Anticlue

Six Sigma: Defects per Unit / Defects per Opportunity

Quantifying what you are going to measure is important for setting a baseline. It simply let’s you know that the changes will have a positive or negative affect. As one looks to remove defects from a business process, one must measure them. Six Sigma has two tools for measure defects, defects per unit (dpu) and defects per opportunity(dpo).

Defects Per Unit, DPU, evaluates the average number of defective units which occur the total sample size. A unit is the item being processed, such as a incident ticket, or the product being coded, or the service being rendered. DPU counts each unit as either defective or not defective. If a unit which is defective has one or several defects, it is counted as a single defective unit.

Calculating DPU
Calculating a process's DPU involves three steps:

1. Determine the total number of units you will sample.
   The first step determines the size of your sample group. A sample group should be small enough to be manageable, yet large enough to reflect whatever problem is undermining the process.
2. Count the number of defective units that occur within that sample.
   The second step shows how many units in the sample group contain at least one defect, or error.
3. Divide the number of defective units by the number of total units.
   The third step gives you the DPU as a decimal number, which can be converted to a percentage.

Defect per Opportunity depicts the average number of defects which occur in the total number of opportunities in a sample group. An opportunity is the chance for a defect to occur within a unit. DPO counts each defective opportunity within a unit as one defect. So in examining a service desk response, some opportunity for defects are:

• Response does not occur within 4 hours.
• Response categorized within the wrong support team.
• Team working the response doesn’t have the customer approval of a working machine afterwards.
• Wrong customer contact information gathered.
• Wrong problem detail reported.
• Wrong problem resolved.

Within our service desk example, each of these opportunity. So if the incident response is mis categorized to the wrong response team and the customer is not responded to within 4 hours. We have two defects within one unit.

Calculating DPO
Calculating a process's DPO involves five steps:

1. Determine the total number of units to be sampled.
   The first step determines the size of your sample group. The sample size should be small enough to be manageable, yet large enough to reflect whatever problem is undermining the process.
2. Determine the number of defect opportunities per unit.
   In the second step, opportunities are determined by creating a list of potential defects customers will care about; focusing on places where something can go wrong, not on the ways it can go wrong; focusing on routine, rather than rare defects; and grouping related defects into one category.
3. Determine the total number of defect opportunities for the sample.
   In the third step, the total number of defect opportunities is determined by multiplying the number of units in the sample group by the number of defect opportunities per unit.
4. Count the defective opportunities within the sample group.
   The fourth step shows how many opportunities within the sample group contain defects, or errors.
5. Divide the total defects by the total opportunities.
   The fifth step gives you the DPO as a decimal number, which can be converted to a percentage.

So when should you use DPU or DPO as a technique? This depends on how many performance standards exist within a process. A performance standard is an expectation from someone inside or outside the organization has for the process, product, or service.

There are ways to choose which measurement technique to use in particular circumstances:

• DPU measurements are easier to obtain and understand than DPO measurements.
• The DPU technique is best used when there is a single performance standard to measure because DPU measurements may hide the actual number of defects in a unit.
• DPO measurements provide more information than DPU measurements about how many actual defects exist in a unit.
• The DPO technique is best used when there are multiple performance standards to measure because obtaining those numbers is complicated.

When choosing between the DPU and DPO measurement techniques, it is necessary to determine the number of performance standards that exist for that process. The DPU technique is best used when there is only one performance standard. The DPO technique is best used when there are multiple standards.