Anatomy of a Turnaround - Part 5: Process Control

Process Control is a process. It is one of the most vital processes in any organization. It is also widely misunderstood.

I would also argue that being good at Process Control is more important than being good at process design or Process Improvement. To illustrate this, allow me to lead with a case study.

The Process Audit Database Case Study:

Some years ago, I had a quality manager approach me with the need to show improvement on our key quality metric. This metric was based upon issues with product and/or process noncompliance, we called it "Deviations."

It was essentially a log of errors and non-compliant practices that the quality team had found within a given month.

We had a list of what deviations were most prevalent, and it generally matched the pareto rule, a few processes were generating most of the deviations.

I started my effort with the assertion that "People fail to comply with a process for one of two reasons: 1) They don't know the standard or 2) they think nobody cares if they comply with it.

The second part of this gets a little argument from time to time. I would reference Pearson's Law to back it up: "That which is measured improves; that which is measured and reported improves exponentially." 

So, if that's true, then all I have to do is check if people know the standard, record the result, and ideally make the auditing experience a positive experience. Don't make them sweat, don't weaponize the results. No consequences. If they don't know something, just retrain them and thank them for their time. 

So I wrote a database that allowed me to enter an unlimited number of questions for each role. However, it only presented 5 of those questions at a time to be asked and it both recorded the questions/answers as well as rotated them so it would be different questions each day. A quality inspector would walk the floor in a pre-determined route and interview each player on the team using the database to ask questions and record how the operators (and other players) did. If they got a question wrong we recorded that fact, but then we just retrained the operator and kept moving.

Now, some questions were to check the state of something or whether a task had been completed correctly (cleanliness of equipment, paperwork filled out properly), but an awful lot of them centered on critical process knowledge. 

Validation tasks are a good example of a hard-to-audit task that looks like it's done on paper but which people will skip for years without making an error. If you aren't auditing these types of steps you can be almost certain they're being pencil whipped. 

The general approach I used for critical process knowledge was to ask 3 questions. 1) How do you do it? 2) Why do we do it that way? 3) Do you actually do it in accordance with the process?

Almost every time I tell someone this, I'm told that the employees will just lie. But that simply is not what happened. 

They may well lie if you just ask one time. I don't know for sure, but I designed the database to catch liars. 

I designed the audit database so that when someone said, "I haven't been trained," after they were found doing something incorrectly, I'd be able to retrieve the 5 times in the last 6 months that someone asked them that critical step and they explained it correctly. However, in more than 2 years of running this process in a pretty large facility that never happened. 

We never found a single instance of a person saying, "I do the process correctly," and then later found evidence that they weren't. 

Interestingly enough, though, we did find some instances of people who could not learn a job. By doing the individual questions on a rotation, every few weeks the questions would repeat. We found several individuals who would miss the same questions every couple weeks because they simply couldn't retain the information.

I think we found 3 or 4 people who couldn't retain the critical information for jobs they held and who ultimately had to be moved to positions where they could be more successful.

The absence of such an auditing program lines you up with failure to detect this issue or forces you to deal with it as a discipline issue which it is not. The humane thing to do is to verify this is going on, talk about it with the person, and then arrange for that person to move to a job that lets that person be successful every day. 

A last note on this auditing process. The operators loved it. Too many businesses hire an employee, give them 2 weeks of shadowing someone, and then abandon them forever on a machine with no further attention or assistance. The operators liked the attention and liked that someone cared how well they were doing their jobs. And with just a little prep and attention, they could prove how well they were doing.

I couldn't tell you how many times an operator stopped me and told me they hadn't missed an audit question for a week.

In fact, because the operators knew they were getting audited they would prep for everything that they expected they might be asked. I remember a specific labeler that always had label residue on it (which annoyed the quality manager). Once we added an audit to make sure it was clean it was spotless from then on because the operators started checking/cleaning it daily in their audit prep.

So how did the quality portion of this auditing experiment work out?

In the first month our rate of deviations dropped by two thirds and that drop sustained to the point that "deviations" had to be redefined because we weren't producing enough of them to make the metric relevant. 

I still held the keys to the database, so I started putting corrective information for all other problems into the database. Every problem I found the solution to and put into the database went away.

We had multiple individual lines that more than doubled output over the 2+ years we ran that audit system.

I also rebuilt the safety team around an audit based system using essentially the same software. Stunning success. World class results. 

Noteworthy technique: I had the safety team performing Job Safety Analysis (JSA) check-sheets to generate audit questions. I just did it because when we were standing up the process, we didn't have enough audit questions. The process was so successful at fixing real problems that employees started training each other to do JSAs outside the safety team and turning them in so I would generate the audits. I have never seen this much excitement or interest in JSAs.

So, why do I say that process control is more important than process?

If you don't have a process control system, your process doesn't matter. People are all doing whatever they know, understand, and feel-like.

Realize that people run off habits. If nothing interrupts them and everything is going well, they will repeat a habit with muscle-memory precision over and over forever. If it's the wrong habit, you'll get persistent poor results.

Changing a habit is hard and it takes a person committing to do it. I believe this style of system is the gentlest way to change habits because it influences them to change their habits without fear or frustration.

When you change a process and ask everyone a single time to change, I do not believe that is a successful way to do it. Making someone read an SOP, sign it, and then try to do the work, also not a good way to do it.

However, the operators learn the audits because they want to be measured and found to be good. So they put in thought and preparation and that makes them able to change without high levels of frustration (a critical error state) because at that point they're managing their own change processes.

Shortly: If you change a process and tell everyone once, you'll get inconsistent change. If you have no written process at all, but write audits around clear expectations and critical knowledge you will successfully guide a team and create positive change.

Thus, process control is a more critical process than even having a written process. However, you still need both.

Note: I recommend against ever weaponizing auditing. You need your people to commit to the creative venture that building a world class organization really is. Coercion isn't an important part of that plan.