🔎 Summary
In this session we unpacked why performance measurement in engineering is so hard, and why many companies get it wrong by copying approaches from sales or other operational functions. Dave shared lessons from 20+ years as a tech leader, including what actually defines a high performing engineer, how to measure impact without fear, and why shared language between engineering, people teams and the business matters more than any tool. The conversation focused on context, clarity, continuous data, and human judgment.
🧠 Why Engineering Performance Is Different
Engineering is a creative, context driven function. Unlike sales, there is no single obvious benchmark.
Quote
“When you go to engineering, product design, these are creative functions. You go from zero to one. You are building something.”
What you could do next
💡 Stop using the same performance measures across all functions
💡 Acknowledge complexity and creativity when designing engineering frameworks
💡 Educate non technical leaders on why benchmarks are harder in engineering
🎯 What a High Performing Engineer Actually Looks Like
Dave challenged the idea that great engineers are defined by perfect code. Context matters more than elegance.
Quote
“The good engineer is the engineer that understands the context. What is the problem, what is the budget, and what happens if I do A or B.”
What you could do next
💡 Shift conversations from code quality to problem solving and impact
💡 Assess decisions made, not just outputs delivered
💡 Reward engineers who balance quality, cost and speed
🔁 From Fear to Fairness in Measurement
Measurement often triggers anxiety because people have been burned by bad systems in the past.
Quote
“When you say I want to start measuring, people start getting worried. Like, do you not trust me?”
What you could do next
💡 Be explicit about why measurement exists and how it will be used
💡 Co create goals and success criteria with engineers
💡 Replace surprise evaluations with shared expectations
🗣️ Bridging the Language Gap Between Engineering and the Business
A major challenge is translation. Engineers and business leaders often talk past each other.
Quote
“If you are not able to translate this thing to something people can understand, it’s a lost battle.”
What you could do next
💡 Help engineers frame technical work in business terms
💡 Encourage leaders to ask “what does this mean for customers or delivery?”
💡 Use shared vocabulary across teams
🧩 Competency Frameworks Only Work When Built Together
People teams or engineering teams cannot build effective frameworks alone.
Quote
“The time where it worked the best is when I worked myself with my chief people officer. It needs to be fifty fifty.”
What you could do next
💡 Build engineering frameworks jointly with people and engineering leaders
💡 Ensure frameworks reflect real work, not generic templates
💡 Make values and competencies reinforce each other
🧾 Continuous Performance Beats Annual Snapshots
Most engineering impact happens between review cycles.
Quote
“Things don’t happen in quarters. They happen in weeks.”
What you could do next
💡 Encourage engineers to log weekly contributions and reflections
💡 Use continuous data collection with periodic retrospectives
💡 Avoid relying only on recent wins or failures
📅 Why One to Ones Matter More Than Tools
No framework works without strong manager habits.
Quote
“If you don’t have a well defined process for one to ones, you become lazy.”
What you could do next
💡 Introduce consistent 1:1 agendas for engineering managers
💡 Cover blockers, delivery, career goals and hard conversations
💡 Train managers to spot issues between the lines
⚖️ Measuring Both Delivery and Enablement
Not all value shows up as features shipped.
Quote
“The most senior people are not doers. They are enablers.”
What you could do next
💡 Recognise coaching, reviewing and decision support as performance
💡 Balance delivery metrics with collaboration and impact
💡 Adjust weighting based on company stage and values
🧠 Data Over Gut Feel
Engineers trust evidence more than opinions.
Quote
“If you go with impressions, engineers get defensive. If you go with solid data, you can discuss facts.”
What you could do next
💡 Gather multiple data points before evaluating performance
💡 Use examples and evidence in feedback conversations
💡 Reduce subjectivity where possible
🔄 Calibration Is Non Negotiable
Consistency across managers is critical for fairness.
Quote
“You and me need to calibrate what good and bad is for us.”
