Normalized Reporting Overview
Normalized reporting can provide a more accurate evaluation of products, stores, or any other data element of a case. It does this by considering factors such as the number of units sold, the total value of items sold, or the number of customers served.
How Normalization Works
Here is a simple example. Imagine a company that sells four products. Consider the following data produced from a standard Frequency report:
Table 1: Product Complaints Per Month
Product | Complaints Per Month |
|---|---|
Product A | 15 |
Product B | 20 |
Product C | 10 |
Product D | 12 |
Totals | 57 |
Based solely on the number of complaints received, it appears that Product B is performing poorly, and Product C is performing very well. Now look at the sales figures for the same products:
Table 2: Units Sold Per Month
Product | Units Sold Per Month |
|---|---|
Product A | 50 |
Product B | 200 |
Product C | 20 |
Product D | 80 |
Totals | 350 |
We see here that Product B is the best-selling item, and Product C sells the fewest number of units. The normalization options on Emplifi Agent reports combine these two tables so that the complaint information is weighted based on the sales data.
In order to normalize the frequency data, an Average Norm Factor must be calculated. This is done by dividing the Total Count (in this example, the total number of complaints) by the Total Norm Value (total units sold in our example).
Total Count | / | Total Norm Value | = | Average Norm Factor |
|---|---|---|---|---|
57 | / | 350 | = | 0.1629 |
Rows with Norm Set Data of "0.0" or that are missing Norm Set Data are not included in the totals for the reports.
Also, when different categories have the same code or folder value, the correct Normalized values are reported. In addition, when a category is the same as another Company ID (not the Company ID of the reporting data), the Normalized values are taken from the “same as” Company ID for the category and the frequency counts are taken from the company for which the report is being ran.
This represents the average number of complaints per units sold for all products. In other words, on average there were 0.1754 complaints for every unit sold of any product.
Next, the Norm Factor for each product must be calculated. This is done by dividing the Count (number of complaints received) by the norm value (units sold).
Product | Count | / | Norm Value | = | Norm Factor |
|---|---|---|---|---|---|
Product A | 15 | / | 50 | = | 0.3 |
Product B | 20 | / | 200 | = | 0.1 |
Product C | 10 | / | 20 | = | 0.5 |
Product D | 12 | / | 80 | = | 0.15 |
Each product's Norm Factor is then compared to the Average Norm Factor to determine how that product is performing compared to the "norm" (or average). This value, called % of Norm, is created by dividing the product's Norm Factor by the Average Norm Factor, then multiplying by 100 to compute a percentage.
Product | Norm Factor | / | Avg Norm Factor | x 100 | = | % of Norm |
|---|---|---|---|---|---|---|
Product A | 0.3 | / | 0.1629 | x 100 | = | 184.16 |
Product B | 0.1 | / | 0.1629 | x 100 | = | 61.39 |
Product C | 0.5 | / | 0.1629 | x 100 | = | 306.94 |
Product D | 0.15 | / | 0.1629 | x 100 | = | 92.08 |
If a product is performing "normally" compared to the rest of the products, then its Norm Factor will be close to the Average Norm Factor, and its % of Norm will be close to 100.
In our products, we see that, on average, Product B is our best performing product, whereas Product C is the worst. This is exactly the opposite conclusion that would have been drawn from the data on the standard frequency report (Table 1).
In this example, a product with a higher % of Norm value is considered to be performing poorly. If Table 1 represented complements instead of complaints, then a higher % of Norm would be considered good.