Loading ...
Sorry, an error occurred while loading the content.
 

Re: [XP] Six Sigma and Extreme?

Expand Messages
  • Bob.Jarvis@chase.com
    We used six sigma as the project framework for our XP implementation (at JPMorgan Chase) with excellent results. Six sigma forced us to really understand what
    Message 1 of 10 , Jul 26, 2004
      We used six sigma as the project framework for our XP implementation (at
      JPMorgan Chase) with excellent results. Six sigma forced us to really
      understand what was important to the customers (better, cheaper, faster,
      quality of life), to establish baseline metrics against those factors, and
      finally allowed us to demonstrate that we achieved significant improvement
      on all fronts.

      Without it, we would have achieved the same results, but would not have
      been able to compare to baseline results, and would have had a much weaker
      argument to expand XP's implementation.

      As Deming said, "In God we trust. All others must provide data." The data
      we gathered because of six sigma have been invaluable in making the case
      for additional XP projects.

      BJ



      "nwaston2003"
      <nwaston@taylor-h To: extremeprogramming@yahoogroups.com
      obson.com> cc:
      Subject: [XP] SixSigma and Extreme?
      07/26/2004 06:16
      AM
      Please respond to
      extremeprogrammin
      g






      We are trying to becoming more 'agile' and adopt various extreme
      programming techniques in our software development. We have 'agile'
      process for software development which integrates into a set of
      engineering processes. Our new buyers, Ametek, are encouraging us to
      use the SixSigma process. This seems to be a heavyweight process
      that is at odds with agile development.
      Have you any comments on SixSigma and extreme programming processes?



      To Post a message, send it to: extremeprogramming@...

      To Unsubscribe, send a blank message to:
      extremeprogramming-unsubscribe@...

      ad-free courtesy of objectmentor.com
      Yahoo! Groups Links
    • Ron Jeffries
      ... Can you tell us a bit (or a lot) about the data you collected, how you collected it, compared it, patted it pricked it marked it with a D? Thanks, Ron
      Message 2 of 10 , Jul 26, 2004
        On Monday, July 26, 2004, at 7:06:42 PM, Bob.Jarvis@... wrote:

        > As Deming said, "In God we trust. All others must provide data." The data
        > we gathered because of six sigma have been invaluable in making the case
        > for additional XP projects.

        Can you tell us a bit (or a lot) about the data you collected, how you
        collected it, compared it, patted it pricked it marked it with a D?

        Thanks,

        Ron Jeffries
        www.XProgramming.com
        A firkin is 9 imperial gallons, or 40.905 liters, or
        81.81 half-liter cans, allowing for 5.8 cans per day over two weeks.
        Therefore, a firkin per fortnight is a six-pack per day.
      • Bob.Jarvis@chase.com
        Ron, Sorry bout the delayed response ... it s been a bit hectic. I ran the project as a Design for Six Sigma (DFSS) project, since we were not improving an
        Message 3 of 10 , Jul 30, 2004
          Ron,

          Sorry 'bout the delayed response ... it's been a bit hectic.

          I ran the project as a Design for Six Sigma (DFSS) project, since we were
          not improving an existing process but rather (essentially) building a new
          one. We gathered "Voice of the Customer" (VOC) / "Voice of the Business"
          (VOB) / "Voice of the Employee" (VOE) data through open-ended questionnaire
          techniques, to determine what was really important to our customers /
          management / developers. We then used the VOx data to derive specific
          quality factors (called "Critical to Quality", or CTQs), and from those
          defined CTQ measures (quantifiable measures that will link back to the VOx
          data). We then used Quality Function Deployment (QFD) techniques to rank
          the CTQs and CTQ measures. The specifics of the above techniques are well
          beyond the scope of this discussion, but the results are not.

          The top CTQ measures were # total defects, % committed functionality
          delivered, # "working as designed" defects, #days duration and $. We used
          new functionality QA test cases as our size measure (denominator). One
          could argue that test cases are not the best measure of size / complexity /
          etc., but we don't use function points or any other standard measure of
          size, and our QA group has a very consistent approach to the system /
          integration tests. Not a perfect measure, but the best we had.

