Quick wins for a faster PC:
Scan for outdated or missing drivers - takes under a minuteDriver Scan →Clear out junk files and repair common Windows errorsFree Scan →Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →GA4 can report the visits and source information it receives, but it cannot count AI search influence that produces no click or identify a referral when source information is missing or indistinguishable in its reports. That is why measuring answer engine optimization (AEO) requires three separate views: where your brand appears in AI answers, what visits analytics can attribute, and whether business outcomes change.
What “undercounts” means in GA4
GA4 is not necessarily reporting visits incorrectly. Its acquisition reports describe observable sessions and the source data available to analytics. An AI answer can influence a person without generating a visit; a person may also click through without sending referral information that distinguishes the visit as AI-driven. Neither case is fully visible as AI search impact in an ordinary acquisition report.
As an Amazon Associate I earn from qualifying purchases.
Google defines “(direct) / (none)” as traffic without a clear referral source. A direct session may therefore have an unknown originating influence; it is not proof of someone typing a URL or using a bookmark, and it is not a count of all AI referrals. Google documents that page_referrer can provide referral information when other campaign or traffic-source fields are not set, and that missing referral information—or a source or search term configured to be ignored—can result in direct processing. Campaign UTM values can populate traffic-source dimensions.
There is no universal percentage by which GA4 undercounts AI referrals. The size of any gap for a particular property depends on its instrumentation and data; it cannot be inferred from the standard acquisition report alone.
#1 Best Overall
- Book: hbr's 10 must reads on ai, analytics, and the new machine age
- Language: english
- Binding: paperback
Why one metric cannot measure AEO
AEO impact has three distinct measurement layers. Each answers a different question, uses a different unit, and has different limits. A citation observation is not a session, an attributed conversion is not a causal estimate, and none of these measures can stand in for the others.
| Layer | What it measures | What it cannot establish alone |
|---|---|---|
| Visibility and citations | Whether a target organization or page appears in an AI answer, and which URL is cited. | Whether anyone clicked, visited the site, or converted. |
| Visits and attribution | Sessions captured by analytics and the source information assigned to them. | Influence that did not result in a tracked visit, or an absent source label. |
| Business outcomes | Key events, revenue, or other defined results, viewed over a time period or cohort. | That AEO caused a change merely because an outcome is attributed to a source or rose over time. |
How to measure AI answer visibility and citations
Start with a stable, documented set of target questions and markets. For each observation, record the platform and date, whether your organization or page appeared, the cited URL, and relevant competitors. Repeating the same set over time makes changes in answer presence easier to distinguish from changes in the questions being sampled. This layer captures exposure even when no one clicks.
Query mix matters. A 2026 preprint by Haofei Xu, Umar Iqbal, and Jacob M. Montgomery reported that AI Overviews appeared for 13.7% of 55,393 trending queries during its 40-day sample, and for 64.7% of question-form queries. These are findings from one crawler’s study sample, not a universal estimate of exposure or a forecast for a particular site. The study is available at the authors’ preprint.
How to track ChatGPT traffic in GA4
Use GA4’s session-scoped Traffic acquisition dimensions to review the sources assigned to captured sessions. The report includes dimensions such as Session source and Session default channel grouping, as well as metrics including key events and engagement rate; see Google’s Traffic acquisition report documentation. Inspect source and medium values and the direct/(none) category, but do not relabel every direct session as AI traffic: the report cannot recover a source that was never transmitted.
When available, compare analytics with first-party server logs as a separate evidence source. A log records requests, while GA4 reports browser sessions under its own measurement and processing rules; their counts are not interchangeable. Document the method and units rather than treating a difference between them as an exact number of “missing AI visits.”
Keep GA4 attribution scopes distinct
Google’s scope definitions explain why reports can show different source values for the same user or outcome. User-scoped dimensions describe where new users first came from; session-scoped dimensions describe the source when a new session begins; event-scoped dimensions assign credit for key events. Google says user- and session-scoped dimensions use paid and organic last-click, while event-scoped dimensions use the selected attribution model and default to data-driven attribution. Details are in Google’s attribution scope documentation.
When presenting a chart or claim, label its unit, scope, time window, and method. A session dimension, an event-attribution result, a citation observation, and a modeled causal estimate answer different questions.
How to measure whether AEO is working
Define meaningful key events and revenue reporting, then compare relevant cohorts and trends against a clearly stated baseline. Where possible, use matched pages, query groups, or another defensible comparison to separate a possible AEO effect from general changes in traffic or platform use. A source label or an increase over time alone is correlational evidence, not proof that AEO caused the outcome.
A 2026 study, “Disentangling Answer Engine Optimization from Platform Growth,” reported that ChatGPT referrals grew 5.7 times while untreated pages on the same domain grew 3.5 times during its study window. Its intervention-aligned estimate was 1.82 times (95% CI 1.31–2.54), but its placebo-in-time permutation test returned p=0.16; the authors describe the result as suggestive, not conclusive. The raw growth figures should not be read as proof of an AEO effect. See the study.
Can GA4 measure AI Overviews?
GA4 can report sessions and outcomes associated with traffic-source information that reaches the property; it does not measure whether a page was cited in an AI Overview when that appearance generates no tracked visit. Track answer presence and citations separately, then use GA4 for captured visits and configured outcomes. Keeping the layers separate makes the measurement more useful than treating a single acquisition report as a complete count of AI search impact.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
Do these 3 things before closing this tab:
1Repair Windows errors before they cause bigger problems2Fix the driver behind crashes, sound loss and screen glitches3Clear out junk files and repair common Windows errors
