritization of the requirements. A good test manager should be a good listener and familiar with the product at the beginning. When prioritizing requirements, the overall risk is determined by the probability of occurrence and the effects: If the described use case occurs quite frequently, but the effects of a resulting defect are minor, this requirement may be prioritized lower than a use case that occurs less frequently, but has massive effects on the product if a defect occurs.
2. Defects are not taken into account correctly when estimating the Effort involved
One question that will be particularly pressing for the controlling and finance departments is an estimate of the effort: How much time (and therefore money) does testing cost? When estimating the testing effort, it is therefore particularly important to always reckon with the fact that errors occur during the execution of the test cases and that these must be readjusted, documented and also re-tested after a fix. And don’t forget: Once there is at least one defect, the corresponding test case must be repeated. Estimating the effort is a very difficult task. Good test managers get experience values from the specialist department for existing IT systems and can also fall back on their own experience values for unknown or new systems. However, they are well advised to tend to allow for more time. This may not please everyone in the company, but it provides a valuable buffer in the test process: If fewer defects occur, all the better!
3. Processes in the Defect Workflow are not clearly defined
I’m sure he’ll come. But what exactly happens when a defect is found should be clearly defined before the start of the test runs. Here the test manager is in a mediating and planning role: he has to talk to the development department and develop a workflow. It must be clarified which status transitions there are and how the developer is notified of the defect. The defect workflow must also define how the night test is to be performed. Good tools support the testers by recording errors and directly linking them to the test runs as well as clearly defining the status transitions.
4. Request Test Data and Environments too late
The requirements are prioritized, the test cases are written and the testers are guided: We’re ready to go. But only almost, because the software test still requires test environments, access data and test data. A brief request to the IT department – and depending on the size of the company, testing can begin three to four weeks later. To prevent this from happening, the test manager should take care of test data and test environments at an early stage and discuss his wishes with the specialist departments in the company. This is the only way to ensure that adjustments to the firewall, the activation of access for the testers and also important test data (e.g. address and credit card data that can be used for test purchases of a shopping app) are available at the start of the test. A good test manager therefore already takes care of the test data and environment when writing the test cases.
5. Do not schedule Test Automation
Many people shy away from test automation: The high initial effort and the lack of resources in the company often lead to automation being put to one side at an early stage. A good test manager takes a closer look here: Especially test cases that are repeated with each release, e.g. due to a high requirement prioritization, are an optimal use case for test automation. A major advantage is often cost savings: After the one-time creation, an automated test can be repeated as often as desired without or at significantly lower costs. At the same time, automated tests are significantly faster and thus reduce throughput time. The earlier the importance of the corresponding test cases is recognized, the more worthwhile automation becomes: A good test manager suggests corresponding test cases at an early stage and does not only recognize after the fourth release that regression tests are good automation candidates in many cases.
Some companies simply do not have the personnel to cope with the initial effort of test automation. In such cases, it is the task of a good test manager to propose automation to the project management and clearly present the use case and cost savings. External support should be requested from the test manager for implementation.
No master has fallen from the sky yet: As a new test manager, the support of the test process at the beginning is not always 100% smooth. With our expert tips on the five most common errors, prospective test managers get good tips for the correct handling of requirements prioritization, defect workflow, cost estimation, automation planning and contact with specialist departments to set up test data and environment.
Are you a tester yourself and do you know other typical error sources? Contact us and we will publish new tips and tricks to avoid errors! Just send a message to wwelcome@andagon.com.
