What John Cutler's Data-Informed Product Cycle Taught Our Usage Intelligence Agent

John Cutler has been writing about the mess of product development since 2015. He spent four years as a product evangelist at Amplitude, co-authored the North Star Playbook, and has since worked as an independent product coach. His central preoccupation is not frameworks for their own sake but the organizational conditions that make good product thinking possible or impossible.

What Is Actually Worth Learning

Cutler's most transferable insight is about the failure mode that precedes bad measurement: most teams jump from high-level strategy straight to feature ideas, skipping every intermediate step that would tell them whether a feature is delivering value. His data-informed product cycle — Strategy, Models, Metrics, Leverage Points, Place Your Bet, Integrate Learning — is the sequence that fills that gap.

The vanity metric test is one of his sharpest tools. A metric is vanity if 'the result is not actionable. Regardless whether the metric goes up or down, we don't change what we do.'

• •Words before numbers: clarify what you are trying to learn in plain language before selecting a metric, not after.
• •Measure to learn, not to control: healthy data-informed teams 'iterate on what and how they measure' rather than locking in a metric as a performance target.
• •Models precede metrics: a metric without a logical model of value delivery is a proxy for nothing.
• •Friction is the enemy: 'from curiosity to insights to action with as little friction as possible.'

How a Method Becomes a Skill

The data-informed-product-cycle skill encodes the six-stage loop as an explicit procedure. The agent anchors first to strategy and stated adoption goals before touching usage data. It then maps observed workflow patterns against the assumed model of value delivery, applies the actionability test to every metric it surfaces, and identifies the two or three leverage points where a small change in behavior or instrumentation will matter.

The one principle we found hardest to encode was the simplest: words before numbers. Getting an agent to articulate what success looks like in plain language before querying for a metric requires forcing the strategy articulation step to happen explicitly. The skill does this by making Step 1 a mandatory anchor against the adoption dashboard and stated goals.

One of More Than 400

This skill is one of more than 400 method-named skills across 19 agents in our swarm. An agent running a procedure is not the same as a practitioner making judgment calls. The skill is a starting point, not a replacement.

A note on this post: This is independent commentary and homage. It distills publicly available writing and talks by John Cutler to illustrate a working method, and every quote is drawn from and verified against the primary sources linked above. The skill it describes is named for the method, not the person, and contains no marketing claims attributed to them. Data Workers is not affiliated with, sponsored by, or endorsed by John Cutler. If you are John Cutler and would like anything adjusted or removed, email hello@dataworkers.io and we will respond promptly.