What is a Data Integrity Report?
Every day, Polar runs automated pipeline tests that compare your connector data against each platform's source API. If a metric drifts outside an acceptable tolerance, a warning appears inline on the affected connector — and you can open a full report to see exactly where and why the difference is occurring.
These tests give you a transparent, daily answer to "are these numbers right?" for every supported connector.
What it looks like
When a test fails outside tolerance, a warning banner appears on your account and the connector card:
View data integrity report
Opening the report shows a breakdown per test:
View details
Status definitions
Each test in the report has one of three statuses:
Status | What it means |
OK | Polar values match the source API within the tolerance threshold. |
Failed | The cumulative mismatch across the period exceeds the tolerance threshold. Polar's support team will be alerted automatically. |
Tolerance thresholds
Tolerances are set per connector to account for differences in how platforms report data. The threshold is calculated as the magnitude of mismatch across the total sum of values for the period — not the number of mismatched days.
Connector | Metric(s) checked | Tolerance |
Shopify | Total Orders, Gross Sales | 1.5% (Orders) · 5% (Sales) |
Facebook Ads | Spend, Clicks, Impressions | 5% |
Google Ads | Spend, Clicks, Impressions | 5% |
TikTok Ads | Spend, Clicks, Impressions | 5% |
GA4 | All available tests | 5% |
Note on GA4 Engaged Sessions: Minor discrepancies between the GA4 API and Polar (e.g. a difference of ~40 sessions per day) are expected due to how GA4 aggregates sessions server-side. These will appear as "OK (within tolerance)" and do not indicate a data problem.
Re-running tests manually
Each connector's integrity report includes a Re-run tests button. This is useful if you've recently reconnected a data source and want to check whether the mismatch has resolved.
Navigate to Connectors and open the affected connector.
Click View data integrity report in the warning banner.
Click Re-run tests at the top of the report page.
Results will refresh — if the issue has been resolved, the warning banner will disappear.
Re-run is rate-limited. If you click while a run is already in progress, the request will be silently skipped until the cooldown clears (usually a few minutes).
F.A.Q.
A test shows "Failed" but my reports look fine. Should I be worried?
A test shows "Failed" but my reports look fine. Should I be worried?
A failure flag means the cumulative mismatch over the reporting period exceeded the tolerance threshold. In practice, this often reflects a known platform discrepancy rather than missing data.
Why does Polar sometimes report slightly more data than the source API?
Why does Polar sometimes report slightly more data than the source API?
Ad platforms and analytics tools can revise historical data after the fact — for example, removing fraudulent clicks or adjusting session counts. Polar may have captured data before that revision occurred. Small overages are normal and fall within the tolerance range.
Does the Data Integrity Report cover all metrics in my dashboard?
Does the Data Integrity Report cover all metrics in my dashboard?
Not yet. The tests cover a curated set of high-signal metrics per connector (spend, clicks, impressions for ad platforms; orders and gross sales for Shopify). Coverage will expand over time.