> ## Documentation Index
> Fetch the complete documentation index at: https://docs.getprova.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Common Data Issues and Prevention

> Learn the most common data issues that affect reporting and modeling, and how to reduce them before they create noise.

## Overview

Most measurement problems do not begin in the model. They begin in the data.

## Common issues

### Missing days

Gaps in spend, impressions, conversions, or revenue can distort timing, trend, and contribution views.

### Zero-spend days that are not real

Unexpected zero rows can make channels look paused or broken when they were not.

### Inconsistent naming

Naming drift makes mapping and interpretation harder and can create false fragmentation.

### Hierarchy shifts

When the same media source moves between channel levels or naming conventions, trend comparisons become harder.

### Partial uploads or stale files

A file can be structurally valid but still incomplete or older than expected.

## Prevention habits

* Keep delivery cadence predictable
* Preserve stable naming
* QA new sources before relying on them in reports
* Review Data Layers and Connector QA after changes
* Treat anomalies as a data question first, not only a performance question