adview_query_id: Google’s AI Mode Attribution Stamp You’re Losing

What Is adview_query_id?

Google is already attributing AI Mode conversions via a URL parameter most WooCommerce stores have never heard of — and aren’t capturing.

Discovered Labs published the first detailed network analysis of Google AI Mode in January 2026. Across 547 AI Mode flows and more than 1,300 background requests, they identified a parameter Google stamps on outbound click-throughs from AI Mode: adview_query_id (Discovered Labs, 2026).

Here’s the thing… This parameter isn’t experimental. It flows through first-party conversion tracking (google.com/pagead/1p-conversion), third-party conversion tracking (googleadservices.com/pagead/conversion), and Google’s analytics and Tag Manager infrastructure. It creates a direct attribution chain: AI Mode query → ad impression → click → conversion.

The catch? Google hasn’t shipped the reporting column. Your store is receiving AI Mode-attributed traffic right now, stamped with adview_query_id, and your analytics setup is discarding the stamp before it can be stored.

You may be interested in: Why ChatGPT Traffic Shows as Referral, Organic, AND Direct in GA4 — And Which Number to Trust

The Scale Problem: 1 Billion Monthly Users, Zero Reporting

AI Mode has crossed the billion-user mark, queries are doubling every quarter, and the purchase-intent signals are outpacing every other query type.

At Google I/O in May 2026, Sundar Pichai confirmed that AI Mode surpassed 1 billion monthly active users globally, roughly one year after launch (Google, 2026). Queries have more than doubled every quarter since launch. The rollout now spans nearly 200 countries across close to 100 languages.

AI Mode has surpassed 1 billion monthly active users with queries more than doubling every quarter since its May 2025 launch — and planning queries (the kind that drive purchase decisions) are growing 80% faster than other query types (Google I/O, May 2026).

Those planning queries are the ones that matter for WooCommerce. When someone asks AI Mode “what’s the best WooCommerce hosting for a store doing $50K/month,” that’s a buying signal. Planning queries in AI Mode are growing 80% faster than other query types (Google I/O 2026). And when that user clicks through to your store, the adview_query_id parameter is attached to the URL.

But your GA4 property can’t tell you it happened. There’s no “AI Mode” channel, no dimension for adview_query_id, no segment to isolate these high-intent sessions. The data is flowing. The reporting is silent.

How the Parameter Works in the Wild

The attribution chain runs from the AI Mode query through a 60-millisecond auction to a stamped click-through — all before the user sees a product page.

Discovered Labs’ network analysis revealed the mechanical sequence. When a user submits a query in AI Mode, Google’s ad infrastructure runs a server-side auction that resolves within approximately 60 milliseconds (Discovered Labs, 2026). That auction determines which ads are eligible for the session, even if the ads aren’t visually displayed yet.

The parameter adview_query_id follows the user through the entire session:

The infrastructure is production-grade. Ads now appear in 25.5% of AI Mode results, a 394% year-over-year increase (Optimyzee, 2026). A new ad placement called “AI Mode Bottom Ads” (aimba) has delivery infrastructure ready in the JavaScript bundles. The question isn’t whether Google will monetise AI Mode. It’s whether your store’s measurement is ready when it happens publicly.

Why Client-Side Tracking Loses adview_query_id

Ad blockers, Safari ITP, and single-page application rewrites strip URL parameters before your client-side tags can fire — and adview_query_id is no exception.

Client-side JavaScript can read adview_query_id from the URL. In theory. In practice, three forces strip it before your tracking code fires:

Ad blockers affect roughly 29.5% of internet users worldwide — an estimated 1.77 billion people (GWI, Q2 2025). The US sits at 32.5%. When an ad blocker kills GTM, every tag inside it dies — including any custom JavaScript reading URL parameters. The parameter arrives on the landing page URL, but the script that would capture it never executes.

An estimated 29.5% of internet users worldwide — roughly 1.77 billion people — use ad blockers, and when those blockers kill Google Tag Manager, every tag inside it dies with the container, including any script that would capture adview_query_id (GWI via Backlinko, 2025).

Safari’s Intelligent Tracking Prevention adds a second layer. ITP caps first-party cookies set via JavaScript to 7 days and applies further restrictions to URL-decorated links from known trackers. A user who clicks through from AI Mode on Safari and returns to purchase 8 days later has already lost the session connection.

