The fine print.
In plain sight.

Official documents

Public policies

Choose the document that matches what you need. Each policy is written as a standalone public reference.

Recommendation transparency

What each number means

STRATA keeps three different kinds of judgment separate so none is presented as stronger evidence than it is.

Official FitScore

An editorial shortlist score out of 100 for hypertrophy-oriented exercise selection. It begins with a rounded weighted baseline—30% target stimulus, 20% useful range, 20% stability, 20% progression, and 10% fatigue cost—then permits a disclosed exercise-specific adjustment of no more than ±4 points. Research informs the factors; no cited source claims STRATA’s exact numerical ranking.

Personal match

A separate reordering of the same library using the goal, experience, equipment, preferences, and optional movement constraints you choose. It does not alter FitScore, diagnose injury risk, predict an individual result, or inspect private workout history.

Account progression suggestion

A private, deterministic next-session prompt derived only after you record a completed result comparable in exercise, measurement, load type, and unit and submit a 1–5 workout check-in. A first result is a baseline. Missing or difficult check-in signals keep the target unchanged. It is separate from FitScore and personal match, does not diagnose pain, recovery, readiness, or injury risk, and never edits a plan without explicit approval.

Community rating

An aggregate of submitted member ratings shown separately from FitScore. It is lived-experience context, not controlled research, and it never silently changes the editorial score.

Recommendation check

An optional answer about whether the shortlist felt useful, relevant, or clear. It stays on the device unless aggregate sharing is enabled; then only the response category increments a daily count. It is not published as evidence and does not train or automatically change the ranking model.

Product activity counts

Opt-in daily counts of allowlisted actions help identify where the product may need attention. They are directional activity—not unique people, conversion cohorts, clinical outcomes, or proof of effectiveness. Each action counts once per account or network per day, and automated actions can still increment anonymous counts.

A ranking is a starting point, not a guarantee. Exercise tolerance, technique, access, progression, and qualified professional advice still matter.