          For comparison, we had six XP releases in two major functional areas to our
          Chase eBanking web site, and we compared them to four similar sized
          releases to other functional areas built with traditional (CMM level 2)
          development processes. The average of the six XP versus the four non-XP
          releases showed the following:

          o Total defects - down 63%
          o Critical defects - down 79%
          o "Working as Designed" - down 38%
          o Duration - reduced 44%
          o Effort - reduced 47%

          So we did demonstrably get better software cheaper & faster.

          Another interesting result ... we used QFD House 2 techniques to evaluate
          the impact of the 12 practices on the CTQ measures, and the clear "winner"
          was on-site customer. The others that scored at the 50% level or better
          (compared to on-site customer) were (in descending order) pair programming,
          planning game, testing, and small releases.

          One last point. We also did a satisfaction survey for both the developers
          and the business partners, and asked them questions on 9 aspects of their
          jobs. They compared life on an XP project versus pre-XP experiences, and
          the business indicated 81% better / much better, and the developers showed
          77% (as opposed to much worse / worse / same).

          Net result - we have quantifiable results that speak to real business
          concerns, and rather easily cut through the noise (on both sides of the
          discussion) that XP usually fosters.

          As I said before, we were about to train the next project team, and our
          merger with Bank One was announced. I expect to use XP for one of the
          merger projects that I'm responsible for, to start making the case for our
          new senior management team.

          I hope that helped.

          BJ



          Ron Jeffries
          <ronjeffries@XProgra To: extremeprogramming@yahoogroups.com
          mming.com> cc:
          Subject: Re: [XP] Six Sigma and Extreme?
          07/26/2004 08:38 PM
          Please respond to
          extremeprogramming






          On Monday, July 26, 2004, at 7:06:42 PM, Bob.Jarvis@... wrote:

          > As Deming said, "In God we trust. All others must provide data." The
          data
          > we gathered because of six sigma have been invaluable in making the case
          > for additional XP projects.

          Can you tell us a bit (or a lot) about the data you collected, how you
          collected it, compared it, patted it pricked it marked it with a D?

          Thanks,

          Ron Jeffries
          www.XProgramming.com
          A firkin is 9 imperial gallons, or 40.905 liters, or
          81.81 half-liter cans, allowing for 5.8 cans per day over two weeks.
          Therefore, a firkin per fortnight is a six-pack per day.




          To Post a message, send it to: extremeprogramming@...

          To Unsubscribe, send a blank message to:
          extremeprogramming-unsubscribe@...

          ad-free courtesy of objectmentor.com
          Yahoo! Groups Links
        • Ron Jeffries
          ... Fascinating! Is there going to be an article about this? Sure should be! Ron Jeffries www.XProgramming.com The lyf so short, the craft so long to lerne.
          Message 4 of 10 , Jul 30, 2004
            On Friday, July 30, 2004, at 9:00:05 PM, Bob.Jarvis@... wrote:

            > Net result - we have quantifiable results that speak to real business
            > concerns, and rather easily cut through the noise (on both sides of the
            > discussion) that XP usually fosters.

            Fascinating! Is there going to be an article about this? Sure should be!

            Ron Jeffries
            www.XProgramming.com
            The lyf so short, the craft so long to lerne. -- Geoffrey Chaucer
          • Steven Gordon
            Bob, I hope you do not mind a little devil s advocacy: Your brief summary could be interpreted to mean that after doing these surveys, you defined requirements
            Message 5 of 10 , Jul 30, 2004
              Bob,

              I hope you do not mind a little devil's advocacy:

              Your brief summary could be interpreted to mean that after doing these
              surveys, you defined requirements in the form of QA tests and CTQ
              measures, and then went to work implementing these requirements, only
              delivering software to the customer when you were done.

              Assuming that you did iteratively deliver working software, obtain
              feedback, and change course according to the feedback during these
              projects, does changing course iteratively based on customer feedback
              fit explicitly into the DFSS methodology? Do you measure against your
              final requirements or your initial requirements? If you measure against
              the final, evolved requirements, one could argue that your metrics
              should always be better than projects that do not change their goals
              because over time your goals will grow closer to what you have actually
              done.

              Steven Gordon
              http://sf.asu.edu/



              -----Original Message-----
              From: Bob.Jarvis@... [mailto:Bob.Jarvis@...]
              Sent: Fri 7/30/2004 6:00 PM
              To: extremeprogramming@yahoogroups.com
              Cc:
              Subject: Re: [XP] Six Sigma and Extreme?


              Ron,

              Sorry 'bout the delayed response ... it's been a bit hectic.

