Page 248 - Lean six sigma demystified
P. 248

226        Lean Six Sigma  DemystifieD


                        and trouble calls in to the customer care center. From two sigma, the company
                        moved to three sigma in 3 months.

                          •   From 17% error rate reduced to 3% in just 6 months
                          •   100% elimination of top five error buckets
                          •   $299,000 per month in savings

                          Cost

                          •   Three days of planning
                          •   Six half-day team meetings

                          •   Two minor software releases


                 Common Problems


                        The core elements of any application involve searching for and then adding,
                        changing, or deleting data. Most applications assume a perfect world, where the
                        data is only created or modified by the system. This is rarely the case. Most
                        systems have a wealth of backdoors used to fix faulty data quickly. Upstream
                        and downstream systems have their own backdoors to fix faulty transactions,
                        so perfect data continues to be a mythological assumption that fosters faulty
                        requirements and designs.
                          The requirements for adding, changing, and deleting data are often too loose,
                        too tight, or nonexistent; this leads to errors and rejected transactions which
                        must be corrected manually by people hunched over computer terminals for
                        8 hours a day. The simplest way to fix these IT errors is by using the Dirty

                        30 process for Six Sigma software.


                 The Dirty 30 Process review



                          1. Pick one top error category.
                          2. Get 100 to 200 specific errors (e.g., in this case by telephone number).
                          3. Investigate each error in your online systems to find the root cause of each
                            individual error.
                          4. Grow a check sheet of causes.

                          5. By 30 a root cause pattern will pop out.
   243   244   245   246   247   248   249   250   251   252   253