Google’s Testing “Sticky Citations” in AI Overviews — Here’s How to Turn Visibility into Clicks

Google is experimenting with sticky citations in AI Overviews (AO) on desktop. As you scroll the AO answer, the first citation stays pinned at the top, keeping that source constantly in view. If you already win an AO cite, this test is a net positive. If you don’t, it just raised the bar.

credit: Search Engine Land

What changed (and where)

• Sticky, not static: Instead of the citations disappearing as the user scrolls the AO, the first source now sticks to the top of the AO module. Early footage shows it in desktop tests.
  
• AO-only behaviour: The sticky effect applies within the AO panel as you scroll its content (not the entire SERP). Barry Schwartz replicated and posted demos.
  
• Context matters: Just last week Google also tested moving citation cards to the bottom of the AO, which predictably hid sources below the fold. Today’s sticky variant does the opposite: it elevates the first cite. Expect Google to keep iterating on layout.
  

Bottom line: If you’re the first AO citation, your brand enjoys more on-screen time. If you’re second or third, you’ve got less visual real estate than ever.

Why this matters (traffic reality check)

Let’s cut the noise. AO visibility ≠ clicks. Field studies show AO citations underperform blue links for click-through, even with big visual prominence. In other words, users read the summary and often don’t click. That’s been observed across tens of thousands of queries and in independent user-behaviour research. 

Sticky citations could shift two things:

1. Brand recall: The pinned source gets persistent exposure. That’s free, repeated brand imprinting during the read.
   
2. Click propensity (maybe): If the pinned source clearly promises extra value (tools, downloads, calculators, richer visuals), you can earn more of the scarce clicks. If your page looks like a me-too summary, users will keep scrolling.
   

How Google seems to choose AO citations (practical view)

You don’t win the pin by accident. Across tests and documentation, the same patterns keep showing up:

• Helpful, people-first content with clear authorship and accountability. Google literally tells you to show bylines, be transparent about who wrote it, and demonstrate topical experience.
  
• Search Essentials basics are non-negotiable: crawlability, indexability, and a page that genuinely solves the task.
  

No, E-E-A-T isn’t a single “ranking factor”, but in AO land, credible, well-packaged answers get picked and pinned. Act accordingly. 

What to do today (if you already win AO citations)

This is where most teams leave money on the table. If you’re seeing AO impressions but thin clicks, fix these:

1) Tighten your E-E-A-T footprint on the winning pages

• Add/refresh bylines with credentials and a short “Why you can trust me” box. Link to an author page that shows real-world experience (roles, certifications, client work).
  
• Strengthen your About and Contact pages; expose an editorial policy and references where relevant. (Trust signals matter more when a summary borrows your answer.)
  

2) Win the click with visible extras

Users only click when there’s obvious value beyond the snippet. Make that value blindingly obvious above the fold:

• Jump links / table of contents mapping the exact subsections AO touches (so users land on the bit they want).
  
• Embedded video (preferably YouTube) for tasks and comparisons—LLMs favour video evidence and users do too.
  
• Downloads that AO will never deliver: checklists, calculators, templates, sample files.
  
• Proof modules: data tables, cited studies, “worked example” boxes.
  

Studies consistently show AO citations don’t naturally drive clicks; you must engineer the click with differentiated value. 

3) Sharpen page structure for machine & human

• Clear H1 that states the outcome (not just the topic).
  
• Lead with the answer in the intro; AO often mirrors your first 2–3 sentences.
  
• Add FAQ sections that clarify edge cases and “it depends” scenarios (great for skimming and anchoring).
  
• Use Article schema with authorship and sameAs where appropriate. (Don’t obsess over deprecated rich result types.)
  

If you’re not the first citation (how to displace the pinned source)

1. Diagnose the gap
   Compare your page vs the pinned competitor on four axes: evidence density, task depth, entity clarity (naming, definitions, variants), and UX clarity (jump links, summary, scannability). The one scoring higher on all four usually wins the cite—and now, the pin.
   
2. Beat them where AO looks first
   

• Make your opening summary more explicit and correctly scoped (define the scenario, constraints, and the “it depends” boundary).
  
• Add fresh primary evidence (original screenshots, benchmarks, short clips).
  
• Tighten entity signals: synonyms, model numbers, regional terms, and canonical definitions that reduce ambiguity.
  

3. Publish the better artefact
   If they have a guide, ship a guide plus a calculator or decision tree. Sticky citations reward the result that adds tangible utility.
   

Measurement & diagnostics

• Check presence: Manually sample priority queries where you rank and where AO appears. Note whether you’re (a) cited and (b) first in the citation rail. Footage and posts confirm the sticky test is live for some users on desktop.
  
• Track impact (directionally):
  
  • GSC: Watch page-level impressions and CTR for AO-winning URLs. If impressions climb with flat CTR, you’re getting “seen but not chosen”.
    
  • On-page analytics: Scroll-depth on AO-landing pages and interactions with your value adds (downloads, calculators).
    
  • Brand lift: Monitor branded search volume for the pinned page’s brand; sticky exposure can move recall even without the click.
    
• A/B the intro: Small changes to the first 50 words can flip whether AO mirrors your copy—and whether users feel they still need to click.
  

Product context: Google keeps moving the furniture

Google’s experimenting aggressively with AO layouts. We’ve seen right-side cards, bottom-pinned cards, and now sticky first citations as you scroll. Expect more UI trials aimed at balancing usefulness with publisher visibility (and less rage from newsrooms). Monitor your critical queries monthly; don’t assume today’s layout persists. 

Implementation checklist (use this today)

• Identify your top 20 AO-exposed keywords (where an AO appears and you rank 1–10).
  
• Verify which of those pages are cited—and whether you’re the first citation.
  
• Harden E-E-A-T on those pages: bylines, author bios, references, policy links.
  
• Add value above the fold: jump links, hero summary, tools, video, downloads.
  
• Re-package the intro for clarity (answer first, caveats second).
  
• Ship one differentiator per page (calculator, template, comparison matrix).
  
• Measure: CTR, scroll, interaction with value blocks; iterate weekly.
  

FAQs

Is sticky citations a full rollout?
No—a test seen in the wild. It’s on desktop AO modules and may not be visible to everyone. 

Will this guarantee more clicks?
No. AO citations often don’t earn clicks by default. You need visible value the AO summary can’t provide. 

What should I prioritise first?
Pages that already win AO citations (especially for money terms). Strengthen authorship, add jump links + downloads, and make the intro decisive. 

Final word

Sticky citations raise the stakes. If you’re the first source, you’ve just won more time in front of the user’s eyes. If you’re not, your content has to be demonstrably better—clearer, more useful, and more credible—so Google has a reason to swap you into that top slot.

Do the boring work: authors, evidence, structure, utility. That’s how you turn AO visibility into actual traffic—and revenue.

Sources: hands-on reporting of the sticky test and prior AO layout experiments; behavioural data showing weak AO link CTR; and Google’s own guidance on helpful, people-first content and authorship.