Monday, June 04, 2007

Why Inspections Fail

A few weeks ago, we had an important document circulate recently to various managers.  We were each to review and sign as approving.  There were 8 or 9 signatures as I recall, mine being one of them.  We all signed off, which then triggered some significant purchases.  A week later, however, one of the original signatories (not me!) serendipitously discovered an error.  An obvious error.  One which, when pointed out, made all of us slam our foreheads with our palms and imitate Homer Simpson.  It was as clear as a few mispeled wurds.  But we all missed it.  And it cost us money and time.  Not an insignificant amount.
Last week, another document circulated, this time containing a list of numeric specifications.  Four names were on this list to sign off.  One person did sign as being correct.  The second person caught an obvious error. 
These two events got me thinking; how did this happen?  How did 9 otherwise intelligent, observant people just plain blow it so badly on the first document?  How did one guy miss an obvious mistake on the second?  And, further, what does this tell me about the inspection process in general?
Two things came to mind. 
First, it was not precise what each signatory was to look for.  The general question was "is the document right?" but we were not crisp or clear in defining what "right" was.  Thus, we ended up violating Philip Crosby's first quality principle:  Quality is conformance not goodness.  We had nothing against which to compare conformance; thus we only assessed goodness.  And, apparently, it was "good enough" for all of us to sign off. 
Interestingly, my colleague found the error on the second document when (gasp) he pulled out the original specifications and compared, one by one, the new document with the authoritative one.  Boom.  The error jumped out and we could take action to correct it. 
Second, with 8 signatories, no one person really will take ownership.  There is the very human tendency to see "if everyone is responsible, then no one is responsible."  Conventional wisdom might say more inspectors will improve the product.  But, clearly, it made it worse.  Crosby's second principle said that prevention, not inspection, was the key to quality.  We actually introduced waste by having all these people sign off and delay the document.  We would have been better off with fewer signoffs with clear criteria to get a conforming document, quickly. 
Keep learning.


Anonymous said...

I found a really useful Lean healthcare blog at

Tom Hopper said...

I have encountered the same situation at work. Lots of people have to sign off on engineering change orders, to approve the implementation of the change in manufacturing. No one's role or responsibility in this is defined.

I adopted the position that my responsibility was based on my job description--ensuring conformance to the product development process (have the proper engineering documents--beside the prints--been updated? Have the cost and timing impacts been assessed?). After months of working with Engineering to try to get them to update the relevant documents with any change order, I finally threatened to refuse to sign-off on the grounds that the work was not complete. I nearly lost my job.

I think that the lesson is that authority must be spelled out, with example decisions. Any organization that refuses to do this will find itself in serious trouble.