---
title: "AI-Assisted Engineering: Measure Outcomes, Not Activity"
url: https://daily.dev/posts/ai-assisted-engineering-measure-outcomes-not-activity-kgafc31vm
source_url: https://thomasthornton.cloud/ai-assisted-engineering-measure-outcomes-not-activity
type: article
source: "Thomas Thornton"
published: 2026-08-04T14:47:20.225Z
updated: 2026-08-04T14:48:02.364Z
tags: ["devops", "github", "productivity", "observability"]
reading_time: 12
upvotes: 32
comments: 5
language: en
---

> ## Documentation Index
> Fetch the complete documentation index at: https://daily.dev/llms.txt
> Use this file to discover all available pages before exploring further.

# AI-Assisted Engineering: Measure Outcomes, Not Activity

**[Thomas Thornton](https://daily.dev/sources/thomasthornton)** · 12 min read · 32 upvotes · 5 comments

## Summary

Activity metrics like token usage, accepted completions, and generated lines of code prove AI tools are being used but not that they're delivering value. Platform teams should instead measure engineering outcomes: workflow success rates, pull request review burden, guardrail failures, cost per accepted outcome, and developer confidence. The post proposes a structured measurement model broken down by team, repository, workflow, and model — with task-specific success definitions (e.g., a Terraform provider upgrade has different success criteria than an incident summary). Key warnings include avoiding vanity dashboards, not treating completed agent runs as successful outcomes, and recognizing that review effort is part of the real cost of AI-generated code.

## Full article

daily.dev links to this article rather than hosting it. Read it at the original source: <https://thomasthornton.cloud/ai-assisted-engineering-measure-outcomes-not-activity>

## Community discussion

Top comments from developers on daily.dev.

**@capestart** · 2 upvotes

> I've noticed that AI makes activity go up almost automatically. That doesn't necessarily mean value went up with it.

**@jhon3484** · 1 upvotes

> **The core argument is right, and the warnings are the best part.** "A completed agent run is not a successful outcome" and "review effort is part of the cost" are the two things most teams get wrong. Leading with them would strengthen the piece.
>
> **The main problem: the proposed metrics aren't outcomes either.**
>
> Workflow success rate, guardrail failures, cost per accepted outcome — these are still activity metrics, just with a stricter gate. A real outcome is a shipped feature, a closed incident, a defect that didn't happen. The post frames this as _activity → outcomes_ when it's actually...

**@agustinbarrientos** · 1 upvotes

> I think cost per accepted outcome is the right headline. Token counts without review time and rework are just a prettier version of lines of code.

**@trevorsuna** · 1 upvotes

> Review burden absolutely belongs in the cost of an AI-assisted change. A run that produces a pull request quickly but needs an hour of careful correction is not a successful shortcut. Cost per accepted outcome is a much healthier signal than tokens or generated lines.

**@gertoska** · 1 upvotes

> Story points, lines of code, commits, PRs… and now tokens and AI-generated code. We keep gravitating towards the easiest thing to measure, rather than what actually represents value. And those metrics are also incredibly easy to game.
>
> Outcomes are harder and more abstract to measure, but they should always be the north star.

## Similar posts on daily.dev

- [Measuring engineering productivity is harder than ever](https://daily.dev/posts/measuring-engineering-productivity-is-harder-than-ever-aityuv2va) · LeadDev · 0 upvotes · 0 comments
- [How AI Is Reshaping Software Development and How Tech Leaders Should Measure Its Impact](https://daily.dev/posts/how-ai-is-reshaping-software-development-and-how-tech-leaders-should-measure-its-impact-crtkldrzx) · Security Boulevard · 0 upvotes · 0 comments

---

Tags: [#devops](https://daily.dev/tags/devops), [#github](https://daily.dev/tags/github), [#productivity](https://daily.dev/tags/productivity), [#observability](https://daily.dev/tags/observability)

[View this post on daily.dev](https://daily.dev/posts/ai-assisted-engineering-measure-outcomes-not-activity-kgafc31vm)