I recently listened to Molly Graham on Lenny Rachitsky’s podcast: “The high-growth handbook: Molly Graham’s frameworks for leading through chaos, change and scale”. If you don’t know her, she helped leaders at Google, Facebook during the hyper-growth years.
As a Head of Engineering, I hear a lot of management advice that sounds nice in theory but falls apart when real deadlines hit. Molly’s advice is different. I found her insights tracked closely with my own experience, errors and learnings. It feels like she has made the mistakes already so we don’t have to.
Below is a list of core concepts I squeezed from the podcast and I tried to make it into a guide so you don’t have to and you can use it.
Table of contents
Open Table of contents
Give Away Your Legos
At its heart, this is just a metaphor for delegation. If you want to grow as fast as your company, you have to stop doing the job you just got good at. You need to hand it off to someone else and find a new, messy problem to solve.
Molly puts it this way: “You have to grow as fast as your company is growing if you really want to take advantage… learning to give away what you’ve gotten good at and move on to the next shiny pile of Legos.”
Think of it like owning a specific part of the codebase. You built the authentication system and you’re the only one who truly understands it. But now the company needs a billing system. You have to hand off the auth system to a new hire - even if they might make mistakes at first - so you can go build the billing system. If you refuse to let go of the code you wrote because you don’t trust others to touch it, you become the bottleneck. In engineering, this means constantly documenting and delegating your previous technical responsibilities so you can tackle higher-level architectural or organizational challenges.
Personally, I find delegating everything immediately is hard. You need to build trust and support with your reports first. You cannot simply assign someone you’ve never worked with to take over a core system you built - that’s mentally challenging. I would stress a lot and it could end with me helicoptering that person, causing us both to fail. I’ve found that delegating small tasks first helps me build an understanding of their capabilities and patterns of operation. Once that trust is established, it’s easier to hand over the big Legos.
I’ve also seen the flip side with “veteran” leaders. You might have someone who built a critical product and is the only one who knows the details. Verbally, they say they want others to work on the codebase, but in reality, it’s a struggle. Every pull request triggers back-and-forth, defensive comments and small conflicts. They say they want help, but they haven’t emotionally let go of the Legos. This halts progress because the team is fighting for permission to contribute rather than building the next thing. Not everyone can give their Legos away easily and it creates significant hurdles.
Bob the Monster
Change creates fear, ego and territorialism. Molly suggests externalizing these feelings into a monster named “Bob”. Bob’s job is to make you the worst version of yourself.
Her rule of thumb is practical: “Anything that lasts longer than two weeks is actually something you should pay attention to.” If the feeling fades in a few days, it was just Bob acting up.
Bob is that voice telling you to send a rage Slack message at 9:00 PM because you got layered or someone changed your roadmap. Don’t send the message. Let the wave roll through. If you are still angry about the same issue two weeks later, then it’s a real problem you need to address.
Personally, this is the area where I struggle the most. I’ve been told I have trouble with emotional intelligence because I react quickly - and not always positively - to change. My monster feels less like a ‘Bob’ and more like a Hydra-type creature. To cope, I try to count to 20 in my head before responding, or simply say “I will circle back” to buy time to cool down. I also rely heavily on taking notes (both in a digital app and a paper notebook) and doing retrospectives to see what triggered me. This analysis helps me decide if a specific issue is actually worth my energy or if it was just the monster reacting.
J-Curve vs Stairs
Most people view careers as stairs: predictable, linear promotions. That’s boring. The “J-Curve” is scarier. You jump off a cliff (take a new, hard role), you fall for a while (feel incompetent or have imposter syndrome) and then you climb out higher than the stairs would have ever taken you.
As Molly says: “The much more fun careers are like jumping off cliffs and you do fall, but then you climb out way beyond where the stairs could ever get you.”
This requires embracing the “Professional Idiot” phase. When you take a role you aren’t qualified for, you will feel stupid for about six months. Ask the dumb questions. That’s how you unlock the learning. If you feel comfortable, you probably aren’t on a J-Curve.
In my career, I approached this differently about three times. I didn’t just jump roles within the same company - I grew into higher roles across multiple companies and then deliberately jumped “down” to an IC role in a new company. I would relearn how to operate, grow and lead from that perspective. It felt extreme and I made plenty of failures, but it gave me multiple perspectives and expanded my network.
You can also apply this using the Toyota principle Genchi Genbutsu (“go and see”). Even if you aren’t changing jobs, go down and see how people under you operate day-to-day. Do the job with them. Delegate a day of your management work to others and focus on what your ICs do today. It might show you why you are “comfortable” in your current role and where the hidden problems lie.
The Waterline Model
”Snorkel before you Scuba”. When a team is struggling, our instinct is to blame the people - deep underwater. Usually, the problem is structural - right at the surface.
Molly says that “80% of problems on teams actually happen because of structural issues… Your only goal as a manager… is clear roles and clear expectations.”
Before you try to fix interpersonal relationships or individual psychology, check the basics:
- Structure: Do they know their goals? Do they know their roles?
- Dynamics: How are decisions made? How are meetings run?
- Interpersonal: Then look at relationships
- Intrapersonal: Finally look at individual psychology
When I joined Stay22 as Head of Engineering, I saw this play out exactly. My goal was to help the organization deliver impact faster while scaling. We didn’t have career ladders or clear role definitions, so I started right at the structural level. I built career ladders with clear progression, reevaluated all engineers to assign specific levels and introduced performance reviews and peer feedback. We doubled the team within a year and despite the new challenges, the clear structure kept us aligned. I also addressed the ‘Dynamics’ layer by transforming our documentation culture (e.g. introduced scalable ADRs), which grew to over 1k articles and refining our meeting structures. Now, I’m implementing weekly Engineering Reviews to keep alignment on goals and incidents.
