Soft dollars vs. the bottom line
Q: Last week I made a proposal to my company’s executive management team that showed how we could save money by targeting certain processes for improvement. Although they agreed that the improvements would be beneficial to the company, the executives argued that most of the savings were in soft dollars. To be given any serious consideration, the savings needed to be dollars that impacted the bottom line. What are they looking for?
A: It would be best to follow up with your executive management team to better understand what they mean by soft dollars. Soft dollars can refer to improvements that result in a reduction of the number of steps or the time it takes to execute a process.
You can calculate the costs of the steps or time eliminated by a process improvement, then multiply it by the number of repetitions per year to come up with an annual savings projection. Some executives might refer to this savings as soft dollars because they might not impact the bottom line—or net income—of the company.
Process improvements that will affect the bottom line will either increase sales or reduce costs. Reducing these costs would likely impact the bottom line: materials or inventory used; maintenance and repairs; utilities; payroll (not a popular one!); and other items that require the company to spend money.
A reduction in process time can impact the bottom line if the time saved can be used to increase production, as long as there is sufficient demand. Process time reduction can also impact the top line (sales) if, for example, design cycle time is reduced and the product is brought to market quicker, which might lead to increased sales.
Process improvements could also improve the quality and/or reliability of products and services, which might lead to increased customer satisfaction, higher sales and increased net income. You might want to work with your accounting department to discuss how your improvements will affect the bottom line and then present a revised proposal to your executive management team.
Ken Cogan, email@example.com
Senior manager, performance management Intelsat
How low can you go?
Q: I do a lot of process mapping with teams at my organization. Usually, we start with a high level map of the process and then break down the process of interest. This approach continues for some time. However, I’m never really sure when to stop process mapping the process. How far should I drill down?
A: What you are describing is illustrated in Figure 1, which represents a generic breakdown of processes within any type of organization. Although only four levels are depicted in the figure, there can be multiple levels of processes and sub-processes.
I don’t believe there is an absolute answer to your question because practical considerations always come into play. However, the following 10 rules of thumb can serve as the basis for determining when to stop:
- An activity or step can be attributed to a specific individual or job family. In this case, accountability has been achieved.
- An activity or step can no longer feasibly be decomposed. Further decomposition might require dropping into the domain of time and motion study.
- The time required to measure the process exceeds the time required to perform the process. (This might not be known in advance.)
- The cost required to measure the process exceeds the value-add of the process. (This might not be known in advance.)
- Responsibility for performing the process transfers out of the organization—when the process is outsourced, for example.
- Responsibility for performing the process falls outside of the process boundaries you established at the beginning of your mapping activity. Rememb er, bound your process mapping activity or you’ll never complete it.
- Root causes become evident.
- The seven (sometimes eight) categories of waste become evident.
- The subject matter experts (SMEs) cannot agree on the
“as is” process. When this situation occurs, it
could be construed as a signal that the current process has
become so complex that it should be discontinued and replaced
with a new one.
Or, perhaps the SMEs are no longer true SMEs. Perhaps organizational turnover has been such that currently designated SMEs no longer possess the necessary process knowledge. I ran into this situation quite a few years ago working on an organization’s short-term and long-term disability processes. All the knowledge experts had been replaced three or four times over. The team could never reach agreement on the “as is” process.
- A process (or subprocess) owner has not or can not be identified. If this situation occurs, there is no one who can authorize process changes.
Remember, these 10 guidelines are rules of thumb and are not absolute. Also, one or more guidelines might be in effect at any given time.
Tom Kubiak, firstname.lastname@example.org
President, TK Performance Improvement Solutions
To tighten, or not to tighten...
Q: I am keeping a statistical control chart on a product measurement. I’d like to tighten the control limits to require production to make a better product. When is it OK to do this?
A: You are confusing statistical control limits with specification limits. The control limits are based on the actual product measurement data and tell you whether the manufacturing process is stable, and, if so, what average and variation it can be expected to maintain as long as the process is unchanged. If control limits are violated, it tells you something has changed.
This could be good or bad, depending on the change. You should not change the control limits unless the chart tells you the process has changed and you are OK with the change.
The specification limits are based on performance requirements and should be determined by design factors and customer expectations. If these requirements change—if a customer wants a more consistent product, for example—then you need to change the specification limits.
The relationship between product specifications and statistical control limits is defined by the notion of process capability. This is usually defined as the ratio achieved when dividing distance from process average to closer specification limit by three standard deviations of individual data points. Typically, a ratio of 1.3 or greater is considered acceptable in terms of consistently being able to produce product that falls within specification limits. (Some critical measures demand larger ratios.) If it is less than 1.3, you probably need to do a lot of sorting and work to improve your process.
Back to your question: Based on your existing situation, calculate your process capability. If it is greater than 1.3, you can comfortably tighten your specification limits. If you have made process changes that have improved your average level or reduced variation, you can recalculate your statistical control limits and then recalculate your process capability. Never change your statistical control limits unless the data from your process tells you to do so. However, you should change your specification limits if there is a product requirement reason to do so. In any case, use the process capability measure to assess the practical impact of a specification change.
Many statistical software packages provide easy calculation of process capability.
Joe Tunner, email@example.com
Q: Mean, median, average, and mode. What’s the difference?
A: The first three terms are measures of central tendency. Most people agree that average and mean are one and the same. While many of us understand that mean is the arithmetic average of all measurements in a data set, there are other variations of mean. Median is the middle number or center value of a set of data in which all the data are arranged in sequence. Mode is the value occurring most frequently in a data set. Many of these definitions can be found in the QP Quality Glossary (www.asq.org/glossary/m.html).