Six Sigma in the Software Industry

M

Momo123

Hello everyone!

I was wondering if you could share your opinion on the following debate statement (by doing so you're helping me out with my thesis research project as well):

Six Sigma quality tools should be implemented in the software industry department in spite of the risk of creativity and profitability decrease.

We know that Six Sigma is aiming at customer satisfaction and quality improvement overall, but this whole process might well play the role of a boomerang and result in profitability inefficiency, for example.

Therefore, it would be interesting to know whether or not a software department that focuses on quality and process excellence should impement Six Sigma quality tools.

Thanks in advance!
 

Jen Kirley

Quality and Auditing Expert
Leader
Admin
Re: Six Sigma in the Software Industry?

Welcome to the Cove!

Some people do not believe Six Sigma can fit well in the software industry, but I did a Google search using key words "six sigma software industry" and found a great many sources, including a presentation sponsored by Carnegie Mellon Software Engineering Institute.

The key seems to be discarding the assumption that Six Sigma can only work for large runs of product so as to generate adequate reliable data for the statistics that are common in manufacturing Six Sigma. In fact, Six Sigma methodology does not require specific statistics. I did a Six Sigma project in document control, which is transactional in nature, and linked it in a thread titled Six Sigma Project - The entire Corrective Action Process.

I hope this helps!
 

Steve Prevette

Deming Disciple
Leader
Super Moderator
Re: Six Sigma in the Software Industry?

Six Sigma has historically been a "hot button" topic here at the Cove and other locations. You may want to search through some of the past postings here on the subject.

You will see that I am not a proponent of Six Sigma, and do believe that it stifles innovation and creativity, and there are better alternatives.
 
M

Momo123

Thank you for the valuable input Jennifer and Steve!

@Jennifer: The presentation that you found on the internet turned out to be very useful. It's funny that I've already done quite some research, but never came across it. :) Thanks for sharing your Six Sigma project as well!

@Steve: It is always important to hear some critical comments! Could you mention some of the reasons why you think Six Sigma stifles innovation and creativity? What are the better alternatives that you have in mind (just for the sake of further research)?
 

Steve Prevette

Deming Disciple
Leader
Super Moderator
Here are two good discussion threads here at the Cove on Six Sigma:

https://elsmar.com/Forums/showthread.php?t=1899&highlight=sigma

https://elsmar.com/Forums/showthread.php?t=50738&highlight=sigma

On innovation - a difficulty of Six Sigma is that it only focuses upon "measurable" quantities. A point made by Dr. Deming is that some of the most important things we need to know are unmeasurable, unknowable. Such as love, respect, devotion. In that list I would add - the ability to innovate. Six Sigma is so focused on dollars and the "bottom line" and short term gains, that it is hard to justify research for research sake, with the "eventual" long term gains that will result.
 

Miner

Forum Moderator
Leader
Admin
I respectfully disagree with the notion that Six Sigma stifles creativity. One of the biggest problems with the use of Six Sigma is when companies try to take a "one size fits all" approach to the methods.

There are two very distinct versions DMAIC Six Sigma for post-launch product/process improvement and Design for Six Sigma (DfSS) for pre-launch Design and Development. A good DfSS program includes innovation tools, TRIZ and Axiomatic Design. Pure research for the sake of new knowledge should not be limited.

On the original topic, I have found that DMAIC Six Sigma does work well in the transactional world and DfSS does in the software world, but the methods must be tailored to the environment. Not all of the standard tools work, and you often need new tools that are not often used in the standard Six Sigma. For example, in transactional, you are often dealing with cycle times, which are rarely normally distributed, so you need nonparametric tools to deal with this. In software, a software version of a Design FMEA works well, but the language must be changed so that software engineers can understand.
 

Bev D

Heretical Statistician
Leader
Super Moderator
Having successfully implemented 'six sigma' in a software company I believe it works. we generated new and revised products faster, cheaper and with much higher quality levels. It wasn't easy becuase it required a 180 degree switch in how developers and their managers thought about software development. A major difficulty is that the process does require discipline. there is a huge difference between coming to work and doing a task however you see fit (what most people think is 'creative') and begin diciplined in your approach to the task. this discipline actually 'frees' creativity and innovation by focusing it and feeding with certain knowledge. Think of some of the incredible engineering feats of our time - putting a man on the moon, the atom bomb, etc. highly disciplined yet highly innovative and fast! OK maybe not cheap, but you get the concept.)

In fact our first lesson about six sigma is that it is a PHILOSOPHY at the level of a paradigm: by focusing on the true needs of the Customer and providign a quality product in the correct manner we will achieve breakthrough levels of improvements in quality delivery and cost.

While many hacks and companies practice and present 'six sigma' as a cost cutting program or a powerpoint/minitab shackled process or some other equally atrocious incarnation, these 'popular' approaches are NOT six sigma. Steve is correct that these incarnations of six sigma are to be avoided.


Six sigma does include a plethora of tools and methodologies - none of which is applicable in every situation. Software will require many different tools than (hardware) product design which requires many different tools than Problem Solving, etc.



It is also important to understand that other 'prgrams' have their place side by side with 'six sigma' - none of which is inherently more important than the other. These programs include but are not limited to: Lean (or TPS), project management, a 'Quality Management System', etc.
 
Top Bottom