TrainingConferencesAbout UsContact UsAdvertiseSQE.com

StickyMinds.com: brain food for building better software

Join

Join

Clarify Your Search Criteria
Tips on Using Our Search Feature(s)
StickyMinds.com Home
ResourcesEventsTopicsPowerPassJobs
Software Testing & QA Online Community  >  Detail: Fear of Automation



A StickyMinds.com Original
Article Picture
Fear of Automation

By Linda Hayes

Send This Content to a FriendGet a Short Link to This ContentPrint This ContentSee User Comments About This Content

Summary: Can a pan create an award-winning meal without the chef? Can a scalpel perform surgery without a skilled surgeon to lead it? Can a paintbrush create a work of art without the artist? These tools and test automation are only as good as those who use them. In this week's column, Linda Hayes debunks the popular idea that streamlining through test automation will mean certain termination for employees. Linda explains that test automation can actually equate to becoming an indispensable team member.


Tricentis
StarWest Speaker Better Software Conference and EXPO 
 
It's hard enough to succeed at test automation given its technical complexity, but it is downright impossible to succeed if team members aren't motivated. Unfortunately some are still afraid of automation because they think it might replace them. Not only is that simply not true, but the opposite is actually true: automation provides a golden opportunity for us to become indispensable. 
 
More Productivity = Higher Value 
 
When I hear a manual tester or analyst say that test automation will replace them, my knee jerk reaction is to wonder if Stephen King is afraid that MS Word will replace him. After all, both are tools in which you capture your knowledge (or imagination) and replicate it. Both literature and test automation are only as good as the writer, so it's your expertise that counts – not the tool. 
 
Look at it this way: Stephen King's net worth increases each time a copy of his work is sold; it doesn't decrease. Your productivity increases each time your automated tests are executed; as your productivity goes up, so does your value to the company. 
 
One of my favorite examples of this formula in action is of an automation consultant who developed thousands of hours of automated executions for several applications – single-handedly. He survived five layoff rounds in the company over a five-year period, and the initial layoff was for all non-employees. Each time the layoff list was drawn up, his name was on it. Each time multiple managers sprang to his defense, mustering the budget to keep him because he was so productive. 
 
It Can't Replace What You Don't Do 
 
There is an even better reason why automation will never replace you: it automates the tests you aren't doing, not the ones you are
 
Let's be honest...one can barely keep up with the rate of change in most systems, testing fixes and enhancements- sometimes over and over – for each release. There is no way you can manually execute all of the tests from prior releases. It's basic math: the inventory of application functionality and the portfolio of applications is constantly on the rise, while your staff and schedule remains flat or shrinks. There's a reason that the most pressing quality problem for most companies is regression testing...or the lack thereof. 
 
Automation is your only hope, not a threat. By using automation for regression testing, it frees your mind and hands from repeating the past so you can focus on the present and plan for the future. Your value goes up as your test coverage expands and product quality improves. You are free to spend more time with the business, understanding their needs and translating them into test requirements and cases, which enhances your contribution and exposure even further. 
 
And even when all of your regression tests become automated – if it ever happens – there is still the ever-present maintenance overhead. Changes usually affect both new and existing functionality. The test library is a living asset that must keep pace with the business and the application, which requires consistent care and feeding to stay viable and valuable. After all, the reason we test software is because something has changed. 
 
Less Staff, More Work 
 
Now I won't deny that some managers believe they can slash their testing staff with tools, but those managers are becoming more and more rare. In most cases staff reductions have already happened anyway, leaving the same amount of work but fewer people. Managers are not trying to figure out how to reduce staff even further; they are trying to figure out how to get the job done with what they have. 
 
In fact, I can count on one finger the number of cases I know of where automation reduced the testing resources significantly. It was a medical products company where their test coverage was FDA mandated and regulated, so they had no choice but to execute tests every time. As a result, they had thirty full-time testers. Automation reduced the full-time testing staff to six. What is interesting is that the other twenty-four testers were microbiologists, and once they were freed from manual execution, they weren't laid off. They were able to return to their real job, which was to work on product requirements, design, and specifications. 
 
