Recommendations & Actions
How Growthract decides what to recommend, and what 'Best Next Move' means
Recommendations are ordered by fixed, explainable rules — severity first, then what the issue affects, then where it is, then whether it was measured.
Short answer
Growthract orders every recommended action with fixed rules. In order, it looks at the severity of the check, what the issue affects (whether search and AI systems can reach your pages, understand them, or recommend them, in that order), whether it is on the homepage, and whether the scan measured it or a model suggested it. "Best Next Move" is the first action in that order, and each action says why it sits where it does.
What this means
The order is deterministic — the same report always produces the same order, and an AI model is never used to re-prioritize it. Severity (how serious a condition could be) and priority (the order to work in) are different: severity comes from the check, priority is the resulting order. Effort is a typical estimate, not a measurement, and the owner type ("Needs developer" vs "Needs content owner") is estimated from the action's own description. Neither is a verified assignment.
Why this happens
Ordering by fixed rules, rather than a model's opinion, keeps the order consistent and explainable between runs. It does not decide whether a condition is intentional: a noindex directive on a login page, for example, may be exactly what you want, so check the captured evidence before changing anything.
What to do
Start with Best Next Move if you want the single highest-leverage fix. Use the Highest-impact, Lowest-effort, or owner-type filters in Actions if you want to work through the list differently. Marking an action "implemented" records that you did the work — it does not by itself confirm the fix took effect. See Verification statuses for that next step.