As Molly suggests, defining the structure and decision processes solved the bulk of our communication and focus issues. The interpersonal and intrapersonal work is still there, but it’s a “TODO” on a stable foundation rather than a crisis.
6 Rules for Creating Goals
Goals are just communication tools. If they don’t create clarity, they are useless. Molly’s approach is strict to help preventing “goal creep”.
How to apply it:
- Max 3 Goals: Even Facebook only had three (Growth, Engagement, Revenue)
- One Goal Wins: If two goals conflict, you need to know which one matters more
- Explain to a 5-Year-Old: If an intern doesn’t understand it, rewrite it. No complex spreadsheets
- Strategy Should Hurt: If you aren’t making painful tradeoffs, you aren’t prioritizing
- One Goal, One Owner: Shared ownership means no ownership
- Goals ≠ Enough: You need a process to track them
When I listened to this part, I actually giggled out of confusion because at Stay22, we definitely have more than three goals. A year ago, it was even worse. Today, our CEO created an OKR framework with four sections, each with a theme and multiple goals inside. We struggle with the rules Molly mentions: some goals have multiple owners, others spread across multiple departments and the process always feels rushed. Tracking is its own comedy - we switched tools once and are planning to switch again. But there is a positive side: we are constantly trying to improve. We learn from every rushed iteration, tweak the process and add new challenges to see if they fail or bring improvement. We aren’t there yet, but the willingness to iterate is our current strength.
Leading Through Change: “I Don’t Know” rule
Leaders often fake confidence because they think they need to have all the answers. In high-growth environments, that’s dangerous.
Molly described it as: “Your job as manager and a leader is not to have all the answers. It is to get good at finding them.”
Stop faking it. When asked a tough question, say: “I don’t know. Who should we ask? How can we learn this? Let’s figure it out”. It builds more trust than a made-up answer.
From my experience, this is the topic leaders and ICs struggle with the most. You have to learn early in your career to say, “I don’t know, but I will find out”. I picked up this skill in school and university where they taught us how to search for answers, not just to memorize them.
Remember when you were a kid and thought you knew everything? Growing up opens your eyes to how much you actually don’t know. It’s similar in leadership. Even if you do know the answer, sometimes it’s worth taking a pause. Ask yourself: Is it helpful to the room to hear “Yes, I know”? Or is it better to say, “I don’t know right now, let me double-check the details and come back”? Often, the latter builds more confidence and accuracy.
Don’t Promise What You Can’t Control
Avoid making promises about titles, long-term roles, or stability just to close a candidate.
Molly calls these “letter bombs” that will explode later: “Do not promise things that you can’t control… There is literally no faster way to demoralize high performers than going back on a promise.”
Don’t tell a new hire they will be the “Manager of X” forever. In a scaling company, you can’t guarantee that. Be honest about the chaos.
In my experience, it is better to be tough and say “No” or “It will not happen” and deal with the upsetting response immediately. I had many such situations in my career and I learned from them. Don’t promise what is not in your control. Just say “No” and work on how to convert that “No” to a “Yes” down the road.
Hiring vs. Firing
We obsess over hiring, but firing is just as critical. Even great managers only get hiring right about 50% of the time. Keeping the wrong people creates “barnacles” that slow the ship down.
Molly talks though it as: “Firing people is as important as hiring people… Getting good at identifying when someone does not belong… is actually a skill.”
Recognize quickly when a hire isn’t working. Removing the drag on the team is better than letting it fester out of politeness.
I built a personal framework to recognize patterns during hiring and the first three months of probation. You need to define what matters for your organization and the culture you are building - this clarity helps you spot Great, Good and Bad fits quickly. I recommend reading Blink by Malcolm Gladwell - it can help you start building the skill to recognize patterns and identify good value in people.
Serve the Business, Not the People
This sounds harsh, but it provides clarity. When making hard decisions like layoffs or reorgs, ask what the business needs, not what will make people happy. If the business wins, everyone eventually wins.
”Serve the business, not the people… The answer is always what’s the right thing for the business.”
Use this mental tool: “If there were no emotions involved, what would I do?“. Make the decision based on that, then execute it with as much kindness as possible. It’s also crucial to remember that business is not a family - it is a sports team, which competes with other teams (businesses, rivals, competitors). Remember it.
Focus on High Performers
Managers default to spending 80% of their time on struggling employees. You get better returns by ignoring that instinct and investing in your best people.
Moly elaborated as: “High performers are actually the future of your company… I run experiments… I’m going to see if they can do this with less guidance.”
Give your best engineer a project that feels slightly too big for them. Remove oversight. Test their ceiling.
I’ve applied this approach many times and it is a key driver for building delegation habits and trust with reports. Your high performers help create a better culture - interestingly, low performers tend to either leave the company or try to catch up with the high performers. Make sure that each of your teams is balanced and has at least one high-performing IC. Identify them quickly and build a relationship with them - you support them and they support you.
Final Thought
Scaling a team and a product is never a straight line. It’s messy, it’s emotional and it often feels like you’re failing even when you’re succeeding. But if there’s one thing I’ve learned, it’s that we don’t have to have all the answers - we just need the right tools and the courage to find those answers together.
These frameworks aren’t silver bullet, but they work as a compass. When the chaos hits (100% it will), come back read this Handbook. Manage your monster, delegate and focus on what matters. We are building something incredible here at Stay22 and the growth - painful as it is - is the proof. Let’s get back to work 🫡