At times when you perform software testing, it’s beneficial and important to have testing tools that can automate testing as it can 
save a lot of time and may achieve tasks that may be difficult to perform with manual workforce. Before applying automated testing in an organization, it must be evaluated first as it may not be applicable to every software or test cases. This article will tackle the differences between manual software testing and automated software testing, their pros and cons, and when to perform manual or automated testing.
Manual Testing
Manual testing is the process of testing software manually for bugs and defects. Testers often follow a test plan and test cases to ensure the completeness of the testing. Testers can also use manual tools like decision tables, test scripts to be used, and a simple checklist of the things to be tested and investigate.
A test plan is a document that describes the intended scope and features to be tested, testing tasks, resources performing each task, schedule of testing activities, and any risks requiring contingency planning.
Test cases are detailed documents of the test input, expected results and execution conditions of a test item. It consists of clear steps in testing to be taken by the testers with a column for the results. It is mainly used to verify that the software runs as expected.
Automated Testing
Test automation is used to control the execution of tests, compare the actual and expected results, the setting up of preconditions, and other test control and test reporting functions through the use of software. It involves automating manual process through writing test scripts that would do the testing and can be run repeatedly. Example of the automated tasks that can be done through test automation:
- Test data creation
- GUI interaction
- Problem detection
- Product installation
- Performance Analysis
- Defect logging
Pros and Cons of Automated and Manual Testing
|
Pros of Manual Testing
|
Pros of Automated Testing
|
|
If the test case can be run in a small number of times, then manual testing can be used. It costs less than automating it.
|
Creates and executes tests quickly. Many testing tools support the execution of performance test and provide record and playback features.
|
|
It allows tester to test more through random testing or ad-hoc. Sometimes manual testing is more beneficial compared to automated testing because it enables the tester to test all features, play more with the program’s functionality and features, and more likely to find real user bugs.
|
Supports partial test automation. It automates parts of a program but not automating all process. This is helpful if, for example, a report cannot be created, then a software tool can be created to find the reason or error behind the process.
|
|
Can be used on multiple testing machines at the same time.
|
|
|
Beneficial on Regression testing as you can catch the new changes and deliver results quickly.
|
|
|
Can be used on stress testing as it can handle and test large amount of data.
|
|
Cons of Manual testing
|
Cons of Automated Testing
|
|
Manual testing is time consuming as test cases and new features need to be re-tested for every release.
|
Software tools for automated testing are expensive. There are new tools available that can help testers in automating GUI. These new tools will not only increase and improve productivity, it can also help reduce test time and cost.
|
|
On automated testing, you cannot test everything; it should still be combined with manual testing.
|
Guaranteeing sufficient test coverage while knowing how much to automate, what to automate and what not to automate holds the key to a successful test automation strategy. A toolkit or platform for test automation that includes process and integration with leading open source and commercial software testing tools is a must to make any test automation strategy successful. The platform should address end-to-end test automation requirements including the 6 points you mentioned but not limited to it. In short the platform or toolkit should effectively address challenges around selection, prioritization and execution of test automation. And yes there is one another important thing the platform must provide quick Go/No-Go decisions for production releases.
I guess for all testing tools there is always an opensource product if any team has to save some bucks. Yeps the cost goes up due to limited tech skills sets between testing community. However if this is overcome i guess on long run automation is less costlier than manual.