← Essays

Measure your efficacy with AI

Want to Prove You're 10x-ing with AI? Measure What Matters.

· 3 min read

TL;DR

If you want to prove you’re 10x with AI:

  • Stop counting inputs.

  • Start tracking decisions made, velocity gained, risk reduced.

  • Document moments of leverage, not moments of usage.

AI can make you faster. Smarter. More scalable.

But most people using AI right now are chasing the wrong scoreboard.

Lines of code. Number of pull requests. Time saved.

All easy to game. All easy to fake.

If you want to actually prove you’re operating at 10x—especially in a leadership or senior IC role—you need to measure efficacy, not activity.

Here’s how to do it.

1. Forget Volume. Track Velocity-to-Outcome.

Real productivity isn’t about doing more. It’s about doing what matters sooner and better.

Ask:

  • Are you delivering the right thing faster?

  • Are your decisions better informed and more defensible?

  • Are teams moving quicker because of you?

Example:

  • Bad metric: “I wrote 20 pages of documentation.”

  • Good metric: “My doc reduced back-and-forth by 70% and unblocked a launch in 2 days.”

2. Define Efficacy for Your Role

Here’s what efficacy looks like depending on your scope:

If you’re an engineer:

  • Time to clarity: How quickly do you move from ambiguity to working code?

  • Time to impact: How fast do your changes make it to production and create value?

  • Reduction in defect rate: Are bugs decreasing because of earlier insights?

If you’re a manager:

  • Speed of decision-making: Are you turning around decisions in hours instead of days?

  • Quality of communication: Are your strategy docs tighter, clearer, and more aligned?

  • Unblocked bandwidth: Are your team members needing less from you because your answers and inputs are sharper?

3. Track Leverage Moments, Not Usage

Anyone can say “I used ChatGPT this week.” That tells me nothing.

Instead, write this down weekly:

  • “What outcome was significantly better, faster, or easier because of AI?”

Examples:

  • “Used AI to explore tradeoffs in a complex system design before committing.”

  • “Summarized 40 pages of meeting notes into a decision doc that aligned 3 teams.”

These are force multipliers. That’s what 10x looks like.

4. Record your Efficacy

Forget dashboards. You don’t need fancy tools.

Once a week, write:

  • What critical problem did I solve or accelerate using AI?

  • What would I have done without AI—and how long would it have taken?

  • Did the result drive clarity, alignment, or momentum?