Managing Managers: The M2 Transition (Senior-Manager / Manager-of-Managers Craft, 2026)

The M2 mental model: your product is your line-managers

The transition from line-manager (M1) to senior-manager (M2) is the second-largest discontinuity in the engineering management ladder per the canonical references. Fournier devotes chapter 7 of The Manager's Path entirely to it. Larson's An Elegant Puzzle 'managing managers' chapter and the related lethain.com posts cover the same material. The structural changes:

• Direct line-of-sight to engineering work disappears. At M1 you knew every PR, every design doc, every retrospective on your team. At M2 you have 2–4 line-managers reporting to you, each with their own 5–10 reports. You cannot read every PR. You cannot attend every design review. The temptation is to compensate by going deep into one of the line-managers' areas — that is the M2-as-line-manager-with-extra-headcount failure mode. Larson is explicit: 'managing managers means letting go of being the manager.'
• Your product is your line-managers. The mental shift Fournier describes: at M1 your output is your team's output; at M2 your output is your line-managers' development plus the output of teams you influence (Grove's manager output equation generalized one level up). A senior-manager who promotes one of their line-managers to M2 over 18 months has done their job; a senior-manager who 'ships projects' but whose line-managers are stagnant has failed.
• Andy Grove's task-relevant maturity becomes load-bearing. Grove's framework from High Output Management chapter 3: management style should be calibrated to the task-relevant maturity of the report, not the seniority. A high-TRM line-manager (one who has done this kind of work before, knows the patterns, can make decisions) needs delegation and air-cover. A low-TRM line-manager (newly promoted, in a new domain, navigating an unfamiliar pattern) needs structure and explicit direction. The M2 who treats all line-managers identically is failing.
• The 1:1 cadence shifts. At M1: weekly 1:1 with each direct report. At M2: weekly 1:1 with each line-manager (4 hours/week). Skip-level 1:1s with every IC under your line-managers on a 6–12 month rotation (variable hours). The line-manager 1:1 is the load-bearing artifact at M2.

The empirical sign you have made the M1-to-M2 transition: when your line-managers run their teams without daily input from you, when they bring you problems already half-solved, and when your skip-level 1:1s confirm the line-managers' read of their teams.

Skip-level 1:1s: the M2 calibration craft

Skip-level 1:1s are the senior-manager's primary calibration and signal-gathering channel. Larson's 'The perfect skip-level meeting' on lethain.com and Fournier's chapter 7 both treat skip-levels as the load-bearing artifact of the M2 job. The canonical mechanics:

1. Cadence. Every IC under your line-managers gets a skip-level 1:1 every 6–12 months. Senior ICs (E5+) trend longer; new hires and early-career ICs trend shorter. New line-managers — set up the first round of skip-levels with their team in their first 60 days; do not wait. Larson recommends 30 min in-person where possible.
2. Agenda. The skip-level is the IC's, not yours. Open with: 'this is your time. I have three questions if we run out, but you go first.' The three fallback questions: (1) what is your manager doing well that I should not change? (2) what is one thing you wish your manager would do differently? (3) what is your career goal and is your work today moving you toward it? Hogan's BICEPS frame helps you read the answers.
3. Confidentiality. The skip-level is not gossip. The IC is talking to their manager's boss. You do not relay specific complaints back to the line-manager. You do, however, look for patterns across multiple skip-levels: if 3 of 6 ICs on a sub-team mention the same line-manager weakness independently, that is calibration data. You take that to the line-manager in a 1:1 as your observation, not as 'I heard.'
4. Action. Don't promise to act on individual feedback. Do log patterns and bring them to your line-manager 1:1s. Do escalate if a skip-level surfaces a serious issue (harassment, bias, severe performance problem) — those go through HR immediately, not through the skip-level mechanic.
5. Failure modes. The senior-manager who uses skip-levels to micromanage their line-managers' teams. The senior-manager who treats skip-levels as morale-building chats with no calibration purpose. The senior-manager who promises action on a complaint and either follows through (undermining the line-manager) or doesn't (losing the IC's trust). Charity Majors's charity.wtf has multiple posts on this; the through-line is 'the skip-level is a privileged signal channel, not a problem-solving meeting.'

