Episode 08

Bryon Kroger (00:05):

We've now covered the entire Mission OS stack. We have clear goals, a rapid path to production powered by a cloud native platform, and a methodology for building applications that users love. But none of it works without people. And when you bring people into the equation, it requires a discussion of the most misunderstood but critical factor in any transformation. And [00:00:30] that's culture. Culture change has become a buzzword in every boardroom. Leaders spend millions of dollars on consultants who come in, run workshops and help them craft new mission and value statements. They send their entire workforce to two-day agile training courses. They put posters up with inspirational quotes. And nothing changes. Because they're all operating under a fundamentally flawed assumption. They believe that you can change a culture by first changing how people think. [00:01:00] They believe that if they just give a powerful enough presentation, people will have an epiphany, adopt a new mindset, and then their behavior will magically change to match it.

(01:10):

This has never worked and it never will. The Mission OS philosophy, the secret to real lasting culture change, is to do the exact opposite. You don't start by trying to change thinking. You start by doing. You change behavior first, and when people see the results, that changes their values [00:01:30] and attitudes, and collectively that changes the culture. And the best story to illustrate this doesn't come from the world of software, but from the factory floor of a shuttered auto plant in Fremont, California. And that's the story of NUMMI. So in the early 1980s, General Motors had a plant in Fremont that was, by all accounts, the worst automobile factory in the United States. The culture was toxic, and the union and the management were in a constant state [00:02:00] of war. Absenteeism was so high that on any given Monday or Friday, they often didn't have enough workers to run the assembly line.

(02:09):

Drug and alcohol use on the job was rampant. Workers would even intentionally sabotage vehicles, leaving like bolts loose or welding up a Coke bottle inside of a door panel so that it would rattle and annoy the customers. This is a complete disaster, right? The culture was so bad that GM finally gave up and shut down the entire plant, firing everyone. [00:02:30] Two years later, Toyota wanted to enter the American manufacturing market, and they decided to form a joint venture with GM who graciously gave them their shuttered Fremont plant. They called this joint venture new United Motor Manufacturing Incorporated or NUMMI. And here's the most important part of the story. While the management was all new, Toyota insisted on rehiring the majority of the same workforce. The same people GM had declared unemployable [00:03:00] due to their toxic culture. The GM executives thought Toyota was insane, but Toyota understood something profound about culture.

(03:08):

They didn't bring the workers back and just give them a lecture about Japanese philosophy. They didn't make them recite new value statements. They just taught them a new way to work. The American workers were immersed in the Toyota production system. They were taught to work in teams. They were taught to focus on quality above all else. And most famously, they were [00:03:30] given the Andon chord. A rope that ran above the assembly line that any worker could pull at any time if they saw a problem. Now, pulling that cord would stop the entire line and a manager would come. But not to yell at them like you might see in an American plant, but to thank them for identifying a problem and to help them solve it. Pretty novel concept, right? But they were placed in a system that was built on respect, teamwork, and continuous improvement.

(03:58):

They were given a new set of [00:04:00] behaviors. And what happened next was a miracle that changed the American auto industry forever. Within a few short months, that same workforce, the one that we would all agree has an absolutely terrible culture, the one that GM had written off as lazy and defiant. They now created the highest quality, most productive auto manufacturing plant in the United States. The problem was never the people. It was the system. When you change the system people work in, when you change their behavior... [00:04:30] their thinking, their values, and their attitudes will change to match. The new culture becomes a byproduct of the new way of working. I call this the NUMMI maneuver, and it is the absolute core of how we approach culture change and government technology. We do not sell culture change workshops. We don't write value statements for our clients.

(04:51):

We build and deliver software. But we do it in a way that is a deliberate NUMMI maneuver. Our approach for this is pairing. [00:05:00] So when we begin an engagement, we insist that our government clients don't just act as stakeholders. They become members of our team. We even pair our product managers, our designers, and our engineers with government counterparts whenever feasible. And just like Toyota did, we immerse them in a new way of working. Pairing is our Andon cord. It's continuous integration and the opportunity for anyone to stop the line, solve the problem, and improve the process without fear. [00:05:30] So a government program manager who's only ever known waterfall development and monthly status reports is now in a daily standup, participating in weekly retrospectives. They're in the room for user interviews, hearing directly from the people they're supposed to be serving. A government engineer who's only worked in a silo, handed a requirements document and told to code, and then throw that over the wall for testing ,and cybersecurity, and delivery, is now pair programming with one of our engineers.

