Measuring quality management system ROI



I am interested in knowing how others have approached in measuring quality management system ROI? What are the measurement that we can put in place?


The logical indicators are reduced scrap and rework and increased customer satisfaction. Also, increased efficiency leading to reduction in direct labor oosts and improved vendor performance.

The value in dollars of these may best be assessed with the help of a good cost accountant and an effort to benchmark from a starting point. Obviously some are DIFFICULT to measure, e.g. new or continued business due to maintenance or improvement of customer satisfaction. But the effort to quantify ROI is worth the effort.



This may sound silly but very often we fail to be able to show the link between QMS initiative and scrap cost or productivity improvement. These are mainly over shadow by cost reduction or productivity initiative. Is there a way to implicate that the cost or productivity improvemet is due to QMS intiative? We are looking for a method to quantify QMS efford and to be able to justify the number headcount we have. Pls. advised.

Anton Ovsianko

be simple


The ROI of a QMS is an issue covered by the whole discipline - the economics of quality.
There is a multitude of factors which are more or less economically significant. And it is not always easy to predict the importance of each of them.
Here are some:
- economy of material resources - due to more efficient use.
- less defect rates - less correction costs, less negative image in the market
- higher consumer satisfaction - higher loyalty, favorable word of mouth among existing and potential clients
- more transparency of management system - better co-ordination, higher flexibility - market success
- better market image (especially by ISO 9000 certification) - easier selling, satisfying formal criteria of large corporate customers
- less controll on the part of our customers - more consumer value, higher turnover, higher price.
- etc.

Speaking of ROI calculation, I would start with identifying goals of the QMS introduction and development. ROI is a parameter of an investment project. So let us look and a QMS introduction (development) as an investment project.
We introduce an quality system element... why? Say, to lower defect rate... What will it bring us?
- production cost decrease by X dollars per piece
- correction cost decrease by Y dollars per piece
- better image and additional sales providing for additional profit of Z dollars per month

Total is X*Q+Y*Q+Z = P

It is definitely going to cost us something:
- control costs = A per month
- additional costs of purchasing better materials at higher prices = B per month
- additional data management costs (computer systems, paperwork etc.) = C per month
- additional workforce = D per month

Total additional cost = A+B+C+D = E

P-E = net additional benefit of a QMS project

P/E might be used as a ROI

It is an ideal model, of cource. On prcatice it is not easy to gather all the above information. Additional profit may be raised with some lag or be diminished throught some other factors not connected to the QMS - for example infavorable market conditions, or be exaggerated through a market boom.

Otherwise, we can estimate the ROI in terms of customer satisfaction and its correspondence with the company's profits. But it is even harder to transform into figures and compare to the costs.

IMHO there are no other ways to calculate a ROI than comparing profits to expenses.

Anton Ovsianko


Just a question?

We are in the process of implementing ISO9001. We have been told that we need to have a measure for each and every process that we define and the derive a composite index for QMS improvement.

Is having a measure for each and every process a requirement of the standard?

Isn't it possible to apply the 80:20 rule where we identify the key (20 to 30 %) processes for measurement?

Would appreciate you response on this.

Thanks and regards
N Varma


Measuring for the sake of measuring

I would never propose measuring something for the sake of have a measurement for each and every process. A trucking company was measuring stuff because their auditor said they needed to measure something (a non-existent requirement of the old 4.20). I asked the president to pay attention to two scenarios. The first is whey she asks: “How many times has this occurred?” The second is when she asks: “Why don’t we have this data?” Both of these are signals that there is data that might need to be tracked.
Measure what makes sense and what benefits you. If you are measuring something and it has value, keep it up. If it doesn’t, quit measuring. If you find something that needs measuring, then measure. It is that simple.

Dave B (the other Dave)



Oh, I forgot to add, thanks for the question and welcome to the cove, unregistered, make sure you register. We need your questions and input for this forum to work.

Dave B (the other Dave)

Neelanshu Varma


( I had earlier posted the comment on measurement as an unregistered user)

Thanks for your comments, I had used the same logic that you have given. The person who is guiding us replied that ISO9001:200 clause 4.1 c requires that the organisation shall determine the effectiveness of all the processes and this is possible only by measuring each process.

Has anyone faced an audit where this issue has come up ??



4.1 c)

Glad to see you registered. Unless my standard is different from yours, 4.1 c) reads:

“determine criteria and methods needed to ensure that both the operation and control of these processes are effective. “

I think the key here is the part that says “determine the criteria and methods”. If that means measurement, then measure. If there is some other way that ensures the operation and control then use that. But wait…there’s more.

4.1 e) requires monitoring, measuring and analyzing these processes. When placed together, it appears that processes must be measured, as you’ve been told. But you could use a single metric to measure several processes. In many cases, it makes better sense.

Part of the problem may be how detailed you define the processes. For example, welding blanks may be considered a single process. I can also break this process down into preparation of the surfaces, staging the parts, performing the weld, inspection and test, packaging and delivery each as a separate process. It may be far easier… no strike that, far more valuable to measure the entire process. Then again it may be much harder, but more valuable to measure each separately. Then again, you might want to measure the entire process, and measure one aspect of the process if it is area of concern.

Remember, it is YOUR system. It needs to work for you!

Dave B (the other Dave)


Right on Dave!

I understand exactly what you're talking about in deciding how far to go in defining your "processes".

I took three ISO9k2k courses and the point that Varma made about being req'd to monitor and measure the processes *and* that actions are taken to achieve PLANNED results and continual improvement, was driven home. The instructor also pointed out however, that 4.1a says that the organization shall "identify" these processes. Not the standard, not the registrar, but the organization would do this.

And then he said that when we went about defining our processes, we should keep in mind that each one we "identified" would have to be measured and have planned actions for continual improvement, etc. He said that if you didn't want to establish individual measurements and planned actions for something, then you should look at it as a small piece of a larger process that you can establish monitoring and measurement activities for. The fewer processes you "identified", the more open you could be with your planning and the areas in that process which could experience improvement.

He showed us a process flow chart that one company had produced and they had broken out every little thing. If this was the final product that they used to demonstrate their "processes",then they would be having to plan for results and monitor and measure, etc. for every little thing that they did!
Not only that, but remember the issue of "control"! You have to control your processes. Aaaccck.

So for my company, starting from ground zero, we are essentially using the megaprocesses of ISO (as described in the note section of 4.1) but have slightly broken down provision of resources and some major steps in product realization. The flowchart looks great, but could use some improvement. Beyond this rudimentary identification of processes, we will have the objectives that generate monitoring and measurement of their own and play contributing roles into the major process improvements. Our objectives are really specific and will change at least annually.

But for the better understanding within a department, we will create flowcharts of smaller "contributory" processes. These will demonstrate all the little bits that make up the whole Official Process.

FYI, in a following course, I mentioned to the instructor what the other instructor had advised regarding making your processes large or at least having the Right and Responsibility to define your own processes as large or small as you wanted. The second instructor completely missed the point and wanted to insist that ISO identifies 23 (or so) processes and you had to use those processes. He didn't see where the production of widgets (as a process) could include the training of the workers or the receiving of the raw materials. And both of these instructors/assessors worked for the same registrar!
Top Bottom