3.3 Pareto Analysis & Prioritization

3.3 Pareto Analysis & Prioritization

In quality improvement initiatives, manufacturing organizations face an abundance of potential problems but operate with finite resources, capital, and engineering time. Attempting to resolve every defect simultaneously results in diluted focus, exhausted budgets, and minimal measurable improvement. Pareto Analysis provides quality technicians with an objective, mathematically rigorous method to separate the vital few sources of process failure from the useful many, ensuring that continuous improvement efforts deliver maximum return on investment.

---

The Pareto Principle: Origin and Quality History

The Pareto Principle originated in 1897 when Italian economist Vilfredo Pareto observed that approximately 80% of the land and wealth in Italy was owned by 20% of the population. In the late 1940s, quality pioneer Dr. Joseph M. Juran recognized that this highly skewed distribution represents a universal law of uneven distribution across industrial systems.

Juran formulated the quality engineering axiom: roughly 80% of process defects, scrap costs, or customer warranty claims arise from approximately 20% of the discrete defect causes. Juran famously termed this phenomenon the 'vital few and the trivial many'. In later years, Juran formally revised his phrasing to 'the vital few and the useful many', acknowledging that while minor secondary defects do not warrant top priority, they are never truly trivial in customer satisfaction.

+-----------------------------------------------------------------------------------+
|                         THE PARETO PRINCIPLE (80/20 RULE)                         |
+-----------------------------------------------------------------------------------+
|              80% of Quality Losses / Scrap Dollars                                |
|             ----------------------------------------                              |
|               Originate from 20% of Root Causes                                   |
|                                                                                   |
|    [ VITAL FEW ]                                 [ USEFUL MANY ]                  |
|    ~20% of Categories                            ~80% of Categories               |
|    Accounts for 80% of defects                   Accounts for 20% of defects      |
|    --> Focus continuous improvement here         --> Address in subsequent cycles |
+-----------------------------------------------------------------------------------+

---

Anatomy of a Pareto Chart

A Pareto chart is a specialized dual-axis hybrid chart that combines a categorical bar graph with an ogive cumulative percentage line curve:

  Count (Y1)                                                    Cum % (Y2)
   300 +---------------------------------------------------------+ 100%
       | [XXXX]                                          *--*---*| 
   225 | [XXXX]                                     *---'        | 75%
       | [XXXX]   [XXXX]                       *---'             | 
   150 | [XXXX]   [XXXX]                  *---'                  | 50%
       | [XXXX]   [XXXX]   [XXXX]    *---'                       | 
    75 | [XXXX]   [XXXX]   [XXXX]   [XXXX]   [XXXX]              | 25%
       | [XXXX]   [XXXX]   [XXXX]   [XXXX]   [XXXX]   [XXXX]     | 
     0 +---------------------------------------------------------+ 0%
         Burrs    Oversize Scratch   Thread   Pinhole  Other
                   Bore              Damage

Critical Structural Rules for the ASQ CQT Exam

1. Horizontal Axis (X-Axis): Defect classifications or failure modes arranged in strictly descending order of magnitude from left to right.
2. Primary Vertical Axis (Left Y1-Axis): Represents the unit of measurement—either frequency count, occurrence tally, downtime hours, or financial scrap dollars. The maximum scale of the left axis should equal or closely bound the total sum of all occurrences ($N$).
3. Secondary Vertical Axis (Right Y2-Axis): Represents the cumulative percentage, scaled precisely from 0% at the baseline to 100% at the top.
4. Column Bars: Rectangular bars whose heights correspond to the individual count or cost on the left axis. All bars must share equal width and touch or have consistent minimal spacing.
5. Cumulative Percentage Line (Ogive Curve): A continuous line connecting points representing the cumulative percentage at each category. The line begins at the top of the first bar (or at its cumulative percentage coordinate) and terminates precisely at the 100% mark on the right axis directly above the final bar.
6. The 'Other' / 'Miscellaneous' Category Rule: Minor defect categories with very low frequencies are typically grouped into an aggregate 'Other' category. The 'Other' category must ALWAYS be placed as the final bar on the extreme right of the X-axis, regardless of whether its combined total exceeds one or two of the minor individual categories.

---

Step-by-Step Construction with Worked Numerical Case Study

To understand the mathematical mechanics of constructing a Pareto chart, consider an ASQ CQT shop-floor scenario where an inspector tallies defects on precision brass fittings over a one-month production run.

Step 1: Collect Data and Tally Raw Frequencies

An inspection run of 2,500 parts yields $N = 292$ total nonconformities across five distinct classifications:

• Burrs on hex edges: 142 occurrences
• Oversize bore diameter: 86 occurrences
• Surface scratches on sealing face: 34 occurrences
• Damaged external threads: 18 occurrences
• Miscellaneous / Other: 12 occurrences

Step 2: Sort Categories in Descending Order

Ensure the categories are sorted strictly from highest frequency to lowest, keeping 'Other' at the end.

