[fine69]

DFSS (DMADV approach instead of DMAIC) is very much applicable to any new product/service creation project.

Lean relies on waste elimination and most companies wouldn't be able to use it during new product development because they don't think around the 5 lean principles when laying down a new product creation strategy.

For all practical purposes, lean makes more sense once the process starts functioning.

@Trojan: Six Sigma is much more than providing statistical input, sorry but your knowledge on the subject sounds very very limited. There are ample tools in its toolkit to create sophisticated solutions as well.

[Trojan]

Originally Posted by fine69

DFSS (DMADV approach instead of DMAIC) is very much applicable to any new product/service creation project.

Lean relies on waste elimination and most companies wouldn't be able to use it during new product development because they don't think around the 5 lean principles when laying down a new product creation strategy.

For all practical purposes, lean makes more sense once the process starts functioning.

Fine69,
My reply to Samurai was with respect to development projects, if you think Lean is no good so be it, it doesn't affect me.
If you think my knowledge is limited so be it, I am neither here for an endorsement of my skillsets here nor does it matter to me what you think.
You have in your previous post questioned the credibility of a very prestigious institution, very good.

If you have an explanation on implementing six sigma in software development projects go right ahead and talk about it.

The question was not on whether something should be implemented before project starts or while the project is in process, the question was whether six sigma is useful and should be used in development projects

[fine69]

Originally Posted by Trojan

Fine69,
My reply to Samurai was with respect to development projects, if you think Lean is no good so be it, it doesn't affect me.

Please read my post again, I've clearly specified when it would work and when it wouldn't.

Originally Posted by Trojan

If you think my knowledge is limited so be it, I am neither here for an endorsement of my skillsets here nor does it matter to me what you think.

You are providing incorrect information which is why I had to jump in and share it with you all that Six Sigma is not meant for getting statistical inputs only, its way more than that!

Also, Six Sigma is most definitely complex - its not the easiest of the subjects yet is not impossible to learn either. But calling it easy is underrating it a lot. Once you understand Measurement System Analysis (Attribute Gage R&R, Variable Gage R&R), Process Capability, Data Distribution, Hypothesis Testing, Design of Experiments, Process Stability (Control Charts) etc. is when you should ideally make a comment about this subject being easy or not.

Originally Posted by Trojan

You have in your previous post questioned the credibility of a very prestigious institution, very good.

I don't think you've done your Green Belt certification from ISI. Let me share with you why I don't give two hoots about the prestige of this institution.

The final exam paper which is held by ISI is a decade old questionnaire in MS Word format. They email it to all the participants and give them 24 hours to revert with the answers. I mean is this how exams are conducted You might simply skip the exam instead of wasting everybody's time doing this formality. Do I need to say that the answers can be googled!

Then the project submission process, while DMAIC is based on statistical validation of hypothesis, the entire concept of hypothesis validation is not talked about in their report guidelines. There is a difference between a PDCA project and a DMAIC project. They want the project on PDCA guidelines while Six Sigma Green Belt certification is based on DMAIC methodology and one of the primary difference between a PDCA and DMAIC project is the validation of hypothesis through statistical tests.

Let me share some more trivia, this prestigious institution doesn't entertain every BB candidate. They say that they will shortlist the profile based on evaluation of their experience/resume and only those would be called for the BB training who they feel are worth their while.

However, one of my ex-boss calls up their main coordinator (the lady who's a Bengali and my ex-boss is also a Bengali) and this so-called criteria goes for a toss. So much for this prestigious institution! Its like any other govt. office, same level of bureaucracy & same level of service/delivery.

Originally Posted by Trojan

If you have an explanation on implementing six sigma in software development projects go right ahead and talk about it.

People who don't understand Six Sigma are the ones who would want to implement Six Sigma in software development projects. One doesn't implement "Six Sigma". What one would do is adopt the DFSS (5-phase DMADV method) approach and start working on a product development using this methodology. When there's an existing process then DMAIC is adopted but for new product creation DFSS is. Simply saying 'Six Sigma' doesn't mean anything, one needs to know when to adopt DFSS and when to adopt DMAIC.

