Six Sigma
Way
Way
BI
Quality
Quality
Copyright © 2007 - 2008 All rights Reserved
BI Capability
Maturity Model
Maturity Model
BI Comprehensive
Improvement Model
Improvement Model
Six Sigma BI Statistics
The statistics are key:
1. We need to prevent problems, rather than react to problems.
The Statistics allow us to monitor the product for quality and set tolerance limits which can trigger corrective action.
2. We need to understand what creates product quality and what are all the causal and impacting factors to the quality product.
The breakthrough formula helps us define and understand the relationships between the quality results and the impacting/causal factors. We will set specifications for each of these and manage to those specficiations.
3. When we make changes to the product, components, process and/or other assets, we need to understand exactly what is happening - i.e. how those changes are impacting quality of the product. How much are we spending in resources for what kind of return? Have we achieved the desired goals?
Statistics allow us to measure and assess the data.
The Performance scale is in whatever units are being used to measure the target, e.g. response time in seconds. The goal in Six Sigma is to come as close to the target as often as possible. Thus, the Critical to Quality measurement may be for response time for a simple query to the information library. The target is to consistently (i.e. all queries) keep the time at 5 seconds or less, i.e. from key entry of the request to the response. In this specific application, the customer perceives this as the response time which represents quality. We set the quality target for response time based on what the customer needs to perform the job, i.e. receive the most value from the information. ( This goal will also be written into a Service Level Agreement which is well documented and shared - so that everyone will understand the response time target.)
Variation
Variation is a deviation from expectations. Variation is the degree to which a part, product, service or transaction differs from all the others in the same class. The reality is that the target will be missed some of the time. No matter how excellent the process and underlying structures are, it is just not possible to hit the target every time. There will always be some variation. The amount of time it takes to process a query varies. This variation shows up as a distribution of actual response times, i.e. performance units, around the target. Normally the distribution is in the form of a bell shaped curve. This normal distribution is symmetrical about the central line (target). The areas to the left and to the right are equal.
Variation
Variation is a deviation from expectations. Variation is the degree to which a part, product, service or transaction differs from all the others in the same class. The reality is that the target will be missed some of the time. No matter how excellent the process and underlying structures are, it is just not possible to hit the target every time. There will always be some variation. The amount of time it takes to process a query varies. This variation shows up as a distribution of actual response times, i.e. performance units, around the target. Normally the distribution is in the form of a bell shaped curve. This normal distribution is symmetrical about the central line (target). The areas to the left and to the right are equal.
Standards and Specifications
Since there are always variations, we add the concepts of standards and specifications. Just how much variation can we allow and still maintain the quality standards desired? Once that question is answered and the tolerances are defined, i.e. the upper and lower limits for performance, then we can measure the quality of the performance.
Now there is a way to indicate the acceptable extent of variation which can be allowed and still have a business information product of acceptable quality to the customer. These performance specification limits provide the parameters for defining, measuring analyzing, improving and controlling quality.
Quality may be defined as ‘getting as close as possible to the target with the least amount of variation’, i.e. hit the target and minimize the variation. Quality is essentially defined by conformance to standards/specifications. Any product (i.e. in this case a query/response) which falls outside the preset limits is considered a defect. Thus, if we set the limits as 3 seconds plus 1 second, any query/response which takes over 4 seconds is a defect. (Note: in this case less than 2 seconds for a response time is certainly acceptable. We might track those queries with lower
Since there are always variations, we add the concepts of standards and specifications. Just how much variation can we allow and still maintain the quality standards desired? Once that question is answered and the tolerances are defined, i.e. the upper and lower limits for performance, then we can measure the quality of the performance.
Now there is a way to indicate the acceptable extent of variation which can be allowed and still have a business information product of acceptable quality to the customer. These performance specification limits provide the parameters for defining, measuring analyzing, improving and controlling quality.
Quality may be defined as ‘getting as close as possible to the target with the least amount of variation’, i.e. hit the target and minimize the variation. Quality is essentially defined by conformance to standards/specifications. Any product (i.e. in this case a query/response) which falls outside the preset limits is considered a defect. Thus, if we set the limits as 3 seconds plus 1 second, any query/response which takes over 4 seconds is a defect. (Note: in this case less than 2 seconds for a response time is certainly acceptable. We might track those queries with lower
The Six Sigma BI
Initiative
Initiative
Should we
apply Six Sigma to BI?
apply Six Sigma to BI?
