Five Ways to Accelerate Root Cause Investigations

Finding the root cause for errors and preventing them from happening again is always good business.  Your quality improves every time you prevent an error.  Costs go down.  And in some industries, like pharmaceuticals or medical devices, it's more than good business.  It's the law. 

Why then do we managers do such a poor job of investigating errors?  Let's examine a typical case. 

Jim is a new manager.  Something has gone wrong and everybody knows it.  Operations have come to a halt and it's his responsibility to investigate.  He has to find out why this error occurred and make sure it doesn't happen again. 

Being a typical manager he's been locked in a meeting room all morning, and by the time he gets back to his desk his in-box is exploding.  He tries to remain calm but his employees know he's not happy.  They are head down and suddenly very interested in their work, wondering who is going to be interrogated. 

He selects his quarry and the chase begins.  "Where were you when it happened?" he asks.

"Oh, I was on break.  I didn't see anything," comes the answer.  And this victim manages to escape.

He pounces on another employee.  "What do you know about it?" he asks.

"I just followed the instructions as they were written.  I don't know anything," she says.  This quarry, too, escapes this day. 

The chase continues until an inexperienced employee seriously thinks about his questions and answers truthfully.   "Well, I saw something unusual.  It seemed trivial at the time and I really didn't pay any attention to it." 

Jim sees his chance and takes it.  He scribbles on his notepad: "Employee needs to be more alert.  Assigned to be retrained." Then he strides back to his desk to write his report, relieved that he could cross that job off his list so quickly.

But can Jim really cross it off?  Not likely.  It will surely come back again.  Nothing really has changed except that the mind of one thinking employee has now been turned off.  That's not all.  Other employees who might have added thoughtful insights to the problem observed the interchange.  And although they appeared to be so very absorbed in their work, they were learning a lesson, don’t get involved.

Could this be you?  When you go out to investigate errors in your organization, does your best source of information about the failure dry up because the employees fear repercussions?  If this sounds familiar, can it be surprising that the same errors keep coming up over and over again? 

Do you want to avoid this happening to you?  Do you want to convince the employees to actively help you find solutions rather than to passively resist?  The key is in recognizing that fear is motivating their behavior.  Here are five time tested ways that will make any manager's life a little easier.

1. Brainstorm, don't blamestorm.  When you investigate errors, focus on repairing the business process, not the people.  Frame your questions around the assumption that there was a weakness in the system and some unlucky employees happened to fall into that trap.  You should rally the employees to work as a team to figure out how the weakness occurred so it doesn't happen to them the next time.

After you have found the root cause of the error you need to implement corrective action to eliminate it so that the error never happens again.  Always try to make your corrective action an engineering or procedural change.  Constantly work to make your business process more resistant to input variability.

These kinds of corrective actions are much more effective and permanent than personnel indictments.  And they have the added advantage that they don't create more fear in the employees.  "Employee needs to be more alert" or "Employee assigned to be retrained" should be last resorts as corrective actions.

2. Develop a Standard Operational Procedure for investigating errors.  If you have an SOP for investigations, then the employees know what's going to happen.  Predictability allays fear.  When you're asking all those questions, you're simply following the procedure.

The procedure doesn't have to be complicated.  Start with the five W's:  Who, What, When, Where, Why.  Don't use names when filling in the Who.  Use position titles. 

The SOP should require that the investigator find the root cause for the error if at all possible.  Corrective actions should be assigned to individuals with target dates.  The SOP should also include means of verifying the effectiveness of the corrective actions.

Ask why five times.  You usually don’t get to the real root cause until uncover four layers of cosmetic solutions. 

Here’s an example:

1. Why does the Washington Monument cost so much to maintain?  

        Frequent washing is damaging the monument.

2. Why does it need to be washed so much?

        Pigeon droppings.

3. Why are the pigeons on the monument?

        To eat the spiders.

4. Why are there spiders on the monument?

        To eat the insects.

5. Why are the insects there?

       They are attracted to the brightly lit surface.

Next time:  the final three ways to accelerate Root Cause investigations, including the most important foundation for gathering critical information about the incident, Driving Out Fear.