Episode 13

Bryon Kroger (00:05):

Welcome back. We've spent this entire series on how to assemble a system for delivering software, transforming culture, and aligning a massive enterprise. But to what end? The answer's really simple. The goal of Mission O/S, the entire point, is to build an organization that can learn. In today's fast moving and competitive world, the organization that can learn the fastest is the one that wins. It's [00:00:30] not the one with the biggest budget, the most people, or even the most brilliant initial plan. It's the one that can sense and respond to a changing reality more quickly than its adversaries. This is a profound and admittedly difficult shift for most large organizations because traditional enterprises are not optimized for learning. They optimize for being right. So think about it... the entire bureaucratic system is built on the assumption that we can predict the future. We spend [00:01:00] years writing detailed requirements documents.

(01:02):

We build five-year plans and we create a complex financial model all in an attempt to de- risk the future by being right on paper. And the system rewards those who create the most convincing plans and it punishes the people whose predictions turn out to be wrong. But at a time when software is eating the world, where there's this constant change and disruption, this model is a total failure. The cost of being wrong is [00:01:30] too high and the pace of change is too fast. So the five-year plan is obsolete in a few months, if not a few days. The organization optimized for being right is too slow, too brittle, and too afraid to adapt. So we have to change the objective. The goal is no longer to be right and to execute the plan. The goal is to be less wrong over time. And the only way to do that is to embrace learning as your core competency.

(01:59):

It's to [00:02:00] embrace experimentation. This isn't some philosophical preference. It's a strategic necessity. And it has one non-negotiable prerequisite that we've discussed at length already. You cannot become a learning organization until you've achieved continuous delivery. Why? Because real learning only happens when an idea makes contact with reality. You can have the most brilliant minds in a room debating an idea for a year, but it's all just theory and potential [00:02:30] until you put it in front of a real user in their real environment doing their real job. That's the moment of truth. Continuous delivery drives this contact with reality. It's the technical capability that allows us to take a hypothesis, like an idea for a new feature or a change to a workflow and test it in production safely, quickly, and sustainably. So without continuous delivery, you're not a learning organization. You're a debate society. You're trapped in a world of theory, unable to generate the evidence you need to [00:03:00] make intelligent decisions.

(03:02):

And that's why, as I've said from the very beginning, establishing that well-oiled IT system is the first and most important step. Once you have that, you need a process to use it. You need a scientific framework for learning. And that framework is the fourth and final step of the Improvement Kata. Experiment towards the target condition. Now, experimentation has become a watered down buzzword, but in Mission O/S, a true experiment [00:03:30] isn't just trying stuff. It's disciplined scientific process and it has three key components. First, it has a clear testable hypothesis. We use a very simple format. We believe that building these outputs, will drive the following user or system outcomes, and lead to mission impact. Here, outputs means the solutions we plan to build. Outcomes, of course, are the changes in user or system behavior, and the mission impact is the [00:04:00] ultimate result generated. And this forces us to be incredibly clear about what we're doing and why.

(04:07):

Second, it has a definition of done. How will we know if our hypothesis was right or wrong? What data will we collect? What change in the user's behavior or what mission metric will prove our assumptions? If you can't measure it, it's not an experiment. And third, it is the smallest possible change required to test the hypothesis. So [00:04:30] we're not trying to complete the whole journey. We're trying to build just enough to learn if we're even on the right road. The micro routine we use for every single experiment is the classic build, measure, learn cycle. You build the outputs from your hypothesis, measure the outcomes and impact, and you use that to learn what you need to do next. So did the data support your hypothesis? If yes, you persevere, you double down. If not, you pivot, you try a different approach, or in some cases, [00:05:00] you might need to abandon the idea altogether because you've learned it as dead end.

(05:04):

And you want to find that out sooner than later in the enterprise. This cycle though: hypothesis, build, measure, learn, it's the heartbeat of a learning organization. It's how you systematically reduce risk and climb the ladder of value one rung at a time. But there's one final crucial ingredient. You can have the best technology and the best processes in the world, but you will not become a learning organization without a culture [00:05:30] where it is safe to learn. So experimentation by its very nature means accepting that you're going to be wrong. A lot. The best product companies in the world report that 50 to 70% of their experiments fail to produce the desired result, depending on the idea stage. So in a traditional enterprise culture, that's a 50 to 70% failure rate, and it's a career ending statistic. If being wrong is punished, people stop taking [00:06:00] the risks necessary to innovate.

(06:02):

They'll only work on the safest, most predictable and least valuable ideas. They'll optimize for not failing instead of optimizing for learning. So they follow the plan. And as long as you measure success as following the plan, people will keep making and following plans, whether or not they're valuable. To build a learning organization, leaders have to fundamentally change their role. You're no longer the judge who evaluates plans [00:06:30] and punishes failures. You're the lead scientist, right? You create the conditions for good experiments. Your job is to foster an environment where hypotheses are encouraged, where the data that comes out of them is revered and where a failed experiment is not a mistake to be condemned, but a valuable piece of learning to be celebrated. In fact, we like to say the only failed experiment is the one with inconclusive data. As long as we get data out of it, whether we're right or wrong about the hypothesis, [00:07:00] that's a successful experiment.

(07:02):

We have to change the reward. Stop celebrating teams that perfectly execute a year long plan without ever checking to see if it was valuable to Warfighters or the mission. Start celebrating the teams that run a two week experiment and prove a foundational assumption was wrong, saving the organization millions of dollars and years of wasted effort. Celebrate the teams that actually ship outcomes and produce mission impact. This is the ultimate goal... to [00:07:30] build an organization that's not afraid of the future because it has the capability, the process, and the culture to learn and adapt to whatever comes next. An organization that doesn't just execute, but evolves. One built on continuous improvement, not being right. None of this happens by accident. To build a learning organization, you need leaders who know where they're going and why. So in our next episode, we'll break down how a clear vision and a living strategy [00:08:00] become the compass for your entire transformation.

‍