Discover How Experiments Boost Learning… By Failing
Giving the right to make mistakes has always sounded… well, risky. As a leader, how do you actually do that? And why should you be open to making mistakes in the first place?
In this age of volatility and uncertainty, organizations have to become agile, which also means developing ways to “try stuff” because they can’t presume on the solutions or how to bring them to customers. In other words, they need the capability of experimenting.
“The principles of experimenting are fairly simple but creating the conditions to do so is not.”
The principles of experimenting are fairly simple but creating the conditions to do so is not, particularly in organizations that have been relying on traditional product and project management for years. The good news is that there are ways to start experimenting and develop this capability without transforming the whole
We’re going to briefly explore four enablers of experimenting.
Contrarily to a project, the goal of an experiment is not to deliver a solution but to answer a question. This question serves to validate or invalidate hypotheses that would make it worthwhile (or not) to deliver a solution later.
The most common type of hypothesis is the value hypothesis. Testing the value hypothesis consists in verifying whether the future solution would indeed bring value to users. Users can be customers, employees or any other stakeholder depending on the solution domain.
Consequently, the definition of success of an experiment differs from that of any other initiative: not only validating but also invalidating a hypothesis is a success. The latter case because it would avoid investing in an initiative that would likely not produce the expected benefits.
It would be tempting to experiment by analyzing or simply by surveying users. There too, lies a big difference between experimenting and classic initiatives. Experimenting requires feedback from real users interacting with a prototype developed specifically for the hypotheses to (in)validate. This is a critical distinction because we know that real user feedback often provides very different and more accurate results than analysis, surveys or any second-hand information would.
Similarly, executing an experiment is very different from executing a traditional initiative such as a project. An experiment has a very small fixed budget, a very short fixed timeframe, and a few dedicated people working entirely autonomously within the experiment’s boundaries.
Although experiments can be derived from existing projects, they are usually placed upstream in the value creation flow because the result of experiments will typically determine whether a project will be initiated or not.
“The leader’s role is to create the environment and mindset supporting experimentation.”
The leader’s role is to create the environment and mindset supporting experimentation.
This builds on servant-leadership, a style of leadership that fosters self-organization, courage, the right to make mistakes and developing people’s potential, among others.
Related content: here is a great article on Authenticity, one of the traits of the servant-leader.
Similarly, experimenting, together with ideation and the ability to scale up successful experiments, is part of a broader intrapreneur role that is championed by agile leaders.
Additionally, to encourage the behaviors that make experimenting possible and ingrain them in the longer-term, leaders often have to adapt the evaluation and reward system. Indeed, although most people are more than willing to experiment, they will just as quickly revert to their formal job description when a crisis or a feeling of personal risk occurs.
Experimenting, just as agility and innovation, is not reserved to a few. Of course, not everyone will be actively involved in experiments, but leaders must ensure that it remains an inclusive process that can tap into a large talent pool. Actually, one of the reasons for developing the capability of experimenting is to better engage workers and attract talent.
By now you probably understand that good intentions and some resource re-allocation won’t be enough to generate a sustainable capability of experimenting. It also requires developing values, skills, practices, and methods.
For people participating in experiments, we consider mostly practices and methods such as design thinking and lean change management, in addition to tools related to specific technologies such as AI, data analytics or even drones depending on the experiment’s domain. Together, these practices should allow running the experiment in a kind of “lab.”
The role of leadership is particularly critical. For leaders, we consider skills such as agile leadership, creative leadership, and more classically management of innovations as key to developing this capability and, as a leader, to gain the legitimacy to lead such efforts.
Note that in this discussion we do not equate leadership with reporting hierarchy. A leader is anyone who has an influence on others’ beliefs and behaviors. Basically, everyone has some leadership. Interestingly, one of the goals of servant-leaders is to create other leaders (and not followers as many would believe).
Why is culture our last point? Well, because in my opinion, successful change agents don’t talk much about culture. They encourage others to try things and, when the new ways become a habit, then culture will have changed. You can’t affect culture directly. Moreover, talking about changing cultures usually raises a lot of barriers and fears, which will hinder the change.
That being said, yes experimenting is changing culture by developing the right to make mistakes and to fail, thus learning at individual and organizational levels. The capability of experimenting is one of the pillars of developing an adaptable organization that will morph progressively and never need a “transformation program” again.
Watch the extended video version of this article.
Developing the capability of experimenting is, in itself, an experiment. What are the minimal questions you have to answer, what is the minimal framework to set up before you can start experimenting and learn along the way? What are your hypotheses to validate?
And of course, does everyone understand why experimenting is important?
Bruno Collet, MBA, MScIT, PMP, ACP, ICP-ENT