Safe-to-fail probes
Safe-to-fail probes are experiments whose consequences are limited enough that failure can be accepted. They let people explore a complex setting, observe what happens, and learn without assuming that every attempt will succeed.
THE IDEA
Safe-to-fail probes
Safe-to-fail probes are experiments in complex settings, designed so that an unsuccessful result has acceptable consequences. Learning depends on observing those results, not on ensuring that every probe succeeds.
Dave often references the Apollo 13 film when he talks about innovation. Technical problems prevented the planned lunar landing, and the crews in space and on Earth had limited equipment with which to respond. But Dave’s reference to the film is also known for the famous quote ‘failure is not an option.’ So famous did it become that NASA even put it on T-Shirts and souvenir mugs. From my perspective, it matches my experiences of running multiple SenseMaker exercises in Welsh Public Services between 2009 and 2016.
One of my key challenges and learning experiences during this time was getting so far and then having the project stall. I would talk with a Snowden enthusiast (the man not the mountain) who had seen Dave speak and wanted to do more. Then there’d be a session with a group of middle/senior managers in the organization and the commitment to use SenseMaker. We’d develop a set of SenseMaker signifiers and suitable prompt questions to look at organizational culture, staff engagement, or how to use staff ideas for improvement.
The final SenseMaker survey would go to senior management for sign off and then – get shelved. I’d be at a loss to understand why? With a bit of digging and the passage of time, it became clear to me that senior leadership wasn’t prepared to deal with what SenseMaker might tell them. Using SenseMaker, they were going to open the door to answers they would struggle to deal with. SenseMaker reveals stories that describe what is happening in an organization. It is an insight into our ‘water cooler conversations.’ Not really wanting to know would stall the project.
I remember a brief but peak conversation with Dave about things not working out. We were at a Cardiff Blues game, and I was agonizing about my failures. Dave said calmly and in a matter of fact way, “But things will fail. You are probing in a complex space. Not everything will work.” And that was the end of that particular conversation. The recognition that some things will fail trying them out in a complex space seems obvious now. However at the time, working for the Wales Audit Office in Public Service Improvement, my world of Audit could not imagine failure as an option. We were only just entering a time when ‘safe-to-fail probes’ were a possibility. Slowly however ‘safe to fail probes’ became translated into ‘audit speak,’ and the Auditor-General for Wales started declaring, “I encourage Public Bodies to take well-managed risks.” Small but significant changes for sure.
When to use it
Trying a new process in a single department before rolling it out company-wide.
Conducting a small pilot study for a new public service offering before full implementation.
Using a prototype to test customer reactions and iterate based on feedback.
Testing a change in company policy in one branch to gauge employee response.
FROM THEORY TO EVERYDAY LIFE
In practice
Stalling SenseMaker Projects
A team developed a SenseMaker survey to understand organizational culture and staff engagement. It went to senior management for approval and was then shelved. The author concluded that leaders were not prepared to deal with the uncomfortable insights it might reveal.
Pilot Product Launch
A company launches a new product feature in a small market to test customer response, explicitly recognizing that the probe may fail. The failure is treated as valuable data for iteration, not a disaster.
A MOMENT TO REFLECT
Put it to work.
What experiments in your organization are considered safe to fail? How do you define 'safe' in that context?
When have you encountered resistance to probing because of fear of failure? How could you reframe the probe to overcome that resistance?
How can you design a probe so that even if it fails, the learnings are captured and shared effectively?
Often confused with
CONTINUE THE CONVERSATION
Source texts
- Cynefin - Weaving Sense-Making into the Fabric of Our World ED02 (Dave Snowden Sonja Blignaut Zhen Goh)