Many times, we have heard from project stakeholders, let’s automate all the testing efforts. My reaction is like “Really?”.
Few questions pop up in mind.
- Is it worth automating 100%?
- Is it possible?
- Will it add any value to the testing?
- Is it required?
Have you ever come across with such questions? I am sure you did.
Let’s understand it one by one. “Let’s automate everything” is a very general statement. I am sure one who expects this from you, he/she hardly has any idea about automation. They may have thought if we automate everything, it will cut down resources, project timelines and will help us deliver the project as early as possible. Well, that’s not possible. Why? Let’s have a look.
How much we should automate?
If we try to do 100% automation, we might end up doing nothing or might mess up with project testing timelines. You need to assess/brainstorm on below points before drawing any conclusion.
- First important thing is to understand project timelines.
- How project will get delivered? Agile or Waterfall Model?
- How/which modules will get delivered for testing and their timelines? which will help in planning your strategy.
- Identify complexity of an application.
- Identify areas where automation is feasible.
- Identify types of testing which can be automated.
After having all above information then you will get to know how much you can automate.
But, Is it worth automating every scenario?
It is very important to understand cost of automation Vs it’s worth Vs its re-usability. Need to maintain balance between these parameters but there are few exceptions.
Let’s understand relationship between these parameters. Most of us, when started learning/implementing automation, first question people ask “Is it reusable?”, then we always think of “Regression”. Is it the only area we should be focusing on?
But, there are many area’s which you can’t test manually or not possible to cover sufficient testing within given timelines, those are the places where automation can add value and is very helpful.
How to find if automation is adding any value? You need to do assessment of the below points.
- Total planned tests Vs critical scenarios covered.
- Time saved.
- Accuracy achieved.
- Percentage of redundant tasks automated.
- Manually impossible but critical scenarios covered.
If you have positive answers for all above questions then go for it. But, I am sure chances are less after considering all above factor’s.
So, It is very important to make wise decision while identifying the areas to be automated, which adds value in testing, rather blindly expecting 100% automation.
Hope information shared above, helps you in brainstorming and making decision on implementing automation.
Have you ever encountered with such question’s any time? and what are your experiences? you can share in comment section.
Thanks for reading! See you soon.
Nice one.
Thanks Ravi