How Google's New AI Layout Impacts Your Brand's Visibility

What did Google change in its AI search layout?

On July 1, 2026, SEO analyst Brodie Clark identified something worth paying close attention to: Google was testing a version of its AI search layout in which citation counts and website favicons were moved from the top of the AI answer to below it. Search Engine Roundtable reported the discovery the same day.

The shift sounds incremental. It is not. Previously, users arriving at an AI-powered search result would see source labels, favicons, and citation counts alongside or above the AI-generated summary. Your brand had a visual presence from the first moment a user's eye landed on the page. In the new layout, the AI answer loads and holds attention first. Sources appear only after users have already read the synthesized response.

The practical consequence is direct: if a user is satisfied by the AI answer, they are less likely to scroll to where your brand now sits. The window for brand recognition collapses. What used to be table stakes, having your logo visible before users read a single word of the AI text, has become a harder target to hit.

Google described this as a test, not a permanent rollout. But the direction is consistent with a broader pattern inside Google's product: AI answers are increasingly designed to be the destination, not a waypoint to somewhere else. Every layout iteration that de-emphasizes sources accelerates that trajectory, regardless of whether this specific test becomes the default.

What does the open web click rate tell us about where this is heading?

The layout change did not happen in isolation. In 2026, SparkToro and Similarweb published a study showing that only 27.6% of clicks from Google searches now reach the open web. In 2024, 60.45% of US searches ended with zero clicks. By 2026, that figure had climbed to 68.01%. Two out of every three searches end before a user visits any external site.

When AI Overviews appear in results, the effect sharpens further. Click-through rates on the underlying organic results drop by nearly 60% compared to pages where no AI Overview is present. That is not a rounding error in your analytics. That is a structural change in how search allocates attention.

The remaining roughly 32% of searches that do generate a click are increasingly split between Google's own properties, including AI Mode, YouTube, Maps, and Images, paid ads, and the open web. The open web's share of that click pool is contracting, and the layout change that pushes source branding below the AI answer compounds the contraction.

Put these numbers together and the picture is clear: traffic from Google is becoming a premium, not a commodity. Brands that depend on volume-driven organic traffic without a citation strategy are operating a model that Google's product roadmap is actively compressing.

Who is most exposed to this layout change?

Every brand that competes through search is affected. But the exposure concentrates in two places.

First: B2B SaaS and technology companies competing on category-level queries. Queries like 'best project management software' or 'top CRM for startups' now return AI answers that synthesize a recommendation from multiple sources. If your brand is not one of those cited sources, you do not appear at all, regardless of your organic ranking. The user reads the AI answer, gets a recommendation, and acts on it. A page-one ranking that is not cited inside the answer contributes nothing to that decision.

Second: local and service businesses competing on geo-specific intent. Queries like 'best accountant in Austin' or 'HVAC repair near me' now frequently trigger AI-generated responses that name specific businesses. The businesses named are not always those with the most reviews or the highest domain authority. They are the ones whose content gives the AI model enough signal to cite them with confidence.

The common thread is this: ranking used to be the primary visibility signal. Ranking still matters as the entry condition for being considered as a citation source, but it is not the finish line. The finish line is being cited inside the answer, with your brand visible in the part of the layout the user actually reads.

What does this mean for how you should think about visibility?

There is no page two in an AI answer. That line used to be a provocation. After this layout change, it is a technical description of how Google now presents information to most users on most queries.

The layout change forces a reframe of the visibility question. Brands have spent years asking 'where do we rank?' The correct question now is 'where do we get cited, and in which AI-generated answers does our brand appear?' Those are different questions with different answers, different tactics, and different metrics.

Citation Share, the percentage of relevant AI answers in a given category that cite your brand, is the metric that captures this shift precisely. A brand can hold position three for twenty keywords and have a citation share of zero if the AI answer for those queries synthesizes information from other sources. Conversely, a brand with modest organic rankings can carry meaningful citation share if its content is the kind of authoritative, specific, citable material that AI models prefer to quote.

The July 2026 layout change makes citation share more consequential, not less. If your favicon and citation count are already below the fold, being cited inside the answer text is the only reliable way to achieve brand visibility in the AI search experience. That is where user attention lives now. That is where brand recognition either happens or does not.

How should you respond to the new AI search layout?

The response is not to wait for the test to be reversed. Whether this specific layout change becomes permanent or not, the direction it represents is durable. Here is what actually moves citation share.

Produce content that answers questions directly and completely. AI models cite sources that give them clean, authoritative responses. Content structured around specific questions, with the answer in the opening sentence rather than buried in paragraph four, performs better as citation material. Every article you publish should be able to answer one question: would an AI model quote this paragraph to support a claim?

Generate original data. A proprietary statistic, a survey result, or a before-and-after case study gives an AI model something to cite that it cannot synthesize from generic content. Original data is the strongest citation signal available. If you are the source of a number, AI systems have to cite you to use it.

Build entity clarity. AI models operate on entity graphs, not just keyword matches. Your brand needs to exist as a coherent, well-described entity across your website, your structured data, and the broader web. Named entities with clear descriptions, consistent contact and location data for local businesses, and clean schema markup give models the confidence to cite you by name.

Track what is actually happening. Citation Share is not visible in Google Search Console. You need to query AI engines directly, across ChatGPT, Perplexity, Gemini, and AI Overviews, on the questions that matter in your category, and measure whether you appear. If you do not know your current citation share, you cannot know whether anything you are publishing is working.

Sources

Google tested moving citation counts and favicons below the AI answer in July 2026, as identified by Brodie Clark and reported by Search Engine Roundtable Rosh Media, 2026-07-01

68.01% of US searches ended with zero clicks in 2026, up from 60.45% in 2024, and only 27.6% of clicks reach the open web Rosh Media, citing SparkToro and Similarweb 2026, 2026-07-01

When AI Overviews appear, click-through rates on underlying organic