Andrea | Code to People

The IC to manager transition isn’t a promotion. It’s a career change.

@Stopworkplacebu Most gatekeepers weren’t born that way. They were never properly developed by their own managers, so knowledge felt like the only leverage they had. It’s a culture problem that gets passed down until someone breaks the cycle.

You’re not leadership material if you gatekeep wisdom like it’s job security. Mentorship isn’t about protecting your throne—it’s about building successors. Hoarding wisdom will get you nowhere. Great leaders don’t fear being replaced over time; They pave the way for it. 🖐

Hot take: being promoted into management should require as much vetting as hiring externally. We don't ask ICs 'do you actually want to manage people?' We just assume the best performers do and promote them. Then we're surprised when great engineers, analysts, and data scientists struggle in management roles they didn't choose. The skills that make you exceptional at the work don't automatically transfer. Agree or disagree?

@milkkarten When leadership outsources judgment to a tool, it doesn’t just slow things down - it tells your team that their expertise and experience don’t count for much. That’s a trust problem, not an AI problem.

it’s important to understand the impact workplace ai obsession has on team morale and boss perception

@randomrecruiter I don’t respond to “hey” until I know what it’s about. My time and your anxiety are both too valuable for that 😂

When your boss sends "hey, got a sec?" with no context

This is exactly why the analyst role isn’t going away - it’s evolving. The institutional context, the judgment about which metrics to trust, knowing what changed in the business last quarter - that lives with people, not models. The best analysts I’ve managed are the ones who can hold all of that and translate it into a decision. That’s still very human work.

Everyone is trying AI analysis with agents. No one has gotten it to be trustworthy enough to open widely to all business users. Dozens of companies I've spoken to in the past few weeks have told me this. Why is that? It turns out, the last 15-30% of quality and trust in analysis is really really hard to get right. This is what your best analysts know that AI doesn't yet: 1. Which metrics to trust, out of the mess of metrics and contradicting dashboards 2. What's changed in the business behind the numbers, including strategy, roadmaps, or just one-off edge cases. 3. How to actually think about the problem, including what its downstream implications and decisions are and how it should be scoped 4. What happened last time, and what feedback signals should alter our future judgements AI can sound smart while being wrong in exactly the ways that waste the most time. Wrote an article explaining this, and what we can do about it, in the latest in Opinionated Intelligence.

@Gavel_on_X Learning to communicate up. Most people optimize for doing great work. Far fewer learn how to make that work visible to the people who make decisions about their career. It’s not politics - it’s just how organizations actually function.

In your opinion, what’s one business skill everyone should learn first?

@1ssve Yes. Had an interview once where the hiring manager couldn’t clearly articulate what success looked like in the role or where the team was headed. That told me everything. If they can’t sell me on the vision, they probably can’t sell their team on it either.

Have you ever been to a job interview and while you are being interviewed you decide you no longer want the job?

There’s a CEO C: - Gets the data infrastructure right from the start - Builds simple, trustworthy reporting on top of it - Uses AI to automate the insight layer - Spends zero time manually entering numbers or arguing about which dashboard to trust The real unlock is building something that doesn’t require heroics to maintain.

I'm seeing this pattern with CEO dashboards now. CEO A: - Spends $50K on fancy BI tools - Has dashboards nobody trusts - Argues with teammates about feelings instead of facts - Systems break every few months - Still firefighting in Slack and email CEO B: - Uses a Google Sheet - Tracks metrics weekly - Makes their team enter metrics manually - Uses green/yellow/red stoplights - Reviews the whole business in 15 - 30 minutes In 2026, guess who scales to $20M?

Your team doesn’t push back on leadership. You do that for them. If you’re silently passing down unrealistic deadlines, you’re not managing. You’re just forwarding emails.

@randomrecruiter As a manager I have one rule: nothing after 3pm on a Friday. Some hills are worth dying on.

