"Some birds aren't meant to be caged, their feathers are just too bright"- Morgan Freeman, Shawshank Redemption. This blog is from one such bird who couldn't be caged by organizations who mandate scripted software testing. Pradeep Soundararajan welcomes you to this blog and wishes you a good time here and even otherwise.

Monday, February 11, 2008

Educating customers on testing

Everyone has heard about testing and unfortunately everyone has heard about it from different sources that say different and contradicting things.
  • One of the customer I worked for thought - testing is about delivering bug free product - which, in my opinion and many other experts with whom I interact is an impractical idea about testing.
  • Another customer I worked for thought - testing is about delivering quality - which, again, is a bad idea.
  • Another customer thought - testing is about executing test cases and based on the test case pass/fail ratio decide to release - even if the test cases isn't helping testers catch bugs.
  • Another customer thought - testing is a job of repeating tests to compare build quality and take a decision to release a specific build to his/her customer.
  • Another customer thought - testing is meeting specification.
  • Another customer thought - testing is satisfying the customer.
Working with multiple customers is a huge challenge for test managers or testers like me and you. I have so far been repeating the word "customer" and whom do you think I am referring to?

I am referring to the manager we report to and the manager we report to is also our first customer. When testers quit their job and join another one, they never ask their customer (manager) what he/she thinks about testing. That's why some people feel they were performing good in their previous company or they are able to perform better in their recent organization - reason being - their idea of testing matched with that of their manager's idea.

If our idea of testing doesn't match with that of your customer's ( manager ), you and I seem to face a lot of trouble in meeting our customer's expectation and are unable to satisfy them.

Everyone, including you and me might be thinking that the idea we have towards testing is the right one. I have been realizing and teaching that there isn't THE RIGHT WAY to do something. Rapid Software Testing taught me cost v/s value and I took it very serious and I am practicing it.

All customers, irrespective of what definition they have towards testing look for cost v/s value. I am striving to deliver a fabulous value to my customers and I am facing a challenge since I realize what value means to me is different from what value means to my customer (manager).

That's why I was looking as a bad tester (despite finding a lot of bugs) to some of my managers in the past. They think the value they wanted to get out of me is - to execute thousands of scripts or test cases.

Unfortunately I wanted to add more value to the testing efforts which wasn't liked and was fired for attempting to do that. They weren't wrong - they just cleared a trap that they dug in when they hired me.

I learned - what I think as value might be a threat to someone else's idea of value and I value this learning I had about value.

Once I (and my team) found a lot of bugs, so much that there weren't as many bugs that the test case document had helped the team catch in the past. The questions from our customer started to focus on why we had a bad test case document and *not* "Why can't we do more of what you guys did?"

A definition, customer has towards "value" doesn't appear to change despite showing enough evidence. Its hard to continue working for such a customer but there is another dimension to think - money.

Some of those who attend my Exercises for a testers mind - A Rapid Software Testing Approach workshop say, "Well, this is great. I am excited to do testing this way but I suspect the problem is - I don't think my manager would be interested on any of these things I do"

I have had different replies to the above:
  • Stop thinking that your manager would not be interested. That's the first step.
  • Convince your manager to take this workshop.
  • Educate your manager on this.
  • Just do it and make your manager ask, "How are you doing these things?"
Here is a list that I suggest you to use, if you realize facing similar challenges: (I practice the same, too)
  • Educating your customer, starts with you.
  • Stop thinking that your first customer wouldn't be willing to see more value.
  • Provide evidence for him/her and ask her to critique the evidence you provide.
  • Request for a discussion about - What you think of value v/s What your first customer thinks of value.
  • If it fails the first time, provide repeated evidence and suddenly do things the way your first customer wants to do - see if your first customer comes back asking why suddenly things appear not so good.
  • If you have been successful educating your first customer, then motivate your first customer to educate his/her first customer.
BTW, what do you plan to educate your first customer on?