              I ran the project as a Design for Six Sigma (DFSS) project, since we were
              not improving an existing process but rather (essentially) building a new
              one. We gathered "Voice of the Customer" (VOC) / "Voice of the Business"
              (VOB) / "Voice of the Employee" (VOE) data through open-ended questionnaire
              techniques, to determine what was really important to our customers /
              management / developers. We then used the VOx data to derive specific
              quality factors (called "Critical to Quality", or CTQs), and from those
              defined CTQ measures (quantifiable measures that will link back to the VOx
              data). We then used Quality Function Deployment (QFD) techniques to rank
              the CTQs and CTQ measures. The specifics of the above techniques are well
              beyond the scope of this discussion, but the results are not.

              The top CTQ measures were # total defects, % committed functionality
              delivered, # "working as designed" defects, #days duration and $. We used
              new functionality QA test cases as our size measure (denominator). One
              could argue that test cases are not the best measure of size / complexity /
              etc., but we don't use function points or any other standard measure of
              size, and our QA group has a very consistent approach to the system /
              integration tests. Not a perfect measure, but the best we had.

              For comparison, we had six XP releases in two major functional areas to our
              Chase eBanking web site, and we compared them to four similar sized
              releases to other functional areas built with traditional (CMM level 2)
              development processes. The average of the six XP versus the four non-XP
              releases showed the following:

              o Total defects - down 63%
              o Critical defects - down 79%
              o "Working as Designed" - down 38%
              o Duration - reduced 44%
              o Effort - reduced 47%

              So we did demonstrably get better software cheaper & faster.

              Another interesting result ... we used QFD House 2 techniques to evaluate
              the impact of the 12 practices on the CTQ measures, and the clear "winner"
              was on-site customer. The others that scored at the 50% level or better
              (compared to on-site customer) were (in descending order) pair programming,
              planning game, testing, and small releases.

              One last point. We also did a satisfaction survey for both the developers
              and the business partners, and asked them questions on 9 aspects of their
              jobs. They compared life on an XP project versus pre-XP experiences, and
              the business indicated 81% better / much better, and the developers showed
              77% (as opposed to much worse / worse / same).

              Net result - we have quantifiable results that speak to real business
              concerns, and rather easily cut through the noise (on both sides of the
              discussion) that XP usually fosters.

              As I said before, we were about to train the next project team, and our
              merger with Bank One was announced. I expect to use XP for one of the
              merger projects that I'm responsible for, to start making the case for our
              new senior management team.

              I hope that helped.

              BJ










              [Non-text portions of this message have been removed]
            • Adam Wildavsky
              ... An article would be grand. In the meantime you may want to have a look at these slides from a presentation Bob gave to the NYC Agile SIG. The presentation
              Message 6 of 10 , Jul 30, 2004
                At 10:20 PM -0400 7/30/04, Ron Jeffries wrote:
                >Fascinating! Is there going to be an article about this? Sure should be!

                An article would be grand. In the meantime you may want to have a
                look at these slides from a presentation Bob gave to the NYC Agile
                SIG. The presentation bowled me over -- the slides were not meant to
                stand on their own and do not do it justice.

                http://www.nycagile.org/index_files/res/Bob_Jarvis.pdf

                --
                Adam Wildavsky Extreme Programmer Tameware, LLC
                adam@... http://www.tameware.com
              • Steven Gordon
                Thanks! Very impressive. But, I did not catch a single mention of Six Sigma. As pointed out in the presentation, XP is effective, but not optimized. By
                Message 7 of 10 , Jul 31, 2004
                  Thanks!

                  Very impressive. But, I did not catch a single mention of Six Sigma.
                  As pointed out in the presentation, XP is effective, but not optimized.
                  By extension, optimizing XP would kill its effectiveness on the next
                  project.

                  -----Original Message-----
                  From: Adam Wildavsky [mailto:adam@...]
                  Sent: Fri 7/30/2004 9:09 PM
                  To: Ron Jeffries
                  Cc: extremeprogramming@yahoogroups.com
                  Subject: Re: [XP] Six Sigma and Extreme?



                  At 10:20 PM -0400 7/30/04, Ron Jeffries wrote:
                  >Fascinating! Is there going to be an article about this? Sure should be!

