Springville Herald Case

The first data we analyzed was which errors occurred most frequently. The above Pareto chart serves to separate the “vital few” errors from the “trivial many”. The first 7 types of errors (from left to right) account for 78% of the total service errors. Concentration on eliminating those types of errors is a good first step in minimizing customer service errors and boosting revenue. If you can eliminate less than half of the error types you can eliminate more than 2/3 of the total errors. Next we looked for correlations between the data above and which errors were most costly.
We again chose Pareto charts to express the relationships between the types of errors and how much they cost the company. The use of Pareto to express the total cost of each error type is valuable to identify which error types are costing the most cumulatively and also offers some correlations. Again we see the first 7 error types (from left to right) make up a large majority of the money spent correcting errors. 79% in fact. We find that 5 error types: Typesetting, Wrong position, Ran in Error, Wrong ad, and Wrong date occur in the “vital few” data of both frequency and total cost of errors.
Further concentration on these 5 error types will not only go a long way in eliminating the frequency of errors, but will also eliminate a large portion of the total cost associated with service errors. Another important finding in this data is that while copy errors occur most frequently (17% of total errors) they are relatively inexpensive to fix (only 6% of the total cost of errors). So eliminating copy errors will go a long way in improving customer service, but will not have the same impact on the cost of fixing service errors.

Examining the cost data further we can see which errors are the most expensive to fix on a per error basis. While Pareto was not necessary to express cost per error (cumulative % is not important in this case), it is the easiest type of chart to read with this much data and serves to show (from left to right) which errors are the most expensive to fix per occurrence. These findings reveal that Ran in Errors are the second most expensive type of error per occurrence. That combined with the fact that we already know Ran in Errors account for the highest total cost of errors (20. %) and are the 4th most frequently occurring (9%) tells us that concentrating most heavily on eliminating Ran in Errors would be the most efficient way to simultaneously improve customer service and cut costs. So let’s took a closer look at Ran in Errors. As you can see, Policy Ran in Errors are the most frequently occurring (53% of total) and by far the most expensive (82% of total). Eliminating these errors as quickly as possible would be the most efficient way to achieve the goal of improving customer service and cutting costs. Some information that would be useful to examine would be how the errors interact with each other.
Do some errors cause others? Even if no error directly causes another it would be useful to know if eliminating errors that occur at the beginning of the publishing time line would prevent others from occurring due to the nature of publishing them. Also, observe the histogram below. As you can see the number of help desk calls per day is concentrated between 40 and 70 per day. It would be useful to know what errors these calls are in regard to. With the average calls per day known, the Herald can also streamline their customer service department to be able to handle this volume efficiently.