Here is a list of heuristics about educating your first customers:

  • Educate yourself before you educate your first customer.
  • Testing is questioning a product in order to evaluate it AND testers provide quality related information to stakeholders/management to help them take better informed decisions -- James Bach & Cem Kaner
  • 100 test cases might not help a tester catch 1000 bugs and if you want more bugs, exploratory testing ( with a mission/charter) is an important approach to be incorporated. Adding another 1000 test cases is not a solution.
  • Specification document is just one or the oracle that helps a tester find bugs. End users or your customers never use specification oracle to spot bugs and hence going beyond specification can help in finding more bugs.
  • Session Based Test Management is a way where Exploratory Testing can be tracked and managed more efficiently.
  • Spending less time on documentation that's going to end up wasteful is a nice way to get more time executing tests to find more bugs.
  • "If you are too bothered about repeatability of tests - then it is good to have a document of test ideas and *not* test cases that tends to have more ambiguity."
  • "Diversification of approaches and techniques helps testers find more and different bugs that otherwise might be caught by a customer."
  • Software testing books are not the *only* books that helps tester find bugs and hence the library should stock a lot of books on thinking, ideas, pattern, philosophy, psychology, humans, communication...
  • We didn't appear to find a bug because of one or all of these:
      • We did not have the skill.
      • Someone did not let us know when they put it.
      • A stakeholder asked us not to report it although we found it.
      • We didn't intend to find that.
      • We thought it is OK to not to find that, if it existed.
      • We were finding other important bugs thinking opportunity cost.
      • We didn't have the right oracle.
  • { You might want to try a lot more, provided some of the above works for you. Sometimes I am aware that I am going to fall into a trap but that's a way I choose to avoid bigger traps }
There might be a trap you may fall into. Some of your first customer's might not deserve/ might be reluctant / be egoistic - to such an education, sense it out smartly and stop educating him/her to get more time to educate yourself. An indication that you fell into a trap is when your first customer fires you.

+ A heuristic is a fallible method of solving a problem
++ An oracle is a principle or mechanism by which we (humans) identify problems.

Pradeep Soundararajan - http://testertested.blogspot.com - +91-98451-76817 - pradeep.srajan@gmail.com

"The test doesn't find the bug. A human finds the bug, and the test plays a role in helping the human find it." --


Anonymous said...

Great post on the very important topic - understanding and educating Software Testing.

Now after my BBST I understand each word of this post so very well.
But I do not still see this as a feasible solution. There are so many factors that block this solution.
For example, my first customer (manager or test manager) will be happy with all the issues I find and defects I log but at the end of the day he will still call me for a one-to-one session and say all this is fine but now tell me why the test cases didn't find them; why did you report a defect that has not come from the test lab. After explaining about things like learning, exploring the product he will say ok fine, now you document each and every step or scenario that you tried because we should be able to give these test cases to a new comer or a trainee and she should be able to execute them without much assistance.
Regarding requirements these customers totally agree with the ever changing requirements and also agree that the old requirement docs are never updated but they also expect we testers to keep our test cases updated so that these changing requirements can be always tracked through test cases in some Test Management system. Note here these manager-customers have problem, its not a problem reported by client-customer.

Next when you try to educate them most of them will say - please I know all this, I have been testing for years and have many successful projects behind me. I have the clients and managements appreciation letters, I have been the awarded tester of the year in my organisation, blah-blah-blah...

This all seems like a lobby which is not easy to break through. And based on my experience I say that smaller companies are more open to experiment with new ideas and suggestions. I recall the Gore Associates example given by Malcolm Gladwell in his book "The Tipping Point". He says in small groups new ideas and information go from one person or one part of the group to the entire group all at once and be enough to tip.

- Sangeeta


Ankur said...

I have heard some stupid arguments about getting a bug free product and people insisting on it. Customer education and also internal management education is very important here. Trouble increases as most of the managers in software company are from development background and are generally not trained managers (or MBA) like we have in other industries.