What Concord verifies before an action reaches you, and what it fixes on its own.
Every platform action is validated before you are asked to approve it. The point is that the thing you review is already coherent and already accepted by the platform’s rules — so approving is a decision about intent, not a proofreading exercise.
Required fields — every field the platform requires for that object type and campaign type, at every tier of the tree.Coherence — budgets against flight dates, start dates against end dates, child objects against their parent’s configuration.Platform rules — allowed objectives, valid bid strategy and budget combinations, formats and character limits, values that exist for that account and that country.Quotas and limits — anything the platform caps, such as the number of targeting criteria on an object.Your conventions — the naming rules and checklists carried by your Skills.Availability for this account — targeting values are resolved against what your advertiser can actually target, rather than assumed. A country or interest your account cannot target is reported, not silently swapped for a neighbour.
The action is not submitted. The failure names the object and the field, so the fix is a message away: correct it in the conversation and Concord prepares the action again.
The most common source of avoidable failures is a rule that lives in someone’s head. Once a check fails for the same reason twice, put it in a Skill — Concord will apply it before you see the action next time.
