akshay deo

πŸ™ Principles

Engineering Management Blog

These are essential principles for being a better engineering manager and setting up better engineering management processes. Most of these principles are from the books that I have mentioned in the introductory post. I deserve very little credit for this.

πŸ–– Consistency is the key

  • Everything that we, as managers, do has to be consistent.
  • Any strategy/policy becomes effective when applied for sufficiently long periods.
  • So pick the tools you are comfortable with, and implement if you pledge to use them consistently.

♻️ Be iterative when it comes to processes

  • Change is the only constant 😌. And every process needs changes and iterations to make it perfect.
  • Keep these processes open for iterations. Do not lock on something unless you can prove it works. Also, do not engage the entire team from the inception of the processes.
  • Have strong opinions loosely held! You will find many counter-arguments for this philosophy, but this is highly important to keep your approach iterative.
  • Don’t add exceptions, count the exceptions, and update the process if an exception repeatedly occurs over a considerable amount of time.

πŸ₯Š Actionable feedbacks

  • The outcome of each engineering management process should be actionable with little or no manual efforts.
    • For the standup flow mentioned above, each standup message processing should give information about
      • πŸƒβ€β™‚οΈProgress of the project
      • πŸ›‘Blockers that an EM needs to look into on high priority.
  • Each actionable output should be converted into a metric and tracked throughout the lifetime of the action.
    • Based on the overall progress and blockers list, the system should give an overall health score based on some pre-defined weighting formula.
  • If it’s too low, the project needs some special attention. πŸ₯‹

⭕️ Always complete the loop

  • Most of the processes adopted by teams do not complete the loop. So no one can learn from these processes and keep on repeating almost the same mistakes in different forms.
  • Completing the loop is the job of an engineering manager. And one of the trickiest and tedious one.
  • For standups, making sure all the blockers are solved, digging deeper if the health score is low are the examples of completing the loop.

πŸ“£ Precisely over-communicate

  • Firmly establish the directions of the culture code/expectations from the team members. And keep on repeating it, even if you think you have done it well in the first go.
  • Sometimes these conversations become unpleasant, repetitive, and impertinent. But its all worth it.
  • Aligning everyone from the team towards the common goal with consistent practices and culture code is the recipe for successful execution.

πŸ€— Open up communication channels

  • Some members, specifically junior members of the team, may not feel comfortable voicing their opinions. Make sure senior members of your team are standing up for junior members.
  • Force team leads/project leads to create private channels, which are not accessible to the managers/leadership where every person in the group can voice all of their problems.
  • Never attribute issues to specific people. If there is a problem, there is a problem 😁. And its EMs job to look into it constructively.

πŸ˜‡ Trust in your teams capabilities

  • More you spend time with the team, you start understanding the strengths and weaknesses your team posses. But take a cons iderable amount of time to build your first capability matrix, ideally 2+ months.
  • Once you know these capabilities, start doubling down on strengths. Start removing yourself from the decisions which involve these strengths.
  • And then start focussing on weaknesses. I will speak about it in detail in another post, but the most effective way of fixing this is, bringing the complementing talent in the team. Either from another team or by hiring from outside.
...
A traditional Capability matrix for your team.

πŸ€– Automated flows

  • Engineers hate redundancy. And they should!
  • Ensure that the processes will leverage advanced APIs, ML, and NLP systems to remove the friction.
  • Long threads of daily standups notes. Everyone from a team pushes what they have worked on, what they will do today and blockers.
  • These are plain text, without forced formatting on the messages. So it depends on individuals to make it concisely insightful
  • Team members (not all team members but most of them) have very little insightful information while pushing these updates
  • Engineering managers keep on losing track as this information is not easily queryable.
  • Hence results in repeated information pulling, more status sync-up meetings, which wastes everyone’s time.
  • Robust and habit forming management flows are built with Trigger ➑️ Action ➑️ Variable Rewards ➑️ Investment.
    • πŸ“£ Timely triggers are crucial for an automated system.
      • Example Sending a Slack message with an action button to push the Standup update every morning at 10:30 am (local time of the team member).
      • Worst case, send an email πŸ˜ƒ(I ❀️Slack)
    • πŸ₯Š Action happens out of Intention. But this intention comes with timely triggers.
    • πŸ’° Rewards are in terms insightful information, higher visibility, automated sync-ups, less distraction for team members.
    • πŸ‘©β€πŸ’» Investment is pushing information into these systems to keep on getting rewards.
Get updates πŸ“₯

I won't spam, I promise ✌️


Comments