Understand & review results
A useful answer should make its data context understandable. Review the result before turning it into a saved layer, an export, or a project change.
Before you start
Keep the original question and source data in view. Confirm which application, project, and workspace produced the result. The labels described here refer to supported Mapplex AI workflows.
1. Check what was included
Review the referenced layers, filters, record counts, and scope. Distinguish a whole-layer result from a selected-feature or map-area result. A preview map is a display of the result; it does not independently define the analytical scope.
2. Check the operation
Read the generated SQL or review logic where available. Confirm the field names, distance units, comparison values, and treatment of missing data.
For spatial outputs, inspect the geometry and source coordinate context. If the result looks displaced or unexpectedly empty, resolve the underlying data or coordinate issue before continuing.
3. Read the evidence and warnings
Supported deterministic field checks can show these evidence labels:
| Label | Meaning |
|---|---|
| Decision-grade | The exact, complete deterministic evidence met the configured quality and freshness conditions |
| Review required | An exact result has a warning, freshness concern, or excluded-data concern |
| Advisory only | The result is approximate, sampled, incomplete, a fallback, or did not meet a validation condition |
An evidence label describes the data path. It does not replace the responsible person’s review or approval.
4. Choose the next action
- Refine: clarify the question or fix source data when the result is not suitable.
- Save or export: inspect the result first, then use an available output supported by your workflow.
- Confirm a suggested field change: review the current and proposed value, source, and rule effects. Apply only through the explicit confirmation flow.
Controlled field suggestions remain subject to roles, record ownership, form rules, and valid values. Undo is available only where the current record state still permits a safe reversal.
Expected result
You can explain what data was used, how the result was obtained, what limitations remain, and why the next action is appropriate.
If the evidence is unclear
Do not infer certainty from a confident sentence or a map preview. Narrow the question, inspect the source records, or ask the project reviewer to check the result. See AI troubleshooting and the full response reference.
AI credits
In Mapplex, a validated technical answer uses one workspace AI credit, including answers produced by deterministic checks. Clarifications, failed operations, suggestions, and conversational replies are not billed as technical answers. A verified answer of zero can still be billable.
See AI credit rules for scope and retry details.