Line Manager / M1 Engineering Manager Guide for Tech Companies (2026)

What line managers actually do, by week

The line-manager calendar at a typical FAANG-tier or AI-lab tech company, 5–10 reports, drawn from Fournier (The Manager's Path, chapter 5), Larson (An Elegant Puzzle 'Effective managers'), and the public schedules Charity Majors and Will Larson have shared on their blogs:

• 40–50% 1:1s and people work. 30-min weekly 1:1 with each direct report (5–10 hours/week). 1:1 with manager (1 hour). Skip-level 1:1s with reports of any tech-leads on your team (1–2 hours). Calibration / performance prep work (variable, heavier in cycle quarters). The 1:1 is the load-bearing artifact — Fournier's chapter on 1:1s and Hogan's BICEPS framework (Belonging / Improvement / Choice / Equality / Predictability / Significance) are the canonical mechanics. The agenda is the report's, not yours.
• 20–25% project oversight. Standups, design reviews, technical decision-making conversations where you are now the tiebreaker not the contributor. Reading PRs (less than you used to, more selectively). Writing the team's quarterly plan with input from senior ICs and PM partner.
• 15–20% cross-functional and stakeholder management. PM partnership weekly. Design partnership weekly. Engineering-peer-manager 1:1s (your line-manager peers, the people you depend on for scope handoffs). Skip-level upward (your VP / director, depending on org).
• 10–15% hiring. Hiring loops, debrief meetings, sourcing conversations with recruiters, written hire/no-hire packets. Hiring is the single most leveraged manager activity per Larson and Fournier both.
• 5–10% IC work. If you are still in tech-lead-manager mode you might keep 1–2 days of code per week. Once you are clean line-manager that drops to under 1 day. Charity Majors's 'manager-engineer pendulum' framing: don't pretend you're still an IC; either you are or you aren't, and the team can tell.

The 90-day calendar shape that Lopp (Rands in Repose) and Larson both recommend: weeks 1–2: every-other-day 30-min 1:1 with each report. Weeks 3–8: weekly 30-min. Week 9+: settle into the stable cadence with a 60-min monthly career conversation overlay.

The first 90 days: the canonical playbook

The first 90 days is the most-written-about period in engineering management. The canonical advice across Fournier, Larson, Hogan, Lopp, and Pragmatic Engineer's coverage of new-EM transitions:

1. Days 1–14: listen, do not act. Run 1:1s with every direct report. Ask the same three questions to each: (1) what is going well that I should not break? (2) what is broken that you would fix if you were me? (3) what do you want me to know about you that I would not learn from your code? Hogan's BICEPS frame helps you read the answers — most people answer the second question with a Predictability or Choice complaint, not a technical complaint.
2. Days 15–30: read the artifacts. Last quarter's PRDs, design docs, retros, postmortems. Last 6 months of perf-review writing if you have access. The team's hiring loop debrief packets if any. Form opinions silently. Note discrepancies between what people said in 1:1s and what the docs say — that gap is where the most useful management work hides.
3. Days 31–60: small bets. Pick 2–3 small things to change. Examples: kill a recurring meeting that nobody defends, restructure the standup so it is 12 minutes not 30, write the team's first published 'how we work' doc. Do not change the tech stack. Do not re-org. Do not change the team's mission. The signal is competence at the small things first.
4. Days 61–90: write your team's plan. One page. What is the team's mission. What are the 3–5 things you will ship this quarter. What is the team's biggest risk. Send to your manager for review. Send to the team for input. This is the artifact that establishes you as the team's manager-in-the-eyes-of-leadership; without it, you are the team's most visible IC.

The dominant first-90-day failure mode per Charity Majors and Larson both: a new EM tries to prove they are still the team's most technically credible person by jumping into code review battles. You will lose authority faster by winning a tech argument with your senior IC than you will by deferring to them. Leverage compounds; show-off does not.

Worked scenario: the first PIP, in 9 months

A 9-month worked scenario — line-manager inherits a team of 7 with one tenured underperformer (E4, 5 years at the company, last review 'meeting expectations' but with consistent 'partial' on cross-functional dimension). The team grumbles about it openly in 1:1s within month 1. Drawn from Larson's 'Useful PiP' essay (lethain.com), Fournier's chapter 6 ('Managing a Team') and chapter 7 ('Managing Multiple Teams') performance-conversation guidance, and Hogan's feedback-equation craft.