Its like me telling somebody, buy a Toyota, without telling them which car type and variant I'm talking about.

[Trojan]

Originally Posted by fine69

You said that using just Six Sigma will only give statistical input and not lead to improvement

Hey hey relax there!
You are going way out of league here, before you make further assumptions cool down and read the previous posts all over again.

You like to pick up on certain words and then form your own amazing conclusions on the same, please go ahead.

I can very well reply on each of your statements but I chose not to, you may consider it my professionalism.

[Samurai]

Relax guys, this is not a Dojo. I thought by earning those belts you have complete control over your emotions.

This is why I keep away from process or methodology wars. I just use common sense processes that have worked for 24 years of my career.

[kaushik51094]

Sorry if i sound stupid, but what is this thread all about? Karate? Corporate world? or something else? Someone please enlighten me, a noob here in want of knowledge!

[fine69]

Originally Posted by kaushik51094

Sorry if i sound stupid, but what is this thread all about? Karate? Corporate world? or something else? Someone please enlighten me, a noob here in want of knowledge!

Hahahaha

Samurai confused you I guess, with his belts and control of emotions line!

There's some worldwide recognized certifications (just like you've got CCNP, RHCE, MCSE) which are on the subject of Six Sigma viz. Green Belt, Black Belt and Master Black Belt and that is what the thread is about.

You might hear new colours in many organizations, like White/Yellow Belt also . These are mainly a step below Green Belt but aren't really recognized across the world, more of a morale boosting thing.

[TheBigH]

Well I never knew Six Sigma could become a popular buzzword and a battlefield on T-BHP. I've been and still am a six sigma practitioner with an experience of around 11 odd years. I'll be happy to help in any way I can. Some background. A certified Green Belt and Black Belt and an aspiring MBB as well. Do PM in case someone needs help.

[kaushik51094]

Originally Posted by fine69

Hahahaha

Samurai confused you I guess, with his belts and control of emotions line!

There's some worldwide recognized certifications (just like you've got CCNP, RHCE, MCSE) which are on the subject of Six Sigma viz. Green Belt, Black Belt and Master Black Belt and that is what the thread is about.

You might hear new colours in many organizations, like White/Yellow Belt also . These are mainly a step below Green Belt but aren't really recognized across the world, more of a morale boosting thing.

Oh! Thanks for the info. couldn't find the 'thank' button for some reason. so, thanking you through this post!

[Trojan]

Originally Posted by TheBigH

Well I never knew Six Sigma could become a popular buzzword and a battlefield on T-BHP

Seriously my thoughts as well!

Well good to know you! Never imagined would find so many SSBBs on the forum


Cheers!
Trojan

[Shreyans_Jain]

Originally Posted by fine69

Getting a Black Belt certification wouldn't really help anybody's cause unless they are trying to land up a Black Belt role.


Attending a GB/BB training wouldn't help at all unless your only objective is getting that certificate. The subject is such that unless you practice it regularly, you wouldn't be able to make sense of whatever was taught to you just 2-3 weeks later.

Agreed 100%

The subject is very vast and technical and unless you practically apply what you learn and do so frequently, you will forget it all pretty soon.

BB is an overkill unless you want to specialize in it or if there is a specific requirement for it.

[TheBigH]

Originally Posted by Trojan

Well good to know you! Never imagined would find so many SSBBs on the forum

Good to know you too trojan!

Quite agree. I was rather surprised myself. If anyone is serious about pursuing their six sigma interests though, TQMI has a tie up with Motorola and their training and certification program is best in class and recognized worldwide. A certification from the American Society of Quality (ASQ) is also highly vaunted. Then come institutes like ISI and QAI. ISI may seem archaic from a testing perspective, but their material is fairly decent. The key is to ensure as to who is the instructor / trainer and find out the most about the instructor. The biggest issue is that certification institutes are a dime a dozen and organizations tend to be choosy about who they hire. This is true for me as well. When I hire for my own team, I typically ask my recruitment team to NOT hire from a few institutes since I have found their training and certification suspect. Has the thank you button been removed? Can't seem to see it?

[Samurai]

I have lot more respect for Six Sigma when compared to other methodology cults, mainly because the former has a firm foundation of statistical methods.