                  An article would be grand. In the meantime you may want to have a
                  look at these slides from a presentation Bob gave to the NYC Agile
                  SIG. The presentation bowled me over -- the slides were not meant to
                  stand on their own and do not do it justice.

                  http://www.nycagile.org/index_files/res/Bob_Jarvis.pdf

                  --
                  Adam Wildavsky Extreme Programmer Tameware, LLC
                  adam@... http://www.tameware.com


                  To Post a message, send it to: extremeprogramming@...

                  To Unsubscribe, send a blank message to: extremeprogramming-unsubscribe@...

                  ad-free courtesy of objectmentor.com
                  Yahoo! Groups Links









                  [Non-text portions of this message have been removed]
                • Bob.Jarvis@chase.com
                  I ve given several presentations (local to NYC), but haven t published anything. Any suggestions? BJ Ron Jeffries
                  Message 8 of 10 , Jul 31, 2004
                    I've given several presentations (local to NYC), but haven't published
                    anything. Any suggestions?

                    BJ



                    Ron Jeffries
                    <ronjeffries@XProgra To: extremeprogramming@yahoogroups.com
                    mming.com> cc:
                    Subject: Re: [XP] Six Sigma and Extreme?
                    07/30/2004 10:20 PM
                    Please respond to
                    extremeprogramming






                    On Friday, July 30, 2004, at 9:00:05 PM, Bob.Jarvis@... wrote:

                    > Net result - we have quantifiable results that speak to real business
                    > concerns, and rather easily cut through the noise (on both sides of the
                    > discussion) that XP usually fosters.

                    Fascinating! Is there going to be an article about this? Sure should be!

                    Ron Jeffries
                    www.XProgramming.com
                    The lyf so short, the craft so long to lerne. -- Geoffrey Chaucer




                    To Post a message, send it to: extremeprogramming@...

                    To Unsubscribe, send a blank message to:
                    extremeprogramming-unsubscribe@...

                    ad-free courtesy of objectmentor.com
                    Yahoo! Groups Links
                  • Ron Jeffries
                    ... You mean, other than publish something ? IEEE Software might have the right mix of rigor and popular reading? Or perhaps in a respected Six Sigma or PM
                    Message 9 of 10 , Jul 31, 2004
                      On Saturday, July 31, 2004, at 6:11:30 PM, Bob.Jarvis@... wrote:

                      > I've given several presentations (local to NYC), but haven't published
                      > anything. Any suggestions?

                      You mean, other than "publish something"? IEEE Software might have the
                      right mix of rigor and popular reading?

                      Or perhaps in a respected Six Sigma or PM journal, and then we could all
                      refer to it and bask in the reflected glory of that journal?

                      Ron Jeffries
                      www.XProgramming.com
                      This is how I develop software.
                      Take the parts that make sense to you.
                      Ignore the rest.
                    • Bob.Jarvis@chase.com
                      Another presentation that included more of the six sigma (& CMMI) aspects ... http://nycspin.org/Resources/XPStory.pdf BJ Adam Wildavsky
                      Message 10 of 10 , Jul 31, 2004
                        Another presentation that included more of the six sigma (& CMMI) aspects
                        ...

                        http://nycspin.org/Resources/XPStory.pdf

                        BJ



                        Adam Wildavsky
                        <adam@tameware To: Ron Jeffries <ronjeffries@...>
                        .com> cc: extremeprogramming@yahoogroups.com
                        Subject: Re: [XP] Six Sigma and Extreme?
                        07/31/2004
                        12:09 AM
                        Please respond
                        to
                        extremeprogram
                        ming






                        At 10:20 PM -0400 7/30/04, Ron Jeffries wrote:
                        >Fascinating! Is there going to be an article about this? Sure should be!

                        An article would be grand. In the meantime you may want to have a
                        look at these slides from a presentation Bob gave to the NYC Agile
                        SIG. The presentation bowled me over -- the slides were not meant to
                        stand on their own and do not do it justice.

                        http://www.nycagile.org/index_files/res/Bob_Jarvis.pdf

                        --
                        Adam Wildavsky Extreme Programmer Tameware, LLC
                        adam@... http://www.tameware.com


                        To Post a message, send it to: extremeprogramming@...

                        To Unsubscribe, send a blank message to:
                        extremeprogramming-unsubscribe@...

                        ad-free courtesy of objectmentor.com
                        Yahoo! Groups Links
                      Your message has been successfully submitted and would be delivered to recipients shortly.