I have seen automation save massive amounts of time – converting weeks of manual labor into days or hours of automated execution. The net result was that it enabled the team to meet the schedule with a stable solution instead of cutting corners and performing triage on test cases. Making the schedule while meeting coverage goals means you get brownie points, not a pink slip. 
 
Let's Get the Facts 
 
Obviously, I am an automation fanatic. So you might think I am biased, and you are probably right. But don't take my word for it. I'll just ask, "Who among you has been replaced by a test tool?"


About the Author
Linda G. Hayes, B.B.A., C.P.A., M.S., J.D. has twenty years of experience in software development, is a frequently published author and highly rated speaker on software quality and test automation. As co-founder of AutoTester, Inc., a leading automated testing software vendor, she pioneered structured software test automation. Her article on integrating automated testing throughout the software development cycle won the Most Significant Contribution of the Year award from the Quality Assurance Institute and was published by Auerbach in the testing chapter of their Systems Development Handbook. Linda can be reached at linda@worksoft.com.

Back to Top
 
 

Member Comments
Add Your CommentExpand Comments
 
Comment:    
by David Ball 8/17/2004

A good column. It seems to me that peeling back the onion reveals another truth about the best automation authors. I would submit that the best automation requires careful test engineering first. An excellent suite of manual tests is more likely to produce valuable automation than ad hoc recordings. The best testers understand the value of time spent with a yellow legal pad before double-clicking on the automation icon.

 
 
Comment:    
by Edward Marsh 8/13/2004

Automated testing is the holy grail of testing. However, after learning and setting up an automated test suite with mercury's win runner, trying to find another job as a test contractor to do win runner programming was next to impossible. Automated testing seems to be a task that is very specific and is not something that seems to be in market demand. The bulk of test market is not automated testing-- if fact very few companies have the dollars and necessary expertise for automated testing. Automated testing has it's place, but most testing is done the old fashioned manual way. That has been my 10 year experince.

 
 
Comment:    
by Oralia Ortiz 8/12/2004

Linda, Good article. I was laid off after 5 years of very intesive manual testing. Not having a trained as a developer, I am having difficulty finding a job without the automation experience. I would like to take programming courses but all the scripting languages seem unique to the various tools out there and traing in those tools is very expensive. Any suggestiions on where to start?

Author's Response:
8/12/2004    
Hi Oralia - most of the proprietary scripting tools are converging on VB and VB Script languages, which are standard. You can pick up VB Script or VB for very little money and/or take courses at almost any college that are also cheap. With these skills you could work with almost all of the next generation of tools. Good luck!

 
 
Comment:    
by Ben Goh 8/11/2004

Linda, another excellent article. I find test automation skills are definitely marketable. Just take a look at the testing job advertisments on newpaper and jobs internet sites. The more I do manual testing the more I find my testing skills obsolete, thus unemployable.

Author's Response:
8/12/2004    
Hi Ben - I agree that automation is a valuable skill, but manual testing can also be if you develop deep business or application expertise. I have talked to many companies who are desperate to find people who understand their business, since that may be more arcane than automation.

 
 
Comment:    
by Ashish Mehta 8/10/2004

Impressive!! When it comes to switch over from manual testing to automation, the tester is always confused, reason - Fear of change. But, it cannot be denied that 'Change always comes bearing gifts'. Absolutely, automation does not replace the manual testing job. It just helps to save time by taking the ownership of the repetitive job from the tester to the tool. In turn, this helps the tester to concentrate more the quality of test cases & down the line - a better work product.

Author's Response:
8/11/2004    
Hi Ashish - amen. Once testers realize that if they can let go of they keyboard and engage their brains, everyone wins!

 
 
Comment:    
by G Lee 8/10/2004

Excellent article, Linda. We are just about to introduce anautomation testing tool and we do not foresee a reduction in testing resources. We are fortunate to have in our testing team, two test analyst that a total of over 20 years software developement experience and with the introduction of automation testing we are planning to utilise their development skills in creating testing scripts.

