As engineering teams scale, the quantity of people that genuinely care naturally decreases compared to folks who are just jobbing. That’s not cynicism, that’s just math. As engineering leaders, we can’t expect every team member to care as much as we do. That care is what separates leaders from jobbers. But the gap between those two groups doesn’t have to keep growing.

I wanna talk about how we can close that gap by propagating excellence.

In 1993, Hélio Gracie sent Royce Gracie to fight in the UFC as a practitioner of Brazilian Jiu Jitsu. At the time, BJJ was relatively known in the mainstream martial arts world. Royce’s success in the UFC disrupted the Mixed Martial Arts community so severely that today, almost every professional fighter is expected to have training in Brazilian Jiu Jitsu. The Gracies didn’t change the sport by convincing every fighter to train differently. They changed it by being so dominant that everyone else had to adapt in order to stay competitive.

As a technical lead of a team of about 25 engineers, part of a broader engineering organization of 300+, I found that it’s really difficult to drive cultural changes across the whole org. You can write docs, send Slack messages, give presentations – and most of it just bounces off. So instead, I decided to follow in Hélio’s footsteps. Be so good with a small group that the broader org can’t ignore what you’re doing.

Here’s how I’ve been approaching it.

Find Your Fighters

This starts with people. Not processes, not tooling, not a wiki page nobody reads. People.

You need to identify the engineers on your team who actually want to grow into leaders. Not everyone does, and that’s fine. But the ones who do – the ones who get frustrated when things are broken, who ask “why do we do it this way,” who voluntarily help junior engineers – those are your fighters.

Your job is to invest in them deliberately. Not just assign them harder tickets. Position them. Give them ownership of a domain so they become the authority on it. Put them in rooms they wouldn’t normally be in. Help them build a reputation beyond your immediate team. One of the engineers I mentor was solid technically, but invisible outside our group. We identified a specific area where they could become the go-to expert, got them presenting at internal sessions, and within a few months other teams were coming to them for guidance. That didn’t happen organically. It happened because someone pointed them in a direction and said “go”.

These are your lieutenants. As they grow, they start carrying the message for you. They set the standard in code reviews. They push back on shortcuts in design discussions. They raise the bar just by being present. And suddenly it’s not just you trying to change the culture – it’s five people. Then ten.

Assign the Training

Here’s where I’m gonna say something that might not be popular: mentorship should be assigned, not optional.

The organic mentorship model – where senior engineers just naturally gravitate toward helping junior engineers – sounds great in theory. In practice, it’s inconsistent. Some people get mentored. Some don’t. The quality varies wildly depending on who happens to sit near who or which Slack channel someone stumbles into. That’s not a system. That’s luck.

What I’ve found works better is deliberate pairing. Take your lieutenants – the folks you’ve already been investing in – and assign them a more junior engineer to develop. Not “hey, be available if they have questions.” Actual structured mentorship. Regular check-ins. Career conversations. Code review with teaching intent, not just approval stamps.

This does two things. First, it makes sure every engineer on your team has someone actively invested in their growth, not just the ones who are naturally assertive enough to seek it out. Second – and this is the part people don’t think ab out – it develops your lieutenants into leaders. Mentoring someone is the fastest way to learn how to lead. It forces you to articulate why you make the decisions you make, to think ab out someone else’s career trajectory, to practice the soft skills that most engineers avoid until they’re suddenly managing a team and have no idea what they’re doing.

You’re not just growing junior engineers. You’re training senior engineers to lead.

Be So Good They Can’t Ignore You

Here’s where the Gracie analogy really kicks in.

All of that internal work – developing leaders, assigning mentorship, raising the bar on your immediate team – that’s your training camp. But the point isn’t just to have a great team in isolation. The point is to be so visibly excellent that the broader organization has to pay attention.

This means making your improvements public. Host lunch-and-learns. Do team presentations on what you changed and what happened as a result. Write it up. When your team ships faster, breaks less, or onboards new engineers in half the time – make sure people know about it. Not to brag. To demonstrate.

When other teams see your engineers confidently owning their domains, when they see your junior engineers ramping up faster, when they see the quality of your code reviews and documentation – they start asking questions. “What are you doing differently?” That’s the moment. That’s Royce Gracie submitting a guy twice his size on live TV. The results do the convincing for you.

You don’t change a 300 person engineering org by sending an email about best practices. You change it by building a team that’s so clearly operating at a higher level that adoption becomes the only competitive option.

The Compound Effect

The thing about this approach is that it compounds. You invest in five people. Those five people each invest in someone else. The standards they set become the expectations for everyone around them. The knowledge they share becomes the baseline. And the culture you wanted to build in a 300 person org starts to take root – not because you mandated it from the top, but because it grew from a seed that was too strong to ignore.

Hélio Gracie didn’t change MMA by writing a memo. He trained his family, sent them to fight, and let the results speak. If you want to change your engineering culture, do the same thing. Train your people, let them compete, and make the results impossible to dismiss.

Excellence propagates. But somebody has to plant it first.