Director of Engineering

This guide is for Directors of Engineering who own multiple teams or domains and need a repeatable way to connect delivery, quality, and investment decisions. It uses Metrics → Delivery, Metrics → Quality, Teams → Iterations, and (where enabled) Resource Allocation and gitStream to create a portfolio view you can use with VPs, PMO, and DevEx.

TL;DR – Director of Engineering:

• Use Delivery and Quality metrics to compare teams and services, not individuals.
• Use Iterations to see how planning, unplanned work, and predictability vary across teams.
• Use Resource Allocation (if available) to connect headcount to initiatives and cost.
• Support gitStream and AI adoption where they remove friction or de-risk flow at scale.
• Run a light but consistent monthly portfolio review with your managers.

---

Start here in 15 minutes

1. Pick 3–5 teams you directly manage.
2. In Metrics → Delivery, set the window to the last quarter and note, per team:
   • Cycle Time and which sub-metric is slowest.
3. For each team, open Teams → Iterations → Completed and ask:
   • “Roughly how much work was unplanned?”
   • “Is carryover a pattern or an exception?”
4. Classify each team quickly:
   • Healthy, Stretched, or At risk.
5. Pick one team in each bucket and write one question you’ll ask that EM in your next 1:1 (e.g., “What’s the smallest change we can try to reduce unplanned work?”).
6. Use these notes as the first slide or section in your next monthly review.

---

Who this guide is for

• Directors owning multiple teams, domains, or services.
• Leaders responsible for both outcomes (delivery, quality) and investment choices (where teams spend time).
• Partners to VP Eng, PMO/Delivery Ops, and DevEx/Platform.

---

What you likely care about

• Which teams or domains are consistently healthy vs. struggling on delivery and quality.
• How unplanned work and incidents affect commitments across your portfolio.
• Whether your headcount and investments match strategic priorities.
• Where standards and automation (gitStream, AI) can scale improvement.

---

Before you begin

• Teams and services are mapped correctly in LinearB.
• Managers understand which metrics you care about and why (Cycle Time stages, key quality/reliability indicators).
• If available, Resource Allocation is configured with relevant custom fields (initiatives, project types, etc.).
• If your org uses gitStream or AI Insights, you know which teams are piloting them.

---

Step 1: Build a simple delivery & quality portfolio view

Goal: Compare teams on a few consistent, system-level signals.

Where: Metrics → Delivery and Metrics → Quality

1. Select a timeframe (e.g., last quarter) and a consistent aggregation mode (Average, Median, or percentile as agreed with your org).
2. Review per-team trends for:
   • Cycle Time and its slowest sub-metric.
   • High-level quality / reliability signals where available.
3. Classify teams into rough buckets:
   • Healthy: stable delivery, predictable quality.
   • Stretched: growing unplanned work, slowing stages.
   • At risk: repeated incidents or highly variable delivery.

Director lens: Your job is to ask “What’s making this system behave this way?” and equip EMs to own the improvement story.

---

Step 2: Use Iterations to understand predictability

Goal: See how planning and unplanned work differ across teams.

Where: Teams → Iterations (Completed)

• For each team, review a few recent iterations and note:
  • Planned vs. delivered work.
  • Level and sources of unplanned work.
  • Patterns of carryover across sprints.
• Use this to draw distinctions like:
  • “Team A has stable commitments but lots of incidents.”
  • “Team B over-commits and regularly pushes work out.”
• Turn those insights into coaching questions for EMs: “What’s the smallest process change we can try next sprint to improve predictability?”

---

Step 3: Connect investments using Resource Allocation (if enabled)

Goal: Show where your org actually spends time and cost.

Where: Resource Allocation dashboard and related Cost/Allocation reports

• Slice by initiatives, projects, or investment categories to see FTE spread.
• Compare:
  • Where teams spend time vs. your strategic priorities.
  • Work type mix (e.g., new features vs. maintenance) for key areas.
• Use this to support decisions like:
  • Rebalancing teams or roadmaps.
  • Justifying more capacity for foundational or reliability work.

---

Step 4: Support standards, gitStream, and AI at scale

Goal: Help managers and DevEx apply low-noise standards where they matter most.

• Identify 1–2 cross-team bottlenecks (e.g., slow reviews, large PRs, high rework).
• Partner with DevEx / Platform to:
  • Define simple standards (PR size, review SLAs).
  • Use gitStream and AI features (where enabled) to automate guardrails.
• Ask teams to measure:
  • Before/after trends on Cycle Time or quality.
  • Developer sentiment where Surveys or Coaching are available.

---

Recommended portfolio rhythm

Monthly (Director-led, with EMs)

• Review a portfolio snapshot:
  • Delivery and quality trends by team.
  • Key Iteration patterns (predictability, unplanned work).
  • Top 1–2 investment questions from Resource Allocation (if enabled).
• For each team, agree on:
  • One system improvement to test (process, standard, automation).
  • How you’ll measure success in the next cycle.

Quarterly (Director-led, with VP / PMO / DevEx)

• Use portfolio data to inform:
  • Roadmap and investment tradeoffs.
  • Org changes (team splits, new ownership).
  • Where to expand standards, gitStream, and AI.

---

Recommended next articles

• CTO Guide
• VP of Engineering Guide
• PMO / Delivery Ops Guide
• DevEx & Platform Engineering Guide
• Team Iterations & AI Iteration Summary
• gitStream Hub