Analyzing Processes

Using Pareto Charts on surveys

We have learned that with Pareto Charts we can identify the main effects that cause a problem. The Pareto principle stays that in many situations 80% of the effects comes from only 20% of the causes. So, with Pareto we can identify those main affects that are causing the problem, and attack them, so, by eliminating those effects the problem will decrease a lot.

Pareto Charts are useful in analyzing processes, but, in this article, I have found out another use of these Charts that is very interesting; also related with the business world.

They use the Pareto Chart on surveys, finding out that, in business, we are likely to find that only 20% of the customers are responsible for 80% of the complains to the firm. So, by focusing in that 20% the firm could reduce complains in a more efficient way than focusing on all of their customers.

So Pareto Analysis can be applied to process defect, but also to customer complaints data to figure out where all those complains come from, so we can focus on those customers who seem unhappy with the services of the firm, and make then happy; have a nice experience.

In order to develop this analysis the first thing to do is collecting data through different ways like, customer surveys, focus groups, review of complains… After all the data is collected, the order to conduct a Pareto analysis for surveys is the following:

  1. Order the data, so we have a count of the number of occurrences of each category. For example, in a survey there are usually questions that need to be tabulated, like when we rank a service between very pleasant or poorly pleasant.
  2. Create a Pareto Chat with the tabulated data.
  3. Examine the chart to find out if the Pareto principle can be applied to the case. The Pareto Chart order the data from the most frequent category to the less frequent. So in this step we need to figure out if the first two (or just the first) category represents more than the 80% of the total data.
  4. Determinate the next step, based on the results. If the Pareto principle applies we should focus on those two categories. And, if it does not apply we should break down the data in a different way. For example, instead of looking at complains about the product type, analyse those complains by, for example, productivity type or customer type.

As I have showed, Pareto Charts have many uses besides analyzing processes, but the process to create these charts is basically the same.

http://www.brighthubpm.com/templates-forms/75631-using-pareto-analysis-on-surveys/