# Resolve experiment report warnings

Restore reporting readiness without hiding data or allocation problems.

Start with the status banner on the experiment's Results page. It identifies which condition blocks analysis.

## No data or incomplete mapping

Confirm the provider connection, mapped external experiment, primary metric, and all variation mappings. Each provider variation needs a one-to-one match. Correct the mapping and sync before checking Results again.

## Unsupported or invalid configuration

The current fixed-horizon report supports rate-per-user primary metrics. Check the metric type, baseline, expected traffic allocation, and observed counts. A descriptive total is not proof that the configuration is valid for inference.

## Fixed-horizon analysis is pending

Confirm that the planned end date has passed and the result status is stopped. While the test is still running or before the planned horizon, totals remain descriptive. Do not mark a running experiment stopped simply to reveal a conclusion.

## Sample Ratio Mismatch

Compare the intended provider allocation with the allocation recorded in ConversionLab, then investigate why observed traffic differs. Check mapping, targeting, tracking, and the provider's experiment setup. Significance remains gated while the mismatch is unresolved; changing expected percentages merely to match observed totals does not resolve the cause.

## Insufficient data, stale source, or unavailable engine

Insufficient data means one or more analysis checks have not cleared. Stale source data needs another sync. Temporary engine unavailability can leave descriptive totals visible while analysis is unavailable; retry after recovery.

When asking for help, include the experiment ID, status message, last sync time, metric type, and the mapping you expected. Keep the observed totals and underlying source available. See [reading results](https://conversionlab.app/docs/experiments/results) for interpretation once the warning is resolved.