How to Track Preferred Sources Conversions in GA4
Google rolled out Preferred Sources into AI Mode and AI Overviews on May 27, 2026. Publishers can add a deeplink button to invite readers to designate them as a Preferred Source. The problem: Google has not added a Search Console report for Preferred Source counts or click-through impact as of May 2026. Publishers who want to measure whether the button is converting have to build their own tracking.
Brass-SEO's recommended setup uses GA4 across three measurement layers: direct button click events, indirect CTR shifts on Google referrer traffic, and a 90-day before/after window comparing organic search performance. This post walks through each layer with the specific GA4 configuration, the analytics events, and the realistic measurement timeline.
The companion post is Google Preferred Sources: How to Add the Button. Read that first if you haven't added the button yet.
Quick Navigation
- Why GA4 Is the Only Game in Town
- What to Measure: Three Layers
- Layer 1: Outbound Click Tracking on the Deeplink
- Layer 2: Indirect CTR Tracking via GSC Referrer Data
- Layer 3: The 90-Day Before/After Window
- Building the GA4 Exploration Report
- Reading the Data Honestly
- Frequently Asked Questions
Why GA4 Is the Only Game in Town
Google Search Console has no Preferred Sources report as of May 2026. The Preferred Sources personalization is user-account-level data that Google doesn't expose to publishers through Search Console. Brass-SEO checked and confirmed: the Performance, Search Appearance, and Discover reports all lack any Preferred Source dimension. There's no count of how many users have added you, no breakdown of clicks by Preferred badge vs. unbadged, and no impression segmentation.
The published timeline from Google's Search Central team has not committed to a Preferred Sources reporting feature in Search Console. Publishers asking when one might ship have received responses that translate to "no timeline." This means the only measurement options in 2026 are indirect ones built in GA4 plus third-party traffic logs.
The good news: the three-layer approach below catches most of what matters. The bad news: none of it provides Preferred-Source-level attribution. You'll see directional signals, not exact attribution.
For the broader question of why AI surface tracking is hard in 2026, see Tracking AI Mode Traffic in GA4: A Setup Guide.
What to Measure: Three Layers
Brass-SEO's three measurement layers, in order of directness:
| Layer | What it measures | Setup time | Confidence |
|---|---|---|---|
| 1 | Button click events (people who click your deeplink) | 10 min | High — direct measurement |
| 2 | CTR shift on Google referrer traffic over time | 15 min | Medium — indirect inference |
| 3 | 90-day before/after on organic clicks for AI-Mode-triggered queries | Ongoing | Medium-low — confounded by many variables |
Layer 1 catches the audience growth side: how many readers are actually clicking through to add you. Layer 2 catches the visibility-effect side: whether labeled citations produce measurably higher CTR than unlabeled. Layer 3 catches the realistic business-impact side: whether the cumulative work is producing real traffic shifts.
Each layer answers a different question. Run all three for the cleanest picture.
Layer 1: Outbound Click Tracking on the Deeplink
The simplest layer. GA4 can capture clicks on any outbound link with an event handler. The setup catches every click on your "Add us as a Preferred Source" button across every page where it appears.
The GA4 configuration:
- Go to Admin → Custom definitions → Create custom event
- Event name:
preferred_source_click - Source event:
click - Matching condition:
link_urlcontainsgoogle.com/preferences/source - Save
GA4 automatically tracks outbound link clicks when Enhanced Measurement is enabled (the default on most setups). The custom event extracts only the Preferred Sources deeplink clicks from the broader outbound-click stream.
For richer data, add custom parameters by configuring the button HTML with data- attributes that GA4 picks up:
<a href="https://google.com/preferences/source?q=yourdomain.com"
target="_blank"
rel="noopener"
data-button-location="post-footer"
data-button-style="official-badge">
Add us as a Preferred Source on Google
</a>
Then register data-button-location and data-button-style as custom dimensions in GA4. The result: you can see clicks broken down by where on the page the button sits and which design variant earned the click.
For the broader pattern of tracking custom events in GA4, see Where Is My Traffic Coming From? A GA4 Guide.
Layer 2: Indirect CTR Tracking via GSC Referrer Data
The harder but more strategically important layer. Google's published claim is that Preferred-badged citations get about 2x the CTR of unbadged citations in the same response. Layer 2 tries to verify this on your own data over time.
The setup:
- In Google Search Console, pull Performance → Search Results for the last 28 days
- Filter Pages by your top 20 highest-impression pages
- Export CTR per page
- Repeat the same pull every 30 days for 6 months
- Compare the trend per page against the date you added the Preferred Sources button
The signal you're looking for: pages with stable or rising impressions but rising CTR over the 6-month window. Pages where impressions are flat and CTR climbs are likely benefiting from accumulated Preferred Source selections among your reader base.
The confounding factors are real. Google's algorithm changes, seasonal traffic patterns, content updates, and SERP feature changes all affect CTR over a six-month window. Treat Layer 2 as directional, not authoritative. The pattern you want is "more pages showing positive CTR trends than negative" — not perfect attribution to Preferred Sources alone.
For more on reading CTR shifts from GSC data, see Title Tags and CTR: What the Public Data Shows and How to Read GSC's Search Appearance Report.
