Reframe Metrics to Leadership's Language
Pick two or three metrics that speak the exact word your leaders keep repeating
- Difficulty
- Moderate
- Time to result
- ~weeks to results
- Steps
- 5
- Confidence
- 90%
A decision framework for choosing which productivity metrics to focus on and how to present them: listen for the theme leadership keeps repeating (market share, profit margin, velocity, transformation), then select and label metrics that map directly onto that theme. It solves the 'what impact are our AI tools having' question by anchoring measurement to what decision-makers already care about. Use it when you must pick just two or three data points and get buy-in.
Origin
Forsgren's 'it depends' answer to measuring AI's productivity impact — she argues you should identify what the leadership chain cares about from the language they use and reframe your metric to resonate, even reusing their exact word ('if they're calling everything developer productivity, go ahead and call it productivity').
Core principles
- 01Start from what leadership repeatedly talks about, not from what's easy to measure
- 02Map the theme to a metric family: market share/competitiveness → speed; profit margin → money saved; velocity → cycle time; transformation → disruption impact
- 03Reuse leadership's own vocabulary so they don't have to work to understand your value
- 04For velocity, pick as broad a span as possible — idea to customer or idea to experiment
How to run it
- 1
Identify the recurring leadership theme
Listen to the messaging you've been hearing and name the overarching narrative: market share, profit margin, velocity, or transformation.
- 2
Map the theme to a metric family
If it's market share or competitiveness, focus on speed (feature to production/customer/experiment). If it's margin, focus on money saved. If it's velocity, measure cycle time. If it's transformation, frame around disruption.
Pro tip For margin, translate savings into recovered headcount cost or reduced vendor spend.
- 3
Pick a broad velocity span
When measuring velocity, choose the widest meaningful span — idea to customer or idea to experiment — and compare how long it takes now versus before AI and friction reduction.
- 4
Relabel to match their words
Reframe your metric using leadership's exact vocabulary so the value is immediately legible to them.
Pro tip If they call everything 'developer productivity,' call it productivity; if they say 'velocity,' frame it as velocity.
Watch out Don't make them work to understand what you're doing and the value you provide.
- 5
Disclose attribution honestly
When gains come from both AI tools and DevX work, say so — credit both rather than over-claiming one cause.
Pro tip Without the DevX improvements the AI tools would likely have produced some gains, but not nearly as much.
In the wild
If leadership keeps talking about losing market share, Forsgren says focus on speed and capture feature-to-production or feature-to-experiment cycle time. If they talk about profit margin, look for money saved and translate it into recouped headcount cost or reduced vendor spend.
→ The chosen metric lands because it answers the question leaders were already asking.
Common mistakes
Measuring what's easy instead of what resonates
Picking metrics detached from leadership's narrative forces them to work to see the value, weakening buy-in.
Over-claiming attribution
Crediting AI alone (or DevX alone) for a gain both are responsible for undermines credibility — disclose that they partnered.
Is it for you?
Best for
DevX and eng leaders who must select a few metrics and win executive buy-in for AI-productivity claims
Not ideal for
Pure internal engineering diagnostics where no leadership audience needs persuading
From the transcript
“it depends on what your leadership chain really cares about”
“Have they been talking about market share, right? Losing market share or or competitiveness in the marketplace. If that's it, focus on speed.”
“if they're calling everything developer productivity, go ahead and call it productivity”
“We don't want to make them work to understand what it is that we're doing and the value that we provide.”
From the episode
How to measure AI developer productivity in 2025
Nicole Forsgren