Author's Response:
8/11/2004    
Hi G - it's a pleasure to hear about a company with realistic expectations! It's better to use automation to get more done than less.

 
 
Comment:    
by John Daughety 8/10/2004

Two things: I agree with the comment above that talks about marketability of test automation skills. I think it can literally save your job when you are in a company whose dominant focus is programming - these people do not respect non-programmers, and even though test automation may not be C++ coding, it can be enough to make you valued. In my current job, we get a lot of new features each release and have two weeks to test them - an impossible task. If I were to automate all of our regression tests, I would still need more manual testers than I have now. For each "symptom" an automated test finds, a manual tester can spend up to two...Read On

Author's Response:
8/11/2004    
Hi John - exactly. You can't skip the "brainless" stuff because it usually is about what we know MUST work. By automating this, you're right - you can focus on the hard stuff.

 
 
Comment:    
by Lizette Grobler 8/10/2004

Hi Linda, I agree with you 100%. However, I am a manual tester and I cannot find the time to work through all the hundreds of test tools out there in order to purchase one. We have a budget for it, but going through trial versions and sales pitches are just too time consuming. We do web application testing (asp, xslt, html, stored procedures), any suggestions?

Author's Response:
8/11/2004    
Hi Lizette - I would be glad to offer suggestions but need to ask a few more questions. Email me at linda@worksoft.com and I'll do my best!

 
 
Comment:    
by Kamran Mirza 8/10/2004

Hi Linda, I am in SQA filed since more than six years now writing from Islamabad Pakistan, and totally agree with you on the point that Automation testing does not reduces the workforce. In my personal experience in some cases it actually added a resource or two in the project. The myth is that the people overlook the maintenance of automation testing scripts. You are absolutely right in saying that it’s the hand behind the Gun which fires not the Gun itself. In my opinion we are at the same stage where we were in the early days of office automation when people feared about their jobs, where as time has proved that yes lay offs would be done...Read On

Author's Response:
8/10/2004    
Hi Kim - wow, Pakistan? Impressive. I like your point that jobs aren't lost, only upgraded!

 
 
Comment:    
by Ronald Goodwin 8/10/2004

Well stated Linda. We are currently collecting metrics on ROI for automation. We find that we are saving in the high K's and low M's for each test we automate. But we have not seen any layoffs. If anything, we are spending some of that savings educating our people on how to automate. We are making our own even more indespensible.

Author's Response:
8/10/2004    
Hi Ron - I'd be thrilled to be included if you are allowed to publish your ROI metrics. That's an area that many teams new to automation struggle with. They think that if they can't reduce headcount, they can't show a return. Any education to the contrary would be invaluable.

 
 
Comment:    
by Gilles Girard 8/10/2004

Greetings Linda !! I agree with alot of the things stated in the article: Assuming that you are a programmer doing testing on the side or a person dedicated full time to test automation. If you're a business analyst creating test cases and doing validation or a data entry clerk typing in the test cases or a manual tester (not an automation specialist)...I see things more like the Webster's New Universal Unbridged Dictionnary - 1983 edition definiton of automation: Any system or method that uses self operating equipment, electronic devices, etc. to replace human beings in doing routine or repetitive work; as automation in...Read On

Author's Response:
8/10/2004    
Hi Gilles - you'll be pleased to know that the latest generation of test automation tools are easily accessible to business analysts and manual testers. In fact, they allow technical resources to empower non-technical ones, thus increasing coverage overall. Example: one company had a team of five people automating a single application. By moving to an advanced architecture, those five people now support 22 applications; the business analysts build the test cases, the automators maintain the framework.