But it doesn't help when the experts speak strictly in jargon, and then fight among themselves in jargon like quoted below.

Originally Posted by Trojan

For development projects, Lean and Six Sigma needs to be implemented together. This is necessary since IT projects are dynamic in nature and no two projects may be similar. So, while six sigma provides the statistical input and what we can really aim for, tangible actions and pain areas are identified using lean principles.

Originally Posted by fine69

Lean relies on waste elimination and most companies wouldn't be able to use it during new product development because they don't think around the 5 lean principles when laying down a new product creation strategy.

Originally Posted by fine69

Also, Six Sigma is most definitely complex - its not the easiest of the subjects yet is not impossible to learn either. But calling it easy is underrating it a lot. Once you understand Measurement System Analysis (Attribute Gage R&R, Variable Gage R&R), Process Capability, Data Distribution, Hypothesis Testing, Design of Experiments, Process Stability (Control Charts) etc.

Then the project submission process, while DMAIC is based on statistical validation of hypothesis, the entire concept of hypothesis validation is not talked about in their report guidelines. There is a difference between a PDCA project and a DMAIC project. They want the project on PDCA guidelines while Six Sigma Green Belt certification is based on DMAIC methodology and one of the primary difference between a PDCA and DMAIC project is the validation of hypothesis through statistical tests.

As head of a software product development company, my interest in any efficiency technique is genuine. However, when I can't differentiate between six sigma experts and navajo code talkers, my interest diminishes.

I am not entirely unschooled in statistics, I did take a course in quantitative methods long back during my masters. Still I was unable to understand most of the jargon mentioned by you.

Six sigma was created for manufacturing industry where automation is a reality. The quality and quantity of input, the output, and the time taken at each step are all measurable. Here you have access to real metrics.

In software development, such predictability is non-existent. More so in product development. The input is user requirements, which can come from customer or marketing team, would keep changing all the time. The output can be quite different from the user's subjective expectation especially if UI is involved. Users often expect to see bleeding edge technology instead of mature time tested technology. Recently a potential client complained about outdated web interface on a product that is 3 years old, they wanted HTML5 everywhere. Finally, the quality of people can affect the project dramatically in either direction. Software development heavily relies on people and it is very hard to measure the quality of people. Most companies try to measure it objectively, which results in very flawed data. Design is very important in software development, and every programmer is a designer at some level. How to measure quality of design? It would be like measuring quality of a painting. Can you make objective comparison between a Picasso and a Monet? Obviously not, it takes subjective judgment and it needs an art expert to do that. Similarly, measuring the quality of a programmer takes subjective judgment from a more experienced programmer.

In the 90s, I was witness to ISO9000 and SEI CMM efforts in TCS and HP. Metrics was the key word. We had enthusiastic non-technical QA people constantly measuring our output, collating the results, and presenting colorful reports. We were told that anything that is not measured, simply didn't happen. But honest measurement was not possible because the process was entirely objective, designed by non-technical QA people. So we all got better and better at fudging the metrics to make them happy. I left the corporate world in 98 to join a startup and escaped all the farce. Later when I studied quantitative methods course, I understood what they were trying to do, and was also horrified at how mindlessly it was applied.

Software development is a creative process, especially products. One can have mind-blocks lasting days, followed by sudden spurt of creative ideas, design troubles, knowledge gaps related to new technologies, etc. The productivity of a programmer can be measured only subjectively, by another programmer. Either that or you can fudge the metrics. Therefore, many years back we implemented a measurement sheet with 50 criteria, where each criteria is given a score subjectively by the respective team leader. Any high score in each criteria has to be defended by the team leader to his manager.

Whenever we see deficient score, we identify those weak areas and provide more training/counselling required to raise their score. I believe in raising the consciousness/awareness of the programmer via constant mentoring, than subject them to more and more stringent processes. A self-aware confident programmer is more efficient than a programmer whose metrics are closely monitored. That has been my experience, and I hope I have conveyed my thoughts without any jargon.

[LTAutoMad]

Originally Posted by tsk1979

I think that is overkill. I mean master black belt. Why would you need that. Anyway, most people in corporate world are pretty non physically active guys atleast in India. you can just go for a beginner kickboxing kind of course, and be strong enough to beat up 2-3 of them at a time.