What you could do next
💡 Run calibration sessions between managers
💡 Align on what meeting and exceeding expectations really means
💡 Reduce bias through shared discussion
🧭 Values Set the Boundaries for Performance
Performance only matters inside clear values.
Quote
“Values are the lines where I hire, fire, and promote people.”
What you could do next
💡 Make values non negotiable in performance decisions
💡 Do not promote people who break cultural norms
💡 Reinforce values through real consequences
🧱 A Practical Checklist for Building Performance Measurement in Engineering
Use this as a readiness check before you introduce metrics, frameworks, or tools.
✅ 1. Get Clear on Why You’re Measuring
If you cannot explain the purpose, don’t start.
Check you can answer:
- What problem are we trying to solve?
- Who is measurement for? Engineers, managers, the business, or all three?
- What decisions will this data actually inform?
If this isn’t clear:
Measurement will be perceived as surveillance, not support.
✅ 2. Define What “Good” Looks Like in Context
High performance looks different in different companies.
Check you have:
- A shared understanding of what good engineering looks like at your company, right now
- Agreement that trade-offs matter (speed vs quality vs cost)
- Recognition that senior engineers create value differently from juniors
If this isn’t clear:
People will optimise for the wrong things or defend themselves instead of improving.
✅ 3. Separate Creativity from Comparisons
Engineering is not sales. Benchmarks are limited.
Check you are not:
- Forcing uniform metrics across creative and non-creative roles
- Ranking engineers against each other by output alone
- Treating productivity as volume instead of impact
If this isn’t clear:
You risk punishing good decision-making and rewarding busy work.
✅ 4. Build Trust Before You Introduce Data
Measurement without trust creates fear.
Check that:
- Engineers understand how data will and won’t be used
- There are no surprise evaluations or hidden scoring systems
- Measurement is discussed openly, not introduced quietly
If this isn’t clear:
People will assume the worst, even if your intent is good.
✅ 5. Involve Engineers in Designing the System
Measurement done to people never works.
Check that:
- Engineers help define success criteria
- Feedback loops exist to adjust what is being measured
- The system reflects real work, not theoretical work
If this isn’t clear:
You’ll spend more time defending the framework than using it.
✅ 6. Use Evidence, Not Impressions
Engineers trust data when it is fair and transparent.
Check you are:
- Using multiple data points, not single snapshots
- Anchoring feedback in examples and facts
- Avoiding recency bias and gut feel assessments
If this isn’t clear:
Performance conversations will feel personal instead of constructive.
✅ 7. Make 1:1s the Backbone of Performance
No framework survives bad manager habits.
Check that:
- Managers have a clear structure for 1:1s
- Performance is discussed regularly, not just at review time
- Blockers, growth, delivery, and context are all covered
If this isn’t clear:
Measurement becomes paperwork instead of development.
✅ 8. Measure Enablement, Not Just Delivery
Senior impact often looks invisible.
Check you recognise:
- Coaching, reviewing, and unblocking others
- Decision-making and system-level thinking
- Long-term value, not just shipped features
If this isn’t clear:
You will undervalue your most experienced people.
✅ 9. Calibrate Before You Judge
Fairness comes from alignment, not formulas.
Check that:
- Managers regularly calibrate what “meets” and “exceeds” expectations mean
- Performance decisions are discussed across teams
- Bias is surfaced and challenged early
If this isn’t clear:
Performance outcomes will depend more on manager style than actual contribution.
✅ 10. Anchor Everything in Values
Performance has boundaries.
Check that:
- Values define what good performance cannot violate
- Behaviour matters as much as output
- Promotions and rewards reinforce cultural standards
If this isn’t clear:
You may reward results at the expense of trust and culture.
✅ 11. Treat Measurement as Ongoing, Not Periodic
Engineering impact happens weekly, not quarterly.
Check that:
- Data is collected continuously
- Reflection happens regularly
- Reviews summarise trends, not surprise people
If this isn’t clear:
Your process will always feel outdated by the time it lands.