What Each Dimension Reveals

Here’s what each of the five dimensions is actually measuring — and what different scores tend to signal.

Dimension 1: AI Interaction Patterns

Are your AI conversations intentional or improvisational?

High scores here mean your team consistently defines a clear outcome before starting an AI conversation, uses structured prompts rather than improvised chat, and treats prompting as a skill worth developing.

Low scores often show up as “AI roulette” — giving AI a rough idea and hoping it figures out what you meant. The output quality varies wildly, and the team doesn’t understand why.

The signal: Teams stuck at Level 1–2 on this dimension often describe AI as “hit or miss.” That’s rarely about the AI. It’s about the quality of the brief.

Dimension 2: Iteration & Rework Discipline

When AI misses, do you course-correct or restart?

High scores here mean your team has a systematic approach to refinement — you diagnose why the output missed, adjust the prompt, and iterate toward the target. AI-generated work typically needs only minor revision before it’s usable.

Low scores look like: getting one output, deciding it’s not right, starting a completely new conversation, and repeating until something acceptable appears.

The signal: Low rework discipline is the primary driver of “AI takes longer than doing it myself.” The time cost isn’t in the generation — it’s in the undisciplined iteration.

Dimension 3: Clarity of Inputs

Are your team’s briefs actually good enough to produce good outputs?

AI is only as good as the context you give it. High scores here mean your team provides rich context, defines constraints, and specifies what success looks like before asking for output.

Low scores mean vague asks, missing context, and frequent “that’s not what I meant” moments — which then get attributed to AI limitations rather than input quality.

The signal: This dimension often reveals writing and communication gaps that predate AI. Teams that wrote vague briefs before AI write vague AI prompts.

Dimension 4: Shared AI Practices

Does your team have a consistent approach, or is everyone doing their own thing?

High scores mean shared prompt libraries, documented approaches, peer review of AI-assisted work, and team-level learning. When someone finds a better approach, it spreads.

Low scores mean knowledge stays individual. One person develops sophisticated prompting skills; their teammates never benefit. The team’s aggregate AI capability doesn’t compound.

The signal: Individual AI excellence doesn’t scale. This dimension is where the difference between team capability and individual capability lives.

Dimension 5: Governance & Measurement

Do you know what AI is deciding, and whether it’s working?

High scores mean your team tracks AI effectiveness, has clear guidelines on what decisions AI should and shouldn’t make, and reviews AI work before it affects customers or stakeholders.

Low scores mean AI is making decisions your team doesn’t fully realize it’s making, with no feedback loop on quality.

The signal: This dimension is often the last to develop — and the most consequential to skip. Teams that don’t measure AI effectiveness don’t actually know if it’s helping.