Understanding Your Score

The AI Adoption Score blends three signals from real usage into a single percentage — here is what goes into it and what stays out.

The score is built from three dimensions of how your organization uses AI coding. Each contributes qualitatively; MegaBrain does not surface a fixed formula, but understanding the inputs explains why the number moves.

What the score is derived from

  • Frequency — how often people reach for AI in their day-to-day work.
  • Depth — how much of the development workflow AI actually touches, from prompts to agent modes to reviews.
  • Coverage — how many people across the organization are active, not just a handful of heavy users.

The 3-day floor and where the data comes from

A meaningful score requires at least 3 days of AI usage. Before that, the AI usage tab shows a “collecting data” empty state instead of a number — MegaBrain will not fabricate a score from too little data. Everything it does show is computed from real MegaBrain Gateway usage attributable to your org, using per-user, per-model, and per-project signals.

What it does not measure

  • It is not a productivity or output metric — it measures adoption, not lines shipped or tickets closed.
  • It is not a ranking of individuals; per-user signals exist to guide enablement, not to grade people.
  • It does not judge code quality or correctness — those belong to Code Reviewer and the Security Agent.
  • It does not capture AI work done outside MegaBrain Gateway, so only Gateway-attributable usage counts.