When comparing automated and manual software testing, there is a clear winner. It’s obvious that automation has outstanding value. To that end, Awetest Customers have seen up to 20x increase in coverage (compared to existing manual testing efforts). At one customer, we were able to execute over 100 tests on 10+ devices in under 2 hours. To put it in perspective, it took the customer 3 Manual Testers 1 week to deliver the same level of coverage.
It’s established that automation can’t be beat when it comes to speed and coverage, so can it one day replace manual software testing entirely?
The short answer: No.
There will always be a need for manual software testing. Certain types of interactions within an application will be too unique, too complex or too costly to develop automated tests for.
Why is manual software testing necessary?
Manual testing allows for finding more real user issues because automated tests,are predefined sets of interactions that represent a single user/user groups interpretation of what should be tested. The whole concept of exploratory testing can’t be automated…today. i.e. automated tests can’t decipher human emotion and visual components. Is the application UI easy to navigate, is the site aesthetically pleasing?
That is specifically something that a computer cannot do (yet), which is one reason why manual testing is necessary. There are also reasons why manual testing is just the better options in certain cases, which are listed here: When (Not) to Automate Testing.
So manual software testing is necessary, but there are certainly times when it can (and should) be replaced.
Strategically Replacing Manual Software Testing
With the rapid adoption of mobile usage and development concepts like Agile Programming, QA now has to deal with the fact that there are more applications to test, on more platforms, in an even more compressed timeframe. The exponential increase in device-OS-Browser combinations coupled with above-mentioned trends is a recipe for manual QA turmoil. Bottom Line, the enterprise doesn’t have enough bodies (or budget) to support the magnitude of testing needed to be done
This is when you need to leverage technology and develop a strategy to deal with this growth. Specifically you need to evaluate the ROI on developing automated tests, instead of manual software testing, for your apps and applications. Under a scenario with limited target platforms, some tests are more economical to just manually test, however under this new paradigm these same tests can now be economically automated. We have seen, on average, building automated tests takes 5-10x the effort required to execute the same tests manually. New situation means new decisions.
First, automated tests are extremely effective in testing permutations of devices, OS, and browsers. A manual software tester would have to cycle through all of the combinations one click at a time to test the app or application. But an automated test that is expensive to develop initially could usually be run against each combination with only minor alterations to meet the discrepancies that arise. Even apps that were developed for specific platforms like iOS or Android would be effectively tested with this approach. Most recently we’ve seen a big push towards re-writing legacy applications to support Responsive Design where our customers are seeing a BIG ROI in implementing a single set of tests that can be executed across Desktop Browsers, Tablets and Phones
Most importantly, the ROI would continue to accrue going forward. Manual software testing costs occur at every release, while automated test costs occur during their development, which mainly occurs upfront when they are initially developed. If you have the belief that your application or app has staying power for near future, then it is worth the investment to develop a long-term strategy to leverage automated testing.