Track ChatGPT Traffic in GA4 | AI Assistant Metrics for Monthly SEO Client Reports

Find visits by AI service in the Traffic acquisition report

Measure monthly traffic by session. A session is a period of activity on a website. Use Session default channel group to classify the source of each session, including visits by new and returning users.

First user default channel group records how a user originally discovered the site. To include someone who first arrived through Google Search and returned through ChatGPT this month, choose the session-scoped dimension.

In the Traffic acquisition report:

1. Open Reports and select Acquisition, then Traffic acquisition
2. Set the date range to the reporting month
3. Choose Session default channel group as the table's primary dimension and read the Sessions value for AI Assistant
4. Use the plus sign beside the primary dimension to add Session source / medium and see the sources within AI Assistant

Source identifies where the visit originated; medium describes the traffic type. For example, chatgpt.com / ai-assistant identifies sessions classified as visits from ChatGPT. Use the source values actually shown in the client's property when reporting traffic by service. (Google Traffic acquisition report)

Google AI Overviews and AI Mode visits count as Organic Search

Visits from Google's AI Overviews and AI Mode are classified as Organic Search. Visits from services such as ChatGPT fall under AI Assistant, so the two groups cover different traffic sources.

AI Assistant includes sessions that GA4 can identify as originating from AI services. This total cannot reveal the origin of visits whose source is unknown. (Google default channel definitions)

GA4 measures activity on the client's site. Site visits alone cannot establish how many people read an AI answer or how often it cited the client's URLs. To report citations, inspect the links in AI answers and retain the question, answer, and cited URL as evidence.

For a measurement plan that includes both citations and visits, see four metrics for measuring AI search visibility.

Calculate the share of sessions with a completed inquiry form

To report completed inquiries, mark the event that records a successful form submission as a key event. A key event is an event designated as important to the business, such as an inquiry or purchase. Google recommends generate_lead for form submissions or requests for information. (Google recommended events)

Use the client's existing submission event when one is already implemented. Ask the measurement team to verify that it fires after a successful submission. Counting a form opening or a failed submit-button action as a completed inquiry inflates the measurement.

In Traffic acquisition, use the selector in the Session key event rate column to choose the completed inquiry form event. A site may also mark downloads or account registrations as key events. The rate for all key events can therefore differ from the rate for completed inquiries alone.

In this guide, lead form conversion rate means Session key event rate for that completed inquiry event. Calculate it for each channel as follows:

Lead form conversion rate
= Sessions in the channel with a completed inquiry event
  ÷ Total sessions in the same channel × 100

The numerator is sessions with a completed inquiry event. A session contributes once to that numerator even when the event fires several times. For AI Assistant, the denominator is also restricted to AI Assistant sessions. (Google key event rate definitions)

Key event counts depend on the counting method: once per event or once per session. Repeated submissions within a session can produce different event counts under those settings. Report the selected event's Session key event rate to express the share of sessions with an inquiry. (Google key event counting methods)

Show clients session counts alongside conversion rates

The following fictional example illustrates how to explain the figures. Assume a B2B service website serving Japan, with sessions filtered to Country = Japan across all devices for September 1–30, 2026. AI Assistant and Organic Search use the same dates, country, device scope, and completed inquiry event.

Channel	Sessions	Sessions with a completed inquiry event	Lead form conversion rate
AI Assistant	60	6	10%
Organic Search	1,200	24	2%

AI Assistant's 10% means 6 out of 60 sessions recorded a completed inquiry. Organic Search's 2% means 24 out of 1,200 sessions did so. AI Assistant has the higher rate in this example, while Organic Search has more sessions with a completed inquiry overall.

With the same total of 60 sessions, increasing the number with a completed inquiry from six to seven raises the rate from 10% to about 11.7%. A single session can noticeably change a rate with a small denominator. Include both total sessions and sessions with a completed inquiry alongside the percentage.

A client-facing explanation could read:

In September, 6 of the 60 sessions from Japan classified as AI Assistant recorded a completed inquiry, giving a rate of 10%. Organic Search recorded a completed inquiry in 24 of 1,200 sessions, giving a rate of 2%. AI traffic had a higher rate, while Organic Search accounted for more sessions with a completed inquiry overall. Next month, we will review the content and inquiry form links on articles and service pages receiving AI traffic.

These figures compare visits in two channels under matching date, country, and device conditions. Visitors' stages in the buying process and their landing pages may also influence the rates. A higher observed rate alone does not establish that the AI referral caused more inquiries. Examine page-level figures and copy to decide which information or form links need changes.

For month-to-month reporting, keep country, device, and event definitions consistent and state both date ranges. Include the number of days when comparing a 31-day month with a 30-day month, and note measurement changes beside the affected figures.

Review page content and inquiry form links on AI landing pages

Compare sessions and lead form conversion rates by landing page. A landing page is the first page viewed in a session.

Keep the selected inquiry event visible in Traffic acquisition and adjust the report:

1. Add a report filter at the top using Session default channel group with the value AI Assistant
2. Replace the Session source / medium secondary dimension with Landing page + query string
3. Compare page-level sessions and rates using the same dates, country, and device scope as the monthly report

The table displays the landing URL without the domain. For /pricing?lang=ja, /pricing is the page path and lang=ja after the question mark is the query string. (Google landing page reporting)

The rate for a landing page measures the share of sessions that started on that page in which an inquiry was completed. When a visitor starts on a comparison article, moves to the pricing page, and submits an inquiry within the same session, that session remains in the comparison article's row. Different query strings produce different rows, so include the URL when reporting a page's figures.

Page receiving AI traffic	What to review	Example change to propose
Product comparison article	Selection criteria and each product's features	Add supported use cases and features to the comparison table and link to each product's details
Pricing page	Setup fees, contract terms, and conditions requiring a quote	State costs and terms and link to the quote request page
Case study	Customer size and implementation steps	Describe the company's size and rollout process and link to an implementation consultation

For example, a comparison table for time and attendance software may list only product names. Add supported work arrangements and company sizes, then link to each product's features and pricing. Readers can assess which product fits their organization and check its costs and capabilities.

A pricing page should state the monthly fee and the features or services it includes. Explain when a custom quote is required and what information to provide so readers know what to ask. Record the URL and proposed changes in the monthly report, then compare that page's sessions and rate in the next reporting period.

Explain channel differences and propose specific page changes

Compare AI Assistant and Organic Search session counts and lead form conversion rates in the monthly client report. Explain how many sessions from each channel included a completed inquiry. Include AI service source values, dates, country and device filters, and the event used to record a completed inquiry.

Specify the target URL, content to add, and link placement in each page recommendation. The client can then identify the actual changes to make.

Please Cite Me! is a reporting tool for SEO agencies that checks whether a page is properly published and whether its content is complete. Its page audit helps you review access and whether readers can understand, verify, and cite the information on the page.

Try Please Cite Me! to audit access and content on the articles and pricing pages receiving AI referral traffic!

Analyze a page with Please Cite Me! →