"I need a mental health day because my manager scheduled a 4pm Friday meeting" Meanwhile your ancestors on a random Tuesday:

@JoinBlind A lot of this is what happens when nobody prepares you for the actual job. Micromanaging and redoing people’s work aren’t signs you hate management - they’re signs you were handed the title and left to figure it out alone.

Promoted to manager, demoted in happiness 😅

The ones who got into leadership by default (promoted because they were the best IC, not because they wanted to lead) often never made the identity shift to being a manager in the first place. Going back to hands-on work isn’t regression. It’s just honesty about what they actually want.

Every day I’m talking to 1-2 L7s and L8s who are trying to find a truly hands on project to become a pure IC again. The dynamics of engineering has changed.

@1ssve Agreed. And as a manager, I’d add: if you find yourself wanting to ask why - the more important question is why you don’t already trust your team enough not to need the answer.

If your manager approves your leave without asking for a reason, it means you work in a good culture. Personal leaves should not be questioned. It's basic human rights.

Same is true for how managers develop their team. You can’t announce “I’m investing in your growth” and have that mean anything. Development isn’t a favor you grant - it’s a relationship that forms through the challenges you give, the feedback you deliver, and the trust you build over time. The managers who say it the most are often the ones doing it the least.

If you’re asking someone to be your mentor, you’re doing it wrong. Mentorship is not a favor you ask for. It’s a relationship that develops when you demonstrate that you are a cannon worth pointing at a big problem.

The common thread across all 7: managers who still think their job is to have the answers. Showing up unprepared, doing status updates, avoiding challenge - it all comes from not yet understanding that the 1:1 exists for them, not for you. The shift from “what do I need from this meeting” to “what does my team member need from me” fixes most of this list.

7 Reasons Your 1:1s Are a Waste of Time (And how to fix them before your next meeting)

This is the most underused rule in management. The feedback conversation you’ve been avoiding for weeks? It almost always goes better than you imagined. And the delay costs far more than the discomfort. Most managers wait until the situation is already broken. The ones who get better learn to have the conversation early - when it’s still just uncomfortable, not urgent.

A general rule of life: the more afraid you are to talk about something, the better you’ll probably feel after talking about it.

@THEROSSHARKNESS Agreed, and this matters even more when you’re managing people. AI can’t fix how your team thinks or operates either. And as a manager, that’s now your problem to solve - not just for yourself, but for everyone around you. Your leverage multiplies. So do your blind spots.

The one thing AI can’t fix is how you think, focus, operate and execute. Funnily, that’s where your biggest advantage lies.

The “reluctance to speak without something to say” is exactly what new managers struggle with - they feel pressure to fill every silence in 1:1s and meetings. And depth over breadth with your team is exactly how trust gets built - not by being liked by everyone, but by really knowing a the people well.

Things I wish someone had told me when I started in my career: The reluctance to speak without having something to say was actually quality control. The preference for depth over breadth in relationships was actually the foundation of a network built on trust rather than volume.

I spoke to someone yesterday who’s been a senior architect for years. Burned out. Exhausted. Wondering what comes next. She said she didn’t thrive in management because she hated giving critical feedback. Here’s what I told her: That’s not a personality trait. That’s a skill nobody taught you. The architects, engineers, and analysts who feel stuck often have exactly the instincts management requires - strategic thinking, systems awareness, deep domain knowledge. What they’re missing isn’t capability. It’s the framework for translating those instincts into leading people. That’s the gap nobody talks about. And it’s 100% learnable.

Worth adding: if you ever move into management, coding drops to near 0%. But clear communication, understanding trade-offs, ownership mindset - those stay. They just apply to people and decisions instead of code. The skills that make a great senior engineer and a great manager overlap more than most people expect.

I asked 10 senior engineers: "What skill makes a developer stand out?" Almost everyone said the same things: 1. Debugging skills 2. Reading other people's code 3. Clear communication 4. Understanding trade-offs 5. Writing simple code 6. Knowing system fundamentals 7. Curiosity 8. Consistency 9. Ownership mindset 10. Learning quick. Coding is only 20% of engineering