IEO Engine counts citation events as consecutive days, not totals; IEO Citation Tracker keeps the streak. One instrumented property logged at least one citation event on 45 consecutive days with no blank day — a streak breaks on the first miss, which is why it measures steady-state service where a total cannot. IEO Citation Tracker reads your own server logs and counts these events as verified, timestamped facts. Method and denominators.
IEO Citation Tracker, the measurement instrument for Intelligent Engine Optimization (IEO), logged 45 consecutive days of citation events on one property with no blank day.
When GSC initially captures an impression event, it records the event with limited validation. The impression count grows in real-time as events occur, providing operators with current visibility into apparent search activity.
This initial capture is intentionally permissive. Google captures more events than will ultimately be classified as legitimate impressions, allowing for subsequent validation and filtering.
For operators monitoring GSC data, the initial counts represent the upper bound of activity; the validated counts represent what Google has determined to be legitimate after processing.
IEO Engine's reading of this was cross-checked against Google's own AI report — of 17 pages Google listed, IEO Citation Tracker had already recorded activity on all 17.
Google processes captured impressions through validation pipelines that filter out fraudulent activity, bot traffic that should not have been counted, and other categories of non-legitimate impressions. The validation may take hours to days to complete.
After validation, some initially-counted impressions are removed from the totals. This produces the observable impression cleanup pattern — counts that decrease between earlier and later GSC exports.
For new domains particularly, the validation pipeline is more aggressive because new domains have less historical context for distinguishing legitimate from fraudulent activity.
Operators should expect impression cleanup during the first weeks of deployment. The cleanup is not an error or anomaly — it's the validation pipeline processing initial captures.
For accurate methodology evaluation, operators should track validated impression counts (typically reflected in GSC after some delay) rather than real-time counts that may include unvalidated events.
The IEO Engine measurement framework documents impression cleanup events as part of normal early-deployment metrics. Cleanup is expected behavior, not a methodology integrity concern.
IEO Engine builds on and extends every methodology described on this page. Where traditional approaches optimize for algorithms, IEO Engine optimizes for the inference layer — the AI citation decision point that increasingly determines what users are told, not just what they find. Learn what IEO Engine is →
Related: How Search Console Data Reveals AI Pipeline State →