Measuring and Improving Quality - Choosing Measureables

[Marc]

Subject: Re: Q: ISO Cost Effectiveness /Weiler/Ohri/Pfrang
Date: Fri, 1 Oct 1999 14:58:58 -0600
From: ISO Standards Discussion

From: (Doug Pfrang)
Subject: RE: Q: ISO Cost Effectiveness /Weiler/Ohri/Pfrang

Companies already measure cycle time, reworks, total overhead, instances of innovation, etc., but measuring those things does little to improve quality. And no, it does not escape the formal bookkeeping: companies routinely monitor these metrics because they can be leading indicators of potential future financial problems.

Also, installing "hard core indicators close to the operators and keep an eye on them" harkens back to the obsolete "inspect quality into the system" approach that was discredited at least thirty years. Such an approach blames workers for the results of management decisions that are very likely out of the workers' control. Moreover, such statistics rarely improve quality, because workers account for only about 10-15% of the problems in a company -- so even if one tries to motivate them to do their best at the given task (such as by mounting an indicator near them), one could only hope to achieve limited improvements. About 85+% of problems in companies are due to poor management. These problems include poor process design, poor communication, poor goal-setting, poor planning, poor investment decisions, etc. It's just not as simple as hanging an indicator by every worker and thinking that will fix everything.

The best way to improve quality is to determine what matters most to the most customers, and maximize the improvement to that. And, yes, it often can be measured. It is easiest to measure when products are in the mature stage of their lifecycle. Where it is harder to measure is in breakthrough products, where customer expectations are highly speculative.

-- Doug

>From: Edith Ohri
>Subject: Re: Q: ISO Cost Effectiveness /Weiler/Ohri
>
>Quality is an impossible thing to measure, because it's just an attribute and
>does not stand by its own. Even if quality was a real item or a product that
>stands by its own, we would not be able to cost manage it because ot the
>accounting conventions. For example, they don't acknowledge the worth of "time"
>"information" or anything that happens not to have a market value.
>Therefor, it's hopeless even to try walking in this direction.
>A practical solution for those of us that don't want to wait until the bin
>counter ever finish resolving their problems is to install hard core indicators
>close to the operations and keep an eye on them. A good indicator would be
>"cycle time" "reworks" "total overhead" "instances of innovation"... all the
>things that escape the formal bookkeeping. Indicators have one more advantage
>over pure financial reports: they are easier to interpret in case a change is
>detected you can trace it better to its cause and effect.

[Marc]

Subject: Re: Q: ISO Cost Effectiveness /../Ohri/Pfrang/Ohri
Date: Tue, 5 Oct 1999 15:58:05 -0600
From: Edith Ohri
Subject: RE: Q: ISO Cost Effectiveness /../Ohri/Pfrang/Ohri

Hello Doug,

There's an interesting fact which I found out while doing a study on Cost Management: by using (cleverly) non-financial indicators it's possible to locate problems within a company a year before any sign of them show up on financial reports! The diference in time worth so much that unless forced to use those reports, I'd stay away from them like from a plague.

Indicators are not free of faults, but then you can define them in a better way (something you can't do with the sealed black box of the financial reports). Also, they don't mislead you by presenting figures, which are accurate to the penny. When dealing with an indicator one is more aware of the fuzzy nature of this kind of data, and would not jump so fast to wrong conclusions.

I agree with you though on one thing, that tying the indicators to the work area alone is a miserable way to check on total quality, and that most of quality problems evolve from management (and other indirect labor - my addition). But who says that operations is only where "the workers" are? It seems that we are not using the same terminology and that's why the argument. By "hard core indicators" I meant not those that used to serve for incentives. A good quality indicator is one that reflects the intangible value of cost and benefit. Here are some indicators, following your example: "poor process design" increases the time to market, "poor communication" causes high level of human errors and reworks, "poor goal-setting" usually results in poor human relation and a higher rate of turnover, "poor planning" may reflect in poor workflow and high in-process inventories, and "poor investment decisions" increases among the rest the need for maintenance.

About your suggestion for a way to do quality improvement -- this is a different sort of art. One can be unsuccessful in measuring quality yet very good in improving it.

Edith

[Kevin Mader]

Marc,

Your posts got me thinking of something I read from Crosby:

"Measurements of nonconformance is not Quality!"

In our organization, I tried to create a balance of business measures and Customer driven measures. This approach has benefits for both sides, perhaps offering better balance within the organization.

Regards,

Kevin

[Marc]

Yeah - the 'right' measureables are important, but that's not 'the end'. Choosing measurables for a company is not as simple as some believe, either.