W13_AK_Pareto Chart and Matrix Diagram method

1.Problem Definition

My colleague in charge to support implementation one type of VAS features in our operator client. From daily report we find out that rate for failed called for this features is quiet high. So that’s why we try to collect all data and try to analyze the performance of this features and try to find the main root cause so we can prepare the solutions and make this VAS features performance become better.

2.Feasible alternatives

From our monthly report, here what we do step by step :

–          Using check sheet, we group and calculate amount of  each error cause

–          By pareto , we will find the biggest error cause and this is become our “Root Cause Problem”

–          After we have the root cause, then we formulate options to decrease the root cause

–          By Matrix Diagram we choose the best option to increase the performance

Here is the error cause data that we have base on our monthly report :



3. Develop the outcomes

By pareto method, we can calculate percentage cumulative for each error cause, so we can know which error cause that become main “Root Cause”. Then we can take appropriate action to solve and reduce the error and increasing this VAS features performance.



4. Select Acceptable Criteria

From pareto chart we can see that the first 5 causes can cover 95.4 % from total defects. So that’s why we focus to solve and decrease this 5 big causes. Here the option that we can take to minimize it :

  • Opt 1 : Add A party threshold limit
  • Opt 2 : Increase storage capacity for the VAS server
  • Opt 3 :Add new VAS server , so we can configure cluster system and do load balance and create better performance
  • Opt 4 : Upgrade OS to more stable version
  • Opt 5 : Add network capacity backbone

5.  Compare the outcome

We used Matrix Diagram to evaluate each option influences for the error caused and choose the best option to increase the VAS Features Performance.


6. Best criteria from mimimum

From step 4 and step 5 , we can do summarize :

We can see form total valued that the best option is for Option 1 (Add A party threshold) and Option 2 (Increase storage capacity) is the second best option. Beside that for Option 1 and Option 2, they solve for Error cause 1 and Error Cause 2 which is impact to 81,8% for all system defect. Which means this two error cause is the big issue for the VAS system performance.

Conclusion : the best option is option 1

7. Performance monitoring and post result

For performance monitoring and post result , the parameter that we can used for performance is the error cause rate for error cause 1 and error cause 2. If in the future after we implement the solution and the error cause become decrease, then the VAS performance become better.


When to Use a Matrix Diagram, retrieved from https://www.qualityamerica.com/knowledgecenter/qualityimprovementtools/when_to_use_a_matrix_diagram.asp, web 14 March 2013.

Cheung, Laurie A., Introduction to Matrix Diagram(2001), retrieved from http://www.freequality.org/documents/knowledge/matrix%20diagram2001.pdf , web 14 March 2013.

Tutorial:Frequency Counts and Pareto Chart, retrieved from http://wiki.originlab.com/~originla/howto/index.php?title=Tutorial:Frequency_Counts_and_Pareto_Chart, web 14 March 2013.

This entry was posted in AndiKur. Bookmark the permalink.

4 Responses to W13_AK_Pareto Chart and Matrix Diagram method

  1. drpdg says:

    AWESOME case study, Pak Andi and you did a nice job on the analysis, with one exception. The Pareto Analysis does NOT provide you with a Root Cause Analysis. The Pareto Analysis (which you did a great job on!!) only PRIORITIZES which are the problems you need to focus your attention on.

    The next step AFTER you do the Pareto Analysis is to THEN determine what the root cause was/is.

    Here are some root cause analysis tools for you to consider:



    Once you have done your root cause analysis THEN you can identify what INTERVENTIONS (process improvements) you can make. After you make those changes to the process, you measure again to see if they have been effective or not. (see Memory Jogger 2, page 128 for an example)

    Keep up the good work and looking forward to seeing more blog postings on this case study. It is a really good one!!

    Dr. PDG, Jakarta

  2. Pingback: W3_WIRA_Root Cause Failure Analysis | TOPAZ SMART

  3. Pingback: W5_HP_Pareto Chart for Analyzing Customer Satisfaction of Project Inspection | TOPAZ SMART

  4. Pingback: W6_HP_Root Cause Analysis | TOPAZ SMART

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s