Single-page application architectures introduce the third failure mode. When a WooCommerce store uses a headless front end or any JavaScript-based routing, the initial URL parameters can be overwritten during client-side navigation. The parameter was there on the first request. It’s gone by the time the purchase event fires.

You may be interested in: GTM Is the Single Point of Failure — When Ad Blockers Kill Your Tag Manager, Every Tag Dies with It

Server-Side Capture: The Only Reliable Pattern

Middleware reads adview_query_id from the HTTP request before any client-side interference occurs — and persists it through to the WooCommerce order.

The reliable architecture captures the parameter at the server layer. When a request hits your WordPress/WooCommerce server, middleware reads adview_query_id from the query string at the HTTP level — before ad blockers, ITP, or JavaScript routing can interfere. The value is stored in the server-side session and attached to the WooCommerce order as meta when the purchase completes.

This pattern has three structural advantages over client-side approaches:

First, it survives ad blockers entirely. The parameter capture happens at the web server, not in the browser. A user with uBlock Origin, Brave Shields, or any other blocker still sends the full URL to your server. The server reads it. Done.

Second, it bypasses ITP’s cookie restrictions. Server-side session storage uses HTTP-only cookies that ITP doesn’t cap at 7 days. The attribution chain holds across multi-session purchase journeys.

Third, it’s immune to SPA URL rewrites. The server captures the parameter on the initial request. Client-side routing changes after that point don’t affect what the server already stored.

Transmute Engine captures URL parameters including adview_query_id server-side at the point of the incoming request, persisting the data through to the WooCommerce order. Because the capture happens at the server layer, the attribution survives ad blockers, ITP, and client-side URL manipulation that would strip the parameter before a JavaScript tag could read it. Stores running Transmute Engine are already accumulating AI Mode session data that will map directly to Google’s reporting column when it ships.

What the Data Tells Us About AI Mode Commerce

The click behaviour shift measured by Pew Research and the ad expansion data from Optimyzee point to a commerce channel that’s being built in plain sight.

Pew Research Center’s analysis of 68,879 Google searches found that users click traditional search results 8% of the time when an AI summary appears, versus 15% without one — nearly halving the click-through rate (Pew Research, 2025). Users ended their browsing session entirely after 26% of AI summary pages, compared with 16% without.

Translation: AI Mode is absorbing the research phase. When a user does click through from an AI summary, that click carries higher intent. The session has already been pre-qualified by the AI response. The user isn’t browsing — they’re acting.

The advertising data reinforces this. Ads in AI Mode results grew from 5.17% coverage to 25.5% year-over-year — a 394% increase (Optimyzee, 2026). Google Search revenue reached $60.4 billion in Q1 2026, up 19% year-over-year, with AI Mode cited as a contributor to incremental query growth.

For WooCommerce stores, the convergence of these signals points in one direction: AI Mode is becoming a meaningful acquisition channel. The stores capturing adview_query_id now will be the ones that can measure its contribution when Google surfaces the data officially. The stores that aren’t will be left reconstructing the picture from incomplete GA4 channel data, months behind.

Pew Research found users click traditional results 8% of the time with AI summaries versus 15% without — meaning the clicks that do come through carry substantially higher intent, and adview_query_id is Google’s mechanism to attribute those high-value sessions to the AI Mode origin (Pew Research, 2025).

Key Takeaways

• adview_query_id is live now: Discovered Labs confirmed Google is already stamping AI Mode click-throughs with this parameter across first-party conversion, third-party conversion, and GTM/analytics layers.
• 1 billion+ monthly users, no reporting column: AI Mode has massive scale, but adview_query_id doesn’t appear in GA4 or Google Ads reports yet — stores must capture it independently.
• Client-side capture fails for 30%+ of visitors: Ad blockers, Safari ITP, and SPA URL rewrites strip the parameter before JavaScript can read it — server-side is the only reliable path.
• Early capture builds irreplaceable historical data: When Google ships AI Mode as a standard reporting dimension, stores with server-side capture will have months of cohort data ready for comparison.
• High-intent channel signal: AI Mode clicks carry higher purchase intent (Pew’s 8% vs 15% click-rate data) and planning queries are growing 80% faster — this is commerce traffic worth measuring.