• Months 1–2. Listen. In 1:1s with the underperformer (call them Casey), ask the BICEPS questions. Casey is unhappy: feels their work is not visible to the org, feels stuck at E4 for two cycles, suspects (correctly) that the team thinks they are slow. Avoid promising anything. Read Casey's last 6 months of work directly. Form an honest opinion: the work is technically OK, but the throughput is half the team's median and the cross-functional reputation is poor. Document your observations weekly in your private 1:1 notes.
• Month 3. First direct feedback conversation, using Hogan's equation. Observation: 'In the last quarter you shipped one project; the team's median was three.' Impact: 'It means the team has been picking up the on-call rotations you used to share, and our cross-functional partners have started routing requests around you.' Request: 'I want to work with you to define what 'meeting expectations' looks like by behavior, in writing, by next week.' Casey is shaken but engaged. Document the conversation in writing within 24 hours and send a follow-up email recapping.
• Months 4–6. Structured improvement period. Weekly 30-min 1:1 dedicated solely to the improvement plan. Specific behavioral targets: ship two projects this quarter at expected scope, attend the cross-functional standup weekly, get one positive 360-review note from a non-eng partner. Casey hits two of three. The team notices the change and morale improves. You are still uncertain whether this is sustainable.
• Month 7. Casey misses the second project's deadline by three weeks; the cross-functional standup attendance lapses. You consult HR on a formal PIP. The HR partner walks you through your company's process; you write the document jointly. The PIP is 60 days, structured, with named exit criteria and a named coach (a senior IC on a peer team). You and Casey have the conversation in person. Casey understands.
• Months 8–9. Casey clears the PIP — barely. Their work meets the bar at month 9; the team's morale has recovered; you have learned the canonical lesson Larson names in 'Useful PiP': a properly-scoped PIP either restores someone or lets them exit with dignity. The failure mode is the manager who delays — the team notices and quits around you. Six months later Casey transfers to a different team with your blessing and a clean reference. Two of your strong ICs tell you in 1:1s they almost left during months 1–6 because the underperformer issue had been ignored for a year before you arrived.

The lesson for new line-managers: delaying a performance conversation by one quarter costs your team one month of their best ICs' patience. The math is brutal. Larson's 'Useful PiP' is the 30-minute read; Fournier's chapter 6 is the 90-minute read; both are mandatory before your first PIP.

Compensation: the real bands at line-manager

Total comp at line-manager (M1) FAANG-tier and AI-lab in 2026 (US, per levels.fyi self-reports — caveat that levels.fyi is self-reported and noisy at the manager level because manager comp varies more by team and tenure than IC comp does):

Three structural facts at line-manager comp:

1. The management premium over senior IC at FAANG is small or zero. Meta E5-IC and E5-manager track the same band; the management title does not bump cash or stock at most FAANG-tier. The reason engineers move into management is rarely comp at line-manager — it is leverage and energy fit.
2. Netflix's single-band comp model pushes nominal numbers higher. Netflix gives you the choice of cash vs. stock annually; the 'all cash' default produces higher reported comp than peer FAANG, but the underlying value is comparable.
3. AI-labs sit materially above FAANG. Anthropic's engineering-manager comp is the top of the line-manager market in 2026 per levels.fyi self-reports, with reported peak total comp on equity vesting cycles exceeding $1M. The risk is equity concentration in private-company stock; the upside is durable.

For negotiation: the best public reference is levels.fyi's compare page (levels.fyi/?compare=Meta,Google,Stripe,Anthropic&track=Engineering%20Manager). The Pragmatic Engineer's 'Trimodal nature of software engineer compensation' framework (Tier 1: traditional companies, Tier 2: public tech, Tier 3: AI labs and frontier) extends cleanly to engineering management.

Failure modes at line-manager: what gets you replaced

1. Hiding from people work. The new EM who keeps coding 60% of the time, treats 1:1s as overhead, and reports up the ladder mostly with technical detail. The team senses the avoidance within a quarter. Fournier's chapter 5 is the explicit warning.
2. Delaying performance conversations. The single most common new-EM failure mode. The team notices the underperformer is being protected; the team's strong ICs interview elsewhere. Larson's 'Useful PiP' is the prevention reading.
3. Being the team's smartest engineer in the room and showing it. Charity Majors's 'manager-engineer pendulum' frames this directly: the moment you start one-upping your senior IC in design review, you have lost the leverage that the manager job is built on. The senior IC's confidence is the team's most valuable commodity; do not corrode it.
4. Re-organizing in the first 90 days. Larson's chapter 'Organizational Design' in An Elegant Puzzle is explicit: re-orgs are expensive. New EMs who re-org early to 'put their stamp on the team' typically un-do their own re-org by month 6.
5. Hiring desperately. The team has an open req. The new EM hires the first acceptable candidate. Six months later the candidate is the team's biggest performance management problem. Larson and Fournier both: 'no hire is better than the wrong hire.' This is the hardest discipline at line-manager because the org is pushing for headcount fill.
6. Cargo-culting frameworks. The new EM reads three management books and tries to roll out OKRs, RACI, BICEPS, and the feedback equation in the first month. The team experiences this as buzzword theater. Pick one tool, use it consistently for a quarter, then add another. Frameworks are tools, not identity.
7. Refusing to do the unglamorous coordination. The new EM thinks their job is 'vision and people.' But 30% of the line-manager job is calendar Tetris, escalation routing, and unblock-the-team coordination. Hogan's Resilient Management chapter 4 names this explicitly: the unglamorous work is the leverage.