Hiring and leveling line-managers

Hiring and leveling line-managers requires a different rubric than hiring engineers. Larson's 'Hiring funnel' essay (lethain.com) and Fournier's chapter 7 both flag the line-manager loop as needing a structured behavioral interview rubric distinct from the IC interview. The mechanics:

1. Different signal sources. Engineering interviews look for technical depth, design judgment, coding fluency, and cross-functional partnership. Line-manager interviews look for people-management depth, performance-conversation craft, hiring track record, and cross-functional partnership at the management tier. The technical-coding round is reduced or eliminated; the people-management depth round expands.
2. The 'manage a difficult performance situation' real-time scenario. The canonical line-manager interview round Larson and Fournier both describe. The interviewer presents a structured scenario ('one of your reports has been missing deadlines for three months and the team is grumbling') and asks the candidate to walk through their approach in real time. Strong candidates name Hogan's feedback equation, ask about the underlying cause, propose a structured improvement period with named exit criteria, and engage with the HR-and-process dimension.
3. Behavioral depth with specifics. 'Walk me through a specific 1:1 you ran in the last 90 days.' 'Tell me about a time you put someone on a PIP — what was the outcome, what did you learn, what would you change?' 'Tell me about a hire you made who didn't work out — what were the early signals you missed?' The interviewer is grading on the specificity and the self-awareness of the answer, not on the framework name-drop.
4. Pattern-matching from IC interviews fails. A strong senior IC who has not done the management work struggles at the line-manager interview loop. The 'I haven't done this exactly but here's how I'd think about it' answer is rarely sufficient at the senior-manager loop. Internal-promotion candidates have the implicit credibility advantage; external candidates need to demonstrate explicit prior management depth.
5. Calibration across line-manager hires. The M2 hiring multiple line-managers needs a calibration rubric to compare them. Larson's 'Hiring funnel' essay is the structural reference; the operational mechanic is a structured behavioral packet that lets the hiring committee compare candidates on the same dimensions.

Running staff-meetings as decision-forcing-functions

The staff-meeting (a weekly 60–90 minute meeting with the M2's line-managers) is the team's decision-forcing function and the second-largest M2 calendar item after individual 1:1s. Larson's 'Running staff meetings' on lethain.com and Hogan's Resilient Management chapter on meeting design are the canonical references. The mechanics:

1. The default failure mode. The round-robin status report. Each line-manager spends 8 minutes describing their team's status. The M2 nods. The meeting ends. No decisions are made; no disagreements are surfaced; no cross-team coordination happens. This is a waste of the line-managers' calendar.
2. The high-leverage pattern. Each meeting has 1–2 named decision items, sent in pre-read 24 hours ahead. The meeting time is for discussion of the pre-read, not for reading aloud. The M2's job is to surface conflict, force decisions, and give air-cover to line-managers raising contentious points.
3. Status updates go in writing. A written status update (Slack channel, shared doc, weekly email) replaces the round-robin. Line-managers read the updates async. Meeting time is reserved for what cannot be done async.
4. Disagreement is welcomed. The M2 explicitly asks 'where do we disagree on this?' and gives air-time to the disagreement. Hogan's Resilient Management chapter on meeting design is direct: a staff-meeting where the M2's preferred answer is the only answer voiced is not generating useful information.
5. The decision is recorded. Larson's 'no surprises' principle generalizes: every decision made in staff-meeting is recorded in writing, with the rationale, within 24 hours. Line-managers who were not in the meeting (illness, conflict) can read the decision and the reasoning. The team-of-line-managers' shared written record is the M2's institutional memory.
6. Cadence variance. Some M2s run weekly; some bi-weekly with a longer monthly strategic block. Larson's framing: the cadence should match the rate of cross-team coordination needs, not a calendar default.