For instance, a manufacturer might want to place a value on the potential effect of different defects on the company's reputation, a much more difficult thing to assess. Other attributes may be harder to quantify. For example, a garment manufacturer might know that wrinkles cost $.10 to fix, while dirt specks cost $.50. Weighting requires a valuation: you weight the frequency counts by assigning attributes, such as cost, severity, or detectability, to each defect type. This attribute could be objective, such as the dollar amount it costs to fix each type of defect. And as we saw when we looked at bad PR instead of incident counts, a weighted Pareto chart may change how we see the priority for improvement projects. A weighted Pareto chart accounts for the severity of the defects, their cost, or almost anything else you want to track. Setting Up Data for a Weighted Pareto ChartĪ weighted Pareto chart doesn't just look at how often defects occur, but also considers how important they are. Because the most common problems aren't always the most important ones, a weighted Pareto chart can give extra emphasis to the most important factors. There's another approach to getting more insight from a Pareto chart: we can look at the data in conjunction with another factor, like a cost, to create a weighted Pareto chart. However, this is really just a different type of frequency data: in effect, we're counting the number of complaints rather than the raw number of defects. For example, even though it was the most frequent defect, the band starting late was barely mentioned in negative reports. This is very important information for the PR manager, because it shows which types of incidents resulted in the biggest number of negative mentions.These results are quite different from the raw counts of defects. Here's how that data looks in a Pareto chart: This gives us insight into how the different types of incidents played out in the media. We can add a column of data to our worksheet that tallies the number of news reports, online reviews, and social media mentions about the various incidents that took place on tour. For example, the Zero Sigmas' public relations manager gathered all the coverage about the recent tour, and she wants to know how it corresponds to things that happened while the band was on the road so she can be ready to handle things that might happen on the next tour! But that's not always what you want to measure. In many situations, you do want to know the number of defects. You might be able to learn more by looking at a different measurement. When different defects have different impacts, a Pareto chart based only on number of occurrences doesn't provide enough information to tell you which issues are the most important. How is that reflected in the Pareto chart above? It's not. In the world of rock and roll, the Zero Sigmas starting a performance late probably has fewer consequences than their getting caught lip-syncing during a performance does. Similarly, if a shirt is sewn with thread that's just slightly off color, the defect is so small the garment might still be usable a shirt with mismatched fasteners will need to reworked or discarded. You can see why this is problematic if you think about defects that might occur in manufacturing a car: a scuff mark on the carpet is undesirable, but it's not on par with a disconnected brake cable. But are they equal? Looking at just the raw counts of the incidents assumes all problems or defects are the same in terms of their consequences. ("Hello, Albuquerque!" was correct on only one night of the tour.)Īll of these are incidents we'd like to happen much less frequently. Value forgot the words to his songs and greeted the wrong city when he hit the stage. It's also useful to see how frequently singer Hy P. We know the Zero Sigmas started every show late, making that the defect that occurred most often, and this information is valuable.