KARI TOOLKIT / PROGRAMMATIC / RECONCILIATION
Placement discrepancy checker
Compare DSP and ad server impression counts, choose your calculation reference, and export the placements that need review.
Free to use. Runs in your browser. No signup or uploads.
Align the reports first. Use matching placement IDs, date ranges, timezones, filters, metric definitions, and fresh reports. A flag is a reason to investigate.
Your review queue
2 TO REVIEWValid placements4
At threshold or zero reference2
Total DSP impressions249,000
Total ad server impressions224,500
Reference: ad server impressions. Review threshold: 10%.
| Placement | DSP | Ad server | DSP minus ad server | Difference % | Status |
|---|---|---|---|---|---|
| Prospecting Display | 120,000 | 100,000 | +20,000 | 20% | At or above threshold |
| Retargeting Display | 45,000 | 44,500 | +500 | 1.12% | Below threshold |
| CTV Awareness | 82,000 | 80,000 | +2,000 | 2.5% | Below threshold |
| New Placement | 2,000 | 0 | +2,000 | Undefined | Review: zero reference |
Aggregate difference: 10.91% of total ad server impressions. Opposite placement differences can cancel in totals. Small counts can produce large percentages.
Export includes both counts, signed and absolute differences, reference, threshold, and reason. Invalid rows are excluded.
A comparison, then an investigation.
Counting events, reporting latency, filtering, and creative setup can produce different counts. Check the context before escalating a tracking or billing concern.
Google’s guide to reporting discrepanciesUseful once. Even better, connected.
We can explore how this step fits into an automation built around your agency’s tools.