What happens when Data Fix detects too many issues in my data?

If the Data Fix preview flags more than 10% of your readings as needing correction, the engine shows an "excessive fixes" warning. This is an important signal — it means the underlying data source likely has a systematic problem that Data Fix alone cannot properly resolve.

Why more than 10% is a red flag

Data Fix is built for patching isolated anomalies — a missing reading here, a brief negative glitch there. When a large portion of your dataset needs correction, the issue is usually upstream: a misconfigured API connection, a malfunctioning meter, a broken CSV export, or a timezone mismatch in your import settings. Applying Data Fix to a fundamentally broken dataset would create a large volume of estimated values rather than real measurements, degrading data trustworthiness.

What you should do instead

1. Check your data source — Go to Analytics → Data Import and verify the connection status, last successful fetch, and error logs for the affected meter.
2. Check the meter hardware — If readings are consistently wrong, the physical meter may need recalibration or replacement. Contact your facility manager or utility provider.
3. Check import configuration — Verify the correct columns, units, and timestamp format in your import settings. A wrong unit (e.g., MWh parsed as kWh) will cause every reading to appear out of range.
4. Fix the root cause first — Once the upstream issue is resolved, re-import clean data. Then use Data Fix only for the small number of residual anomalies.
5. Run Data Fix on short periods — If you must fix a portion, use a narrow time range (1–2 weeks) and verify results before expanding.

The warning does not block you — you can still proceed with preview and apply. But the warning exists to encourage you to investigate the root cause rather than patching over a systemic problem.

Where to check: Analytics → Data Fix for the warning. Analytics → Data Import for connection status and error logs.