Black belt etc., will require too much investment, and its an overkill. Its like turning up to an indica drag race in a ferrari. All you need is a simple cruze or something like a TSI Polo.

Second part about training. Why would you train others. I mean, if others are also as well trained as you, you are only increasing your competition.

In the corporate work, everybody is competing with everybody. Infact people even compete with themselves. I am told they go to the mirror and try to outrun the reflection to try and beat themselves.

So expecting a black belter who you trained to be in your team always is being naive. I mean if you want to be the most badass guy, make sure you are the only one trained in combat. This is all you need to come out winner in the appraisal meeting.

tsk1979, hahaha! You had me in stitches! When I think that your tongue-in-cheek post will be read by many who have no idea what a six sigma black belt is, I am laughing even more.

[fine69]

Originally Posted by TheBigH

Quite agree. I was rather surprised myself. If anyone is serious about pursuing their six sigma interests though, TQMI has a tie up with Motorola and their training and certification program is best in class and recognized worldwide. A certification from the American Society of Quality (ASQ) is also highly vaunted. Then come institutes like ISI and QAI. ISI may seem archaic from a testing perspective....

I have some comments to make here. When the TQMI folks were explaining Chi-square in their Black Belt class they wanted to conclude the hypothesis by only looking at the p-value. I wouldn't expect anybody to conclude their hypothesis based on p-value when Chi-square and Critical Value comparison is available. As you would know, it has a far more significance when concluding a hypothesis than p-value. No points for guessing, the critical value table was not shared with the class either.

TQMI currently is thriving on the fact that they are the only ones who can provide Motorola certification. In fact, if you go to their website they show-off the certificate because that's what most folks are after.

On the ISI front, they don't touch Minitab in their training. This basically means that a trained Green Belt from ISI has to be taught Minitab usage, hypothesis test conclusions etc. even when they are trained Green Belts. No other institute does their GB training without Minitab because the training should help someone implement the knowledge, for all practical purposes. Also, the statistics that they cover, that's just descriptive and inferential statistics which doesn't really need 3 days worth of class-time, again, my opinion.

To add, ISI professors aren't the most participant friendly folks either. The way some questions were taken up in the class, people stopped asking questions altogether.

To people seeking certification, my advise would be to go get it from anywhere because ultimately knowledge & experience would talk more than the certificate. I've interviewed 100+ Black Belts till date and have seen nincompoops (w.r.t. to Six Sigma knowledge) from the best of organizations & best of institutes while at the same time some very knowledgeable people from roadside institutes. People who've learnt and worked hard on the subject would always stand out, I firmly believe in this.

Originally Posted by Samurai

As head of a software product development company, my interest in any efficiency technique is genuine. However, when I can't differentiate between six sigma experts and navajo code talkers, my interest diminishes.

My response was to someone who would understand what I mentioned which is why I didn't explain it too much. It was very specific to one post so I took the liberty of using some jargons.

Originally Posted by Samurai

Six sigma was created for manufacturing industry where automation is a reality. The quality and quantity of input, the output, and the time taken at each step are all measurable. Here you have access to real metrics.

When I said Yellow/White belts are more of morale boosting things I had a similar logic. Obviously everything cannot be applied to the service industry as it is which is why Six Sigma is being abused in the service industry like crazy. I don't mean that Six Sigma is not applicable, just that its not applicable in a lot of scenarios where it is applied.

As a Black Belt consultant let me openly say that 50% (I'm being very liberal here) of the Six Sigma DMAIC projects are actually a farce because that situation isn't the one where both cause and solution to the problem are unknown.

Originally Posted by Samurai

In software development, such predictability is non-existent. More so in product development. The input is user requirements, which can come from customer or marketing team, would keep changing all the time. The output can be quite different from the user's subjective expectation especially if UI is involved.

Quality Function Deployment exists because of such scenarios only. It is part of Design for Six Sigma approach, translating Voice of Customer in to specific measurable 'Critical to Quality/Delivery/Cost' metrics. The customer being vague is taken care of to a large extent, if not completely.

