Modeled vs observed conversions
Observed conversions are directly recorded from events that the system actually saw. Modeled conversions are statistical estimates that fill gaps left by consent declines, cross-device journeys, or blocked tags. Modern reports blend both, so understanding which conversions are measured versus estimated is essential to reading a total honestly and not treating an estimate as a count.
What this means
Observed conversions come from events the platform genuinely recorded and could tie to an interaction. But consent declines, tracking prevention, and cross-device hops leave gaps where conversions happened but could not be directly attributed. Modeled conversions estimate that missing volume using patterns from the observed data.
Reports increasingly present a blended total. The number is partly fact (observed) and partly estimate (modeled), and the two have very different epistemic status even though they appear in one figure.
Why the distinction matters
Treating a modeled total as if it were exact leads to false precision — comparing two periods whose modeled shares differ, or trusting a small segment that is mostly estimated. The modeled portion has uncertainty and depends on having enough observed data to train on; thin segments model poorly.
Google documents conversion modeling and when modeled conversions are included. Good practice is to know your modeled share, lean on observed conversions for high-stakes decisions, and read modeled totals as directional. Keep an observed first-party baseline so you can see how much of any movement is real versus re-estimated.
- Observed = directly recorded; modeled = statistically estimated
- Blended totals mix fact and estimate in one number
- Modeled shares carry uncertainty and need training data
How it appears in analytics and logs
A reported total that includes modeling is a best estimate for the unobserved portion, not a literal count; the modeled share carries uncertainty.
Diagnostic use case
Distinguish modeled from observed conversions in a report so you know how much of a total is directly measured versus statistically estimated.
What WebmasterID can help detect
WebmasterID emphasizes directly-observed first-party events, giving a clear observed baseline to compare against modeled platform totals.
Common mistakes
- Treating a modeled total as an exact, observed count.
- Comparing periods without noting their differing modeled shares.
- Trusting modeled numbers for thin, low-data segments.
Privacy and accuracy notes
Conversion modeling estimates aggregate outcomes precisely to avoid tracking individuals through gaps. This page is educational, not legal advice.
Related pages
- Modeled conversions
Modeled conversions are conversions a platform estimates statistically rather than observes directly. When direct measurement is blocked — by missing consent, cross-device journeys, or privacy protections — ad and analytics platforms model the likely conversions from observable trends and aggregated data, and report them alongside observed ones. Understanding which conversions are modeled is essential to reading attribution honestly.
- Consent and attribution
Consent is upstream of attribution: under frameworks like the EU's GDPR and ePrivacy Directive, storing or reading identifiers for tracking generally requires the user's consent. When consent is declined or withheld, the touchpoints those identifiers would have recorded never enter the data, so attribution operates on partial paths. Understanding consent is therefore inseparable from reading attribution honestly.
- Attribution Reporting API summary reports
The Attribution Reporting API is a Privacy Sandbox proposal that lets browsers measure ad conversions without third-party cookies or cross-site identifiers. It produces event-level and aggregatable reports; aggregatable reports are combined into noisy summary reports that give campaign-level conversion counts and values while limiting what can be learned about any individual.
- Privacy-first analytics
Keep a clear observed first-party baseline.
Sources and verification notes
- Google Analytics Help — About conversion modelingDocuments modeled vs observed conversions and when modeling applies.
Last reviewed 2026-06-24. Facts are checked against primary/official sources where available; uncertain specifics are marked “Data not yet verified” rather than guessed.