Layer 3: The 90-Day Before/After Window
The longest-cycle layer measures business-level impact. The setup:
- Note the date you add the Preferred Sources button as Day Zero
- Pull GSC Performance for the 90 days before Day Zero
- Note total organic clicks, average CTR, and average position for that period
- Pull GSC Performance for the 90 days after Day Zero (so Day 91 through Day 180)
- Compare
Brass-SEO's expectation for a publisher with a real audience (10,000+ monthly visitors): a 5-15% organic clicks lift over the 90-day window post-button-add, attributable in part to Preferred Source effects. The range is wide because the effect compounds slowly as more users add you and as the AI-mediation share of search increases over the same window.
Publishers with smaller audiences see proportionally smaller absolute lifts and noisier signals. Publishers in industries where AI Mode triggers less often (heavy transactional, navigational, or local) see smaller lifts because the surfaces where the badge shows up are less common.
Layer 3 is the layer most likely to be misread. A 12% lift in organic clicks could come from Preferred Sources, from Google algorithm tailwinds, from your content publishing cadence, or from external factors. The honest framing: this layer tells you if the broader investment is paying off, not whether the Preferred Sources button specifically caused it.
Building the GA4 Exploration Report
A single GA4 exploration combines the three layers into one weekly check-in dashboard. The configuration:
- Technique: Free-form
- Dimensions: Page Location, Event Name, Custom Event Parameter (
data-button-location) - Metrics: Event count, Users, Sessions
- Filter: Event Name =
preferred_source_click - Date range: Compare last 28 days to prior 28 days
- Segment: All Users
The result is a table that shows you weekly clicks on the Preferred Sources button by page location, with a 28-day comparison. Save the exploration. Check it once per week.
The supplementary exploration: same configuration but with Event Name filtered to standard outbound clicks (not just Preferred Sources). This gives you the relative share — what percentage of all outbound clicks from your site are going to the Preferred Sources deeplink. If the share is climbing, the button is competing well against your other CTAs.
For more on building useful GA4 explorations, see The Only 5 GA4 Numbers That Matter for Your Business.
Reading the Data Honestly
Brass-SEO sees four reading patterns most publishers will encounter, with the recommended response for each.
Layer 1 climbs, Layer 2 flat, Layer 3 flat. People are clicking but it's not converting into measurable CTR or business effects. Possible cause: not enough total clickers to move the broader signals yet. Wait 60 more days, recheck.
Layer 1 flat, Layer 2 climbs, Layer 3 climbs. People aren't clicking the button much, but pages are getting Preferred citations anyway (likely through users adding you via the Google Search settings directly). The button is underperforming as a CTA; the underlying audience is converting through other paths. Test button placement or copy.
Layers 1 and 2 climb, Layer 3 flat. The mechanism is working but the absolute traffic effect is below the noise floor. Either your audience is too small for the effect to be measurable yet, or AI Mode triggering is rare in your topic area. Continue.
All three climb. Best case. The button is converting, the underlying signal is producing CTR lift, and the business effect is visible. Recommend doubling down on the original-reporting content that's pulling Preferred Source citations into your pages.
The dishonest reading pattern Brass-SEO sees most often: declaring the button "works" or "doesn't work" after 30 days. The 90-day window exists because shorter windows confound with too many other factors. Set the calendar reminder for Day 91 and don't draw conclusions earlier.
To audit your own pages for the citability signals that AI surfaces use upstream of Preferred Source visibility, start a Brass-SEO trial.
Frequently Asked Questions
Will Search Console eventually add a Preferred Sources report?
Possibly. Google has not announced one as of May 2026. The cadence of Search Console feature additions for new search surfaces (AI Overviews took roughly 8 months to get even partial Search Appearance reporting) suggests a Preferred Sources report could ship in 2027 if Google decides the feature is durable enough to warrant publisher tooling. In the meantime, the GA4 setup is the only practical option.
Can I use UTM parameters on the deeplink to track which page sent the click?
Indirectly. Google's preferences URL doesn't pass UTM parameters through to your analytics — the click goes outbound to google.com and the user never returns to your site through a UTM-bearing referrer. The way to track per-page button performance is the data-button-location attribute approach in Layer 1, which captures the source page at click time in your GA4 events.
Are third-party tools better at measuring this than GA4?
As of May 2026, no specialized commercial tool has launched for Preferred Sources tracking. Most third-party SEO platforms rely on the same GSC data you have access to directly. The GA4 setup above is the practical state of the art.
What's the minimum traffic needed to see meaningful signals?
For Layer 1 (direct button clicks), any traffic level works since you're measuring conversion rate on outbound clicks. For Layer 2 (CTR shifts), Brass-SEO recommends at least 1,000 monthly impressions on the affected pages. For Layer 3 (organic clicks change), at least 5,000 monthly organic clicks before the trends rise above noise.
How do I know if Preferred Sources is contributing to traffic vs. other AI surface changes?
You don't, cleanly. The mature framing is: Preferred Sources is one input to a broader AI search visibility strategy. Measure the broader strategy (manual quarterly visibility tests, citation pattern analysis, audience growth) and assume Preferred Sources is contributing in proportion to the button's click-through volume. For the quarterly testing pattern, see How to Test Your Brand's AI Search Visibility.