The Engineering Manager's First 90 Days: A Playbook

Published on June 6, 2026

Most new engineering managers fail silently. Not in a dramatic way. Their team doesn't mutiny. Their code doesn't break. They just drift—running status-update meetings, chasing Jira tickets, and reacting to fires—until six months in, their skip-level asks why the team keeps missing estimates and why two of their best engineers are interviewing elsewhere.

The first 90 days as an engineering manager are the highest-leverage window of your entire tenure. What you establish in this period—how you run 1-on-1s, how you surface blockers, how you give feedback—becomes the culture of your team. Get it wrong, and you spend the next year undoing it.

This is the playbook.

Days 1–30: Shut Up and Listen

Your only job in the first 30 days is to understand. Not to fix things. Not to implement the process you used at your last company. Not to prove that you have opinions.

You almost certainly don't understand the real problems yet. The things that look broken from the outside are usually symptoms. The actual root causes are buried in context you don't have yet—context that lives in your team's heads and surfaces only if they trust you enough to share it.

Run a discovery 1-on-1 with every direct report in week one.

Don't come with an agenda about what you want to change. Come with four questions:

1. What is the one thing slowing you down most right now?
2. What is the one decision that keeps getting delayed?
3. What did the previous manager do that you want me to keep doing?
4. What did the previous manager do that you should stop immediately?

Write down every answer verbatim. Don't respond. Don't defend. Don't promise. Just listen and take notes.

"The fastest way to destroy trust as a new manager is to spend your first month talking about what you're going to change."

You will hear conflicting things. Two engineers will describe the same situation completely differently. That is valuable information. It tells you where there is misalignment, where there are unspoken team dynamics, and where the real leverage points are.

Do not make any structural changes in month one. No new ceremonies. No new tools. No new processes. The only exception is a change that would cause active harm if you delayed it.

Days 31–60: Find Your Two Leverage Points

By day 30, you will have run 1-on-1s with your whole team and you will be sitting on a pile of raw information. Some of it will be noise. Some of it will be the most important data you collect all year.

Your job in month two is to identify the two highest-leverage problems on your team and build a concrete plan to address each one.

Why two? Because if you pick one, you will solve a symptom. If you pick ten, you will solve nothing. Two forces you to make a real prioritisation decision.

A high-leverage problem has three properties:

1. Fixing it unblocks multiple people, not just one person
2. It is within your control to fix (not dependent on a VP reorg or a roadmap shift)
3. Your team already knows it's a problem and will notice when it improves

Slow code review turnaround is a high-leverage problem. One engineer who struggles with written communication is not.

By the end of month two, you should have:

• A clear written diagnosis of each problem
• A proposed solution you have shared with your team for feedback
• A timeline for the first measurable change

The act of writing the diagnosis publicly—sharing it with your team in a 1-on-1 or team meeting—builds enormous trust. It signals that you were actually listening in month one, that you understood what you heard, and that you are willing to be held accountable to fixing it.

Your 1-on-1s in this phase should shift. Month one was exploratory. Month two is diagnostic. Start bringing a working document to every 1-on-1 that tracks the specific blockers you are working to remove for that person. Ask: "Last week I said I was going to resolve the code review SLA issue. Has that changed anything for you?"

This is the loop that separates managers who build trust from managers who just occupy the role.

Days 61–90: Make Your First Real Changes

By day 60, you have eight weeks of structured 1-on-1 data behind you. You understand your team's real frustrations, you have identified your two leverage points, and your team has seen you follow through on small commitments. You have earned the credibility to make changes.

This is the moment to move. Not tentatively. Not with endless caveats. With clear intent and a feedback loop already in place.

Whatever your two leverage points were, implement your solution now. Change the code review process. Fix the on-call rotation. Restructure the sprint planning ceremony. Do the thing.

Then—and this is the part most managers skip—announce the change explicitly in your next team 1-on-1 cycle, and tell each person: "I'm changing X because of what you and the team told me in month one. In four weeks I want to know if it's actually working."

This closes the loop. It shows your team that the conversations they had with you in month one were not performative. It creates a culture where people bring you real information because they have seen that real information leads to real action.

"Most teams have never had a manager who explicitly traced a change back to what the team said. Do this once and watch how much more honest your next 1-on-1s become."

The 90-day marker is also the right time to have a direct conversation with each report about their growth trajectory. Not a formal performance review. Just a clear, honest 20-minute conversation: where they are, where they want to go, and what the next meaningful milestone looks like.

If you run good 1-on-1s, this conversation should not be a surprise. It should be the natural conclusion of two months of incremental context-building.

The One Tool That Runs Through All Three Phases

If you read this playbook carefully, you will notice that the 1-on-1 meeting appears in every single phase.

That is not an accident. The 1-on-1 is the primary management instrument. Everything else—team meetings, code reviews, performance frameworks—is secondary. If your 1-on-1s are chaotic, your first 90 days will be chaotic regardless of what else you do.

The most common mistake new engineering managers make is treating the 1-on-1 as a catch-up conversation. It is not. It is a structured data-collection and trust-building tool. It should have a persistent agenda, tracked action items, and a follow-up mechanism that holds both parties accountable.

Running them in a blank Google Doc or a Slack DM thread means the institutional memory disappears. The action item you agreed on last Tuesday is buried under three weeks of messages. The thing your engineer mentioned about a team conflict six months ago is gone.

This is why we built Accordia. A quiet, focused operating system for engineering managers who want to run 1-on-1s the right way—with structured agendas, persistent action item tracking, and a clear record of what was said and what was agreed.

Your first 90 days only happen once. Run them with the right infrastructure.