Contrasting Observation and Intervention Programs - Treating Symptoms vs. the Cause


Our loyal readers are quite familiar with our 2010 research into safety interventions in the workplace and the resulting SafetyCompass® Intervention training that resulted from that research. What you may not know is why we started that research to begin with. For years we had heard client after client explain to us their concerns over their observation programs. The common theme was that observation cards were plentiful when they started the program but submissions started to slow down over time. In an attempt to increase the number of cards companies instituted various tactics to increase the number of cards submitted. These tactics included such things as communicating the importance of observation cards, rewards for the best cards, and team competitions. These tactics proved successful, in the short term, but didn’t have sustainable impact on the number or quality of cards being turned in. Eventually leadership simply started requiring that employees turn in a certain number of cards in a given period of time. They went on to tell us of their frustration when they began receiving cards that were completely made up and some employees even using the cards as a means to communicate their dissatisfaction with their working conditions rather than safety related observations. They simply didn't know what to do to make their observation programs work effectively. As we spoke with their employees we heard a different story. They told us about the hope that they themselves had when the program was launched. They were excited about the opportunity to provide information about what was really going on in their workplace so they could get things fixed and make their jobs safer. They began by turning in cards and waiting to hear back on the fixes. When the fixes didn’t come they turned in more cards. Sometimes they would hear back in safety meetings about certain aspects of safety that needed to be focused on, but no real fixes. A few of them even told us of times that they turned in cards and their managers actually got angry about the behaviors that were being reported. Eventually they simply stopped turning in cards because leadership wasn’t paying attention to them and it was even getting people in trouble. Then leadership started giving out gift cards for the best observation cards so they figured they would turn a few in just to see if they could win the card. After all, who couldn’t use an extra $50 at Walmart? But even then, nothing was happening with the cards they turned in so they eventually just gave up again. The last straw was when their manager told them they had to turn in 5 per week. They spoke about the frustration that came with the added required paperwork when they knew nobody was looking at the cards anyway. As one person put it, “They’re just throwing them into a file cabinet, never to be seen again”. So the obvious choice for this person was to fill out his 5 cards every Friday afternoon and turn them in on his way out of the facility. It seemed that these organizations were all experiencing a similar Observation Program Death Spiral.

The obvious question is why? Why would such a well intentioned and possibly game changing program fail in so many organizations? After quite a bit of research into these organizations the answer became clear, they weren’t intervening. Or more precisely, they weren’t intervening in a very specific manner. The intent of observation programs is to provide data that shows the most pervasive unsafe actions in our organizations. If we, as the thought goes, can find out what unsafe behaviors are most common in our organization, then we can target those behaviors and change them. The fundamental problem with that premise is that behaviors are the cause of events (near misses, LTA, injuries, environmental spills, etc.). Actually, behaviors themselves are the result of something else. People don’t behave in a vacuum, as if they simply decide that acting unsafely is more desirable than acting safely. There are factors that drive human behaviors, the behavior themselves are simply a symptom of something else in the context surrounding and embedded in our organizations. Due to this fact, trending behaviors as a target for change efforts is no different than doctors treating the most common symptoms of disease, rather than curing the disease itself.

A proper intervention is essentially a diagnosis of what is creating behavior. Or, to steal the phrase from the title of our friend Todd Conklin's newest book, a pre-accident investigation.  An intervention program equips all employees with the skills to perform these investigations. When they see an unsafe behavior, they intervene in a specific way that allows them to create immediate safety in that moment, but they also diagnose the context to determine why it made sense to behave that way to begin with. Once context is understood, a targeted fix can be put into place that makes it less likely that the behavior happens in the future. The next step in an Intervention Program is incredibly important for organizational process improvement. Each intervention should be recorded so that the context (equipment issues, layout of workplace, procedural or rule discrepancies, production pressure, etc.) that created that behavior can be gathered and trended against other interventions. Once a large enough sample of interventions is created, organizations can then see the interworking of their work environment. Rather than simply looking at the total number of unsafe behaviors being performed in their company (e.g. not tying off at heights) they can also understand the most common and salient context that is driving those behaviors. Only then does leadership have the ability to put fixes into place that will actually change the context in which their employees perform their jobs and only then will they have the ability to make sustainable improvement.

Tying it back to observation programs

The observation program death spiral was the result of information that was not actionable. Once a company has data that is actionable, they can then institute targeted fixes. Organizations that use this approach have actually seen an increase in the number of interventions logged into the system. The reason is that the employees actually see something happening. They see that their interventions are leading to process improvement in their workplace and that’s the type of motivation that no $50 gift card could ever buy.