My book is with the printer. What can be the quality target?

18 February 2014 by Infinite Ideas in Business and finance

by Jan Gillett, author of Making your work work

After nine months of thought, research, writing and checking, Making your work work is in the hands of the printer – it’s too late to change anything now. In a few weeks several thousand copies will arrive.

Readers are likely to have high expectations about the quality of a quality-management related book. But how can we assess the quality of a book? Does the quality management world provide meaningful targets, or do we have to look elsewhere?

Have I got it ‘right first time’?

Well, the answer depends on what you mean by ‘right’, and ‘first time’. It’s a little longer than the original estimate, so in that sense it’s already not right. It’s been read by half a dozen people, and I’ve incorporated most of their observations, so right or not, that’s hardly ‘first time’. After considering proof reading it’s actually more like third or fourth time. Is that an indication of success, or failure?

Actually it’s inevitable. In addition, when the intended market (readers who don’t already know very much about using process or quality management for everyday work) sees it, it’s virtually certain that their feedback will lead to changes for a second edition. So, no, it can’t be ‘right first time’, and in fact having such an aim would paralyse creative work.

But I do of course hope that it is right enough to get a good reaction and for people to recommend it. Then we could do a second edition, which would also take several iterations in development.

So ‘right first timer isn’t a helpful target in this case. But it can be. Consider a self-assembly furniture item. Having a high level of right first time achieved by the customers would be great.

Are there ‘zero defects’?

Another deathless phrase, swallowed without thought by many big companies as a target for a change programme before it collapses in the face of real products and services, real customers. It sounded fine in concept, but application shows the arbitrary nature of virtually all definitions of a defect. Agreement between producer and user over an extended time about what a defect is, is effectively impossible.

No doubt there would be as many definitions of defect in a book as there are readers, so this target is not going to work.

Some zero targets, such as deaths from accidents, are real enough, in fact in that case any other target is an insult, but they are the exception.

Could the book achieve ‘Six Sigma’?

If not zero defects then, what about the six-sigma target of about 3 errors per million opportunities? Since PMI is a leading consultancy and training firm in Six Sigma we should be able to achieve that, shouldn’t we? But, and it really isn’t an excuse, let’s consider this one carefully too.

If, regardless of the above, we assume we could agree what a defect is (the customer’s tolerance) my book, with about 70,000 words in all, would need to have no errors at all. Hmm, anyone who has written more than a page or two will see the nonsense here. A Six Sigma achievement would be one error in the entire contents of a small library. In fact, one error per chapter would be pretty impressive, and that with a lot of effort in multiple reviewing, proof reading and so on. That’s about Four Sigma, apparently not much to shout about.

In fact this shows that ‘Six Sigma’ as a target should be used with great care. When used as originally by Motorola to drive up the quality of mass production of new electronic devices, it was transformational. The problem is that it has been adopted so carelessly that for many it is devalued. As a target for publishing it’s no help. Incidentally, if six-sigma as a target isn’t sensible for an organisation, then Six Sigma as a programme name is also not helpful.

How about an older ambition ‘on target with minimum variation’?

Going back more than forty years, in fact to the early 1960s, we arrive at this phrase, credited to Genichi Taguchi. Although it somehow seems less demanding than the specification-based western rallying cries of zero defects and so on, it is in fact a profoundly rigorous term, that you can apply to every circumstance.

If you seek to get your outputs ‘on target with minimum variation’ you need to:

•  appreciate what your customer values;
• establish criteria that you can measure, both of the customer characteristics and of your output;
• understand the process that leads to the output, and its context (how the work works);
• optimise the operation of the process to get the mean of the outputs close to the target, and with minimum variation about the mean.

If you are making crankshafts this philosophy enables you to achieve better than six-sigma, in fact it’s pretty routine. Those who have watched Don Wheeler’s video ‘A Japanese Control Chart’ will recall that the factory produced many millions of parts with none out of specification.

And it works just as well for a call centre manager trying to do their best with the variety of queries coming in on the phone.

Will Making your work work be on its target, will the variation be acceptable?

I have tried to keep in mind as my target audience everyday managers trying to achieve better and more predictable output from their work. Regular managers don’t have time for complications, but do need some analogies, inspiration and explanations. Thus every topic presented temptations to expand, to illustrate or to go into more detail, but doing too much would be going off target. Time will tell if I have achieved an acceptable compromise.

What about errors? Well here’s a minefield illustrating exactly the problem with defining what a defect is. Obvious ones like spelling surely? Well yes and no, for US or Indian English is not the same as English English. Well how about punctuation and phrasing? Much of that is opinion, with academics disagreeing. Some errors would really be wrong – at one stage in the writing process I found I had written customer when I meant supplier, and I’m glad to have spotted that! But most are not even likely to be noticed! At PMI we found an error in one of our manuals recently that can be traced back to the early 1990s. Nobody had spotted it before. This is a manifestation of the age-old question ‘if a tree falls in the forest, but no-one is there to hear it, does it make a noise?’ which I explore at some length in the book.

None of which means that we have not tried very hard indeed to eliminate errors, but no doubt some will have slipped through. Most readers will have examples of how inspections fail to pick up errors, so I must brace myself for being informed of some that we missed, and to figure out how to respond!

Conclusion: the most important characteristics are unknown, and probably unknowable

This was one of Dr Deming’s many infuriating claims in the 1980s that would cause a regular manager (as I was in those days) much frustration. Only after a lot of reflection did it make sense. We had all been captured by the finance discipline, so that we were used to putting everything in numbers. In fact one particularly useless cliché that still survives is: ‘If you can’t measure it you can’t manage it’. Dr Deming said that was wrong, and that was hard to take. But he was right of course. Numbers are an abstraction of reality, not the reality itself. The appeal of a Lexus over a BMW, or vice versa, is a matter of opinion, worth billions, but not actually measurable.

So I hope my book appeals enough to people to inspire them make their work work better, and that they also come up with ideas that will help me make the second edition better still.