The test effort estimation and test planning depend on lot of efforts that affect the accuracy of the estimation. Following are some of the factors that have been encountered in the different testing projects that were executed.
Testers not involved in project estimate
In many cases, when a project is estimated, very rarely the testing groups is consulted, which results in not taking into account various factors that otherwise should have been taken care with respect to testing.
Complex functionality /No requirements
Not understanding the extra functionality of the application while estimating the effort ends up in an inaccurate estimate. There are times when even documented requirements are not made available to do proper test estimation.
Projects where there are many change requirements end up in having an inaccurate estimate. Having a good change management process with assurance group involvement is the key to manage changing requirements.
How much to test – Incremental development
Estimating for a project, which is an incremental development, could be erroneous as gauging the amount of test effort involved in the change in sometimes a challenge. For example, a small change might require the entire application to be re-tested. Assurance group involvement in the impact analysis meeting will be enhance the courage to the right extent.
Shrinking of testing time due to upstream delays
Most project have development schedule the slippages, but the date of the delivery generally has to be maintained, which invariably affects the execution of test plan. In very rare cases does the test manager has the option to negotiate schedule.
Not focus on the critical area of product
Due to lack of planning, the wrong test strategy would be put in place and ultimately enough time and focus may not be given to the high-risk areas.
Most project managers forget to allocate time for communication and coordination, which especially in an offshore/onsite model would result in more effort being spent in the project that planned.
In case incorrect manual scripts are provided for a project like test automation ,the accuracy of the estimate goes awry; so the accuracy of the script need to be checked while estimating the effort required to automate the manuscript.
If the tester expertise is not taken into account, invariably the learning curve would cause your estimate to be skewed.
No test data/Wrong test data
Some test managers forget to consider where the test data comes from, which is some cases in the major factor for the success of the project.
Test environment setup
Not planning for the test environment setup and not setting up the environment on time might result in delay of testing.
Source : “Software Testing – Effective Methods, Tools and Techniques“ book authored by our director, Mr Pradeep Oak and published by McGraw Hill.