Originally Posted by Samurai

Software development heavily relies on people and it is very hard to measure the quality of people. Most companies try to measure it objectively, which results in very flawed data.

It is hard, not impossible. Some companies do it very well, some not. When they don't know how to do it well, it most definitely relies in flawed data and a lot of top MNCs are committing this crime day in day out.

Originally Posted by Samurai

Design is very important in software development, and every programmer is a designer at some level. How to measure quality of design? It would be like measuring quality of a painting. Can you make objective comparison between a Picasso and a Monet? Obviously not, it takes subjective judgment and it needs an art expert to do that. Similarly, measuring the quality of a programmer takes subjective judgment from a more experienced programmer.

Yes, quality of a painting can be measured, just that your perspective of quality is probably different from mine. Let me make an attempt at explaining something.

Firstly, there is always a customer to a business. The customer decides what he wants, so lets say I'm the customer to this painting. Now I like a Monet better just because I like the sound of the artist's name. I don't care what he made till the time its a genuine one. I'm ready to pay top dollars for it so does it really matter to you (you're the painting seller) whether Picasso is better or not. The customer decides the quality in a business. It doesn't matter what the art expert says, you as the supplier would want to give the customer what he demands.

Now if my requirements are vague (which happens a lot), a Six Sigma guy helps formulate them into measurable elements using 'Quality Function Deployment'.

Originally Posted by Samurai

In the 90s, I was witness to ISO9000 and SEI CMM efforts in TCS and HP. Metrics was the key word......Later when I studied quantitative methods course, I understood what they were trying to do, and was also horrified at how mindlessly it was applied.

As a QMS (ISO 9001:2008) Lead Auditor let me just say that this certification has been a farce in most companies for a long time now. Until the 2008 version of this standard, most companies would hire a consultant who would do the necessary paperwork required for this certification and the same consultant would also become the 'Management Representative' (this guy is crucial and has a lot of responsibility riding on him). The ISO guys took note of it and made a clause that the MR (Mgmt Representative) has to be someone who is a full-time employee of the company and someone from Top Management. While this didn't made everybody very serious about ISO 9001 but it meant someone from Top Mgmt taking responsibility of implementing QMS (Quality Management System).

I've worked with a Tata group of company where the MR was the function head of Quality Department and I used to report to him. This gentleman ensured that not a single activity around ISO 14001 and OHSAS 18001 certification was done as an eye-wash. So some companies do it diligently, others just do the necessary eye-wash needed. My point, ISO 9001:2008 is a very very robust standard and if you implement this diligently, most companies wouldn't even need Six Sigma for 99% of their problems (I can share the rationale behind me saying this if anybody is interested). Sad part is that it doesn't happen and when a certification body takes up a contract with any MNC they know that if they create too many hassles in certifying them then they'll go to someone else so it is similar to what has happened to the school and hospital industry, its a business now.

Originally Posted by Samurai

Software development is a creative process, especially products. One can have mind-blocks lasting days, followed by sudden spurt of creative ideas, design troubles, knowledge gaps related to new technologies, etc...A self-aware confident programmer is more efficient than a programmer whose metrics are closely monitored. That has been my experience, and I hope I have conveyed my thoughts without any jargon.

Businesses thrived even when Six Sigma wasn't around and there are a lot success stories where Six Sigma is still unheard of. When you use subjectivity it doesn't mean that its bad, it might work for you while it might not work in other scenarios.

Six Sigma (DMAIC) is for breakthrough improvement when a problem is chronic (identified through voice of customer or voice of business) and both the cause and solution are unknown. If such is the case (all conditions met), Six Sigma would work, if this isn't the case, Six Sigma isn't even needed. People might still use it and run fake DMAIC projects (which is happening on an incredible scale) but that's only lowering the value that it carried 10-15 years ago.

For new product creation, Design for Six Sigma is a very robust approach that helps translate subjectivity in to measurable elements. This doesn't mean it would apply everywhere but I've seen it myself in one of the Mahindra group of companies, used in software development and till last year it was one of their strongest selling point during RFP (Request for Proposal) defense presentation.

PS: Yes I've worked for both Tata & Mahindra