(06:00):

[00:06:00] They're learning test-driven development. And they're seeing their own code get deployed to production continuously, in a matter of hours, not years. We don't preach to these people about Agile. We don't give them a certificate. We just invite them to work with us in a different way. And what happens is exactly what happened at NUMMI. They see the results. They see that this new way of working actually produces better outcomes. They see that shipping software is possible. They see users who are happy and engaged. And [00:06:30] seeing those results begins to change their values and their attitudes. They start to believe that a different way is possible because they've experienced it firsthand. So culture change isn't the goal. Mission impact is the goal. Culture change is the trailing indicator of a successful transformation of the work itself. So what does this mean for you as a change agent or transformation leader?

(06:54):

It means you can't wait for your organization to grant permission to have a better culture. You have to create [00:07:00] it. Change behaviors to change values, attitudes, and beliefs. Your first balanced team is your NUMMI catalyst. It is a new way of working in the middle of the old enterprise. Your job is to fill that team with innovators, protect it, immerse your customers in it, and use it to deliver a tangible, undeniable success story. That success story is your proof. It's the evidence you use to get the resources to build your second team, and then your third. [00:07:30] And you don't change the enterprise by fighting it head on. You change it by creating pockets of a better future, and then radiating that change outwards until the new way becomes the default way. So I would urge you to stop talking about culture. Start changing the work.

(07:48):

That is the NUMMI maneuver. And then ask, where do we aim this new mindset? Culture is the spark. Alignment is the fuel, and that's what we'll tackle next.

(08:00):

[00:08:00] One example of an Andon cord is how we treat continuous integration and continuous delivery. We build secure release pipelines, and we have trunk-based development and test-driven development. And when you put these things together, it functions as a way for teams to quickly identify when an issue has emerged, [00:08:30] stopping development so that the team can fix that issue before moving on. You say, how is that different than normal development? Well, in a lot of teams where they don't practice trunk-based development, they have long-lived feature branches, there's breaking changes that are being introduced all the time, but you don't even know about them. Or sometimes maybe they're running CI tests, things are breaking, but people don't stop and fix the issues before moving on. They wait until the very end of a big batch release to then work out all those issues. [00:09:00] And much like you see on a manufacturing floor, that creates a graveyard of code, or a graveyard of unfinished vehicles ,with all kinds of quality issues that are hard to resolve after the fact.

(09:15):

And so we would rather fix that now and get it fixed for every future team and every future merge so that we don't have to deal with it later. And that's one example of the team is given the space to [00:09:30] stop and address quality issues, and they're not yelled at for introducing a breaking change. And so these are the kinds of practices that you can start implementing on a technical level to show teams that you trust them, and to build trust with your government partners.

(09:53):

Pairing is a big part of our NUMMI maneuver at Rise8. And one of the first times that I paired with a government partner [00:10:00] at Rise8 was actually a organization that had replicated a lot of the success Kessel Run had... at the application level and even at the platform level, but they were having difficulty getting to that scale part. And it was really, at that point, the leadership needed to grow. And so I did a lot of pairing personally with the leadership, and we implemented new ways of working at the leadership level. Now, the initial reaction was [00:10:30] maybe a little bit skeptical. And one of the first things I wanted to do, just like we do with software, is see a change, produce a result really quickly. And one of the things they were having trouble doing was getting teams aligned on these bigger outcomes that their customer of customers wanted to see, the commander of the organization they were serving.

(10:54):

And so I ended up pairing with the leadership to work directly with that commander to develop a set of OKRs, [00:11:00] get the teams aligned behind those OKRs, and deliver some of the key results within one month. And we kept doing that month over month. And by the end of the quarter, we had delivered on some pretty big outcomes, and we kept it going quarter over quarter. And by the end of the first year, we had completely transformed what was a lot of really great application development that was having really successful outcomes, and even some pretty good mission impact, to something that the commander said he couldn't live without and completely [00:11:30] changed his mind and opinion on government acquisitions. Now that's powerful. And so this pairing can be done not only at the lower level, the team level, which usually we get ... Initially, sometimes there'll be some skepticism, particularly to paired programming and doing it 40 hours a week, but people convert pretty quickly.

(11:52):

At that leadership level, again, where a lot of the transformation can stall out, there's a lot of initial reaction that's negative, especially because these people [00:12:00] just don't have the time to spend doing these new ways of working. And so figuring out ways that you can make small incremental changes to their ways of working, and produce outsized impact ,will give you the leverage you need to then change those smaller things that help you optimize and really get the teams working.

‍