 
 
Comment:    
by Darryl Hurmi 8/10/2004

Linda I liked your article and you are right, a good automation programmer will never lose their job. We started with automation 10 years ago and the fist think we found was that we needed more people not less, you can't stop testing while you develop the automation. In the end we found were auyomation worked well and were it didn't but we didn't lose any testers. Most of the testers now had time to work more closely with the programmers on current code rather than spend weeks or months writing tests for the final product. This has lead to a complimentary relationship between the programmers and the testers and an increase in the number...Read On

Author's Response:
8/10/2004    
Hi Darryl - glad you agree! But I also think that even manual testers need not fear. As mentioned above, the latest automation solutions empower non-programmers to automate and allow programmers to support a much wider set of applications.

 
 
Comment:    
by Prasad Dange 8/10/2004

This is really a good article. I found it very helpful in getting my doubts clear about the Automation, its usage and impact over the product testing cycle.

Author's Response:
8/10/2004    
Hi Prasad - thanks, I'm glad it was helpful. See my other StickyMinds column, Does Automation Save Time and Money? for more information.

 
 
Comment:    
by Jim Hazen 8/9/2004

Linda, Totally agree with you and I think you're right on the money, as the following quote puts it. "Both literature and test automation are only as good as the writer, so it's your expertise that counts – not the tool." But I think you missed one key element of the Automation myth, what the Tool Salesman says. I have found too many times the sales people saying that automation will allow for the staff reduction (among other things), and that "Automation is Automagic". This is what I believe fuels the "fear" by manual testers (and some automation people) of being replaced by the tool. Also, this causes the Management and C-Level...Read On

Author's Response:
8/9/2004    
Hi Jim - I love an educated reader! I could not agree more - false expectations kill more projects than anything else. You might check out my past StickyMinds column "Does Automation Save Time and Money?" for a similar theme. I can tell you that I routinely debunk this nonsense whenever I can. If management want to cut the time and cost of testing, they should just test less and take the consequences. Automation is about testing more!

 
 
Comment:    
by Linda Hayes 8/9/2004

This is for Anne - you might also check into some of the new, cheaper test tools that cost 10% of the usual suspects. Email me for more information.

 
 
Comment:    
by Jim Dougherty 8/9/2004

Excellent Linda - At my former employer we leveraged automation to its fullest. Our initial efforts with one of our flagship web products reduced the test cycle effort from 99 staff days to 24 staff days. As to losing positions - as we reduced the size of our automation team, those resources were dispersed to the test groups and became sources of automation expertise within those groups. Jim Dougherty

Author's Response:
8/9/2004    
Hi Jim - always good to hear from someone who has actually delivered on automation in a challenging environment! Were any manual testers displaced by your efforts?

 
 
Comment:    
by Anne Turner 8/9/2004

My company has made noises about automation in the past, but they don't seem to want to pay for the training it will require. We also have serious licensing problems: three testers, one license. How can we get around this?

Author's Response:
8/9/2004    
Hi Anne - it may make you feel better to know this is a classic problem: companies want the benefits of automation but they don't want the cost. As to your 3:1 ratio, I recommend you build (or buy) a framework that lets you develop the test cases externally - say, in Excel - then submit to a single system for execution.

 
 
Comment:    
by Chris Hinchliffe 8/9/2004

Another excellent article, Linda. There are two other attributes of test automation that are worth mentioning. The first is, if you are laid off, test automation is a highly marketable skill and will make you more attractive in the jobs market. The second is that automation is fun (or at the very least interesting). As testers we like to test new stuff, rather than spending weeks running regression suites, maintaining a suite of automation scripts is quicker and more enjoyable than manually running the scripts yourself and it gives you time to test the new functionality in the system.

Author's Response:
8/9/2004    
Hi Chris - both good points. For those manual testers who are afraid automation will displace them because they don't have coding skills, your second point is right on. They can hone their analytical skills instead of doing mindless repetition.

 
Back to Top



 
Ads By Google
What's This?
 
 



About Us   |   Contact Us   |   Terms & Conditions   |   Privacy Policy   |   RSS Feed



© 2013 StickyMinds.com. All rights reserved.
ASTQB

Tricentis



STARWEST