Notes on leading technical teams

Practical articles for engineering and technical managers, from your first one-on-one to multi-year strategy. Each one is marked with the stage of the certification path it's written for. New posts every Tuesday and Friday.

Earlier posts

  1. A weekly status update people actually read

    A five-part template for the weekly project update, with a before-and-after example, honest status colors and how to deliver bad news early.

  2. A senior specialist track that isn't a consolation prize

    How to define staff and principal levels for a specialist group so the track rewards real impact, keeps your experts and doesn't turn into a parking lot.

  3. Forecasting delivery dates as a range your stakeholders can plan around

    Replace the single estimated date with a forecast from your team's own throughput, a confidence level, and a plan for when scope grows.

  4. Drawing team boundaries before you announce a reorg

    A worksheet for testing proposed team boundaries against real work, cognitive load and dependencies, so your next reorg doesn't need redoing in a year.

  5. How to manage AI change in your organization

    A practical approach to AI change management: pick problems first, set guardrails, redesign workflows and roles, and measure outcomes, not usage.

  6. Retrospectives that change something by the next sprint

    How to run a retrospective that ends in one owned, finishable action, check it next time, and stop the same complaints coming back every sprint.

  7. A hiring loop that ends in a decision, not a debate

    Design an interview loop with one focus per interviewer, a shared scorecard and written feedback before the debrief, so hiring decisions rest on evidence.

  8. Exceptions to engineering standards that expire instead of piling up

    A lightweight exception process for security, platform and architecture standards: a one-page request, risk-based approval, expiry dates and a register.

  9. Managing the people who were your peers last week

    Your first month after promotion inside your own team: the conversations to have, what to change, what to leave alone and how to give feedback.

  10. Build-or-buy decisions you won't regret in three years

    A decision framework for build-or-buy calls: separate core from context, cost the full lifetime, plan the exit and write down what would change your mind.

  11. Using an AI assistant in your work as a technical manager

    Six manager tasks where an AI assistant saves real time, the prompts to start with, and the lines you shouldn't let it cross.

  12. Preparing for calibration so your ratings hold up

    How to build a one-page case for each person, spot your own rating biases before others do, and argue fairly in a calibration meeting without trading favors.

  13. Delegating work without dropping it or hovering over it

    Decide how much authority to hand over, brief the work in five lines, and set check-ins that match the person's experience, not your nerves.

  14. Research bets that end in a decision instead of fading out

    How to frame exploratory research as a bet with the hardest question first and written kill criteria, so every project ends in a clear go or stop.

  15. A one-page technical strategy your executives will read

    Most engineering strategies are lists of goals. How to write one that makes real choices, fits on a page and gets a decision from the people who fund it.

  16. One-on-ones that engineers don't dread

    A practical setup for one-on-ones: who owns the agenda, a running-doc template, a question bank for when things go quiet, and the habits that ruin them.

  17. Getting another team to deliver what your team depends on

    How to find cross-team dependencies early, ask for them in a way the other team can say yes to, track them weekly and escalate without burning bridges.

  18. Engineer to manager: what actually changes in your week

    The switch to management changes how your output is measured, how fast you get feedback and where your time goes. What to expect, and how to test it first.

Follow new posts with RSS