Step 3: Calculate Individual Relative Frequency Percentage ($P_i$)

For each category $i$, calculate its percentage of the total defect pool using the formula: $P_i = \left( \frac{f_i}{N} \right) \times 100\%$ Where $f_i$ is the individual category frequency and $N = \sum f_i = 292$.

• Burrs: $\left(\frac{142}{292}\right) \times 100% = 48.63%$
• Oversize Bore: $\left(\frac{86}{292}\right) \times 100% = 29.45%$
• Surface Scratches: $\left(\frac{34}{292}\right) \times 100% = 11.64%$
• Damaged Threads: $\left(\frac{18}{292}\right) \times 100% = 6.16%$
• Other: $\left(\frac{12}{292}\right) \times 100% = 4.11%$

Step 4: Calculate Cumulative Percentage ($C_k$)

Sum the individual relative percentages sequentially: $C_k = \sum_{i=1}^k P_i$

• Category 1 (Burrs): $C_1 = 48.63%$
• Category 2 (Oversize Bore): $C_2 = 48.63% + 29.45% = 78.08%$
• Category 3 (Surface Scratches): $C_3 = 78.08% + 11.64% = 89.73%$
• Category 4 (Damaged Threads): $C_4 = 89.73% + 6.16% = 95.89%$
• Category 5 (Other): $C_5 = 95.89% + 4.11% = 100.00%$

Master Pareto Tabulation

Interpretation

The top two categories—Burrs and Oversize Bore—account for 78.08% (nearly 80%) of all process defects. By dedicating quality engineering resources solely to fixing tool deburring and boring bar wear, the facility eliminates over three-quarters of its nonconformances.

Frequency vs. Weighted (Cost / Severity) Pareto Charts

A critical trap on the ASQ CQT examination—and in industrial practice—is assuming that defect frequency is always the correct metric for prioritization. When all defects carry identical rework costs, raw frequency is appropriate. However, when failure modes have vastly different economic or safety consequences, a Weighted Pareto Chart must be constructed.

Consider an aircraft engine component machining cell tracking defects over six months:

+-----------------------------------------------------------------------------------+
|                         FREQUENCY VS. COST WEIGHTED COMPARISON                    |
+-----------------------------------------------------------------------------------+
| BY RAW FREQUENCY:                                                                 |
| Rank 1: Surface Scuff (64.3%)                                                     |
| Rank 2: Thread Burrs (30.0%)                                                      |
| --> Top 2 represent 94.3% of all occurrences, BUT only $2,175 (7.2%) of cost!    |
|                                                                                   |
| BY FINANCIAL COST (WEIGHTED PARETO):                                              |
| Rank 1: Fractured Casting Wall ($20,000 = 66.1%)                                 |
| Rank 2: Eccentric Bearing Bore ($8,000 = 26.5%)                                  |
| --> Top 2 represent 92.5% of total economic loss ($28,000 out of $30,175)!       |
+-----------------------------------------------------------------------------------+

If the quality technician prioritized projects based strictly on frequency, engineering resources would be expended polishing out $2.50 scuffs while catastrophic $2,500 casting fractures continued to bankrupt the operation. A cost-weighted Pareto chart aligns quality engineering focus directly with business financial impact and customer risk.

---

Advanced Pareto Techniques: Stratification and Before-and-After Verification

1. Stratification Analysis (Second-Level Pareto)

When the dominant bar of a Pareto chart is identified (e.g., 'Burrs on hex edges' accounting for 48.6% of defects), the quality technician does not stop there. The technician performs stratification—breaking down the dominant category into a secondary, more granular Pareto chart by:

• Machine / Workstation (Machine 1 vs. Machine 2 vs. Machine 3)
• Operator / Shift (Shift A vs. Shift B vs. Shift C)
• Raw Material Vendor (Foundry X vs. Foundry Y)
• Tooling Supplier (Insert Brand Alpha vs. Insert Brand Beta)

Stratification frequently reveals that 90% of the dominant defect originates from a single cutting tool vendor or one night-shift machine spindle.

2. Before-and-After Pareto Charts (CAPA Verification)

Pareto charts serve as indispensable tools during the Check/Verify phase of the Plan-Do-Check-Act (PDCA) cycle and Corrective and Preventive Action (CAPA) investigations. By plotting a 'Before' Pareto chart side-by-side with an 'After' Pareto chart using identical category scales:

• The technician verifies whether the implemented corrective action successfully eliminated or reduced the targeted 'vital few' defect category.
• The technician confirms whether the operational change inadvertently caused a secondary failure mode to spike (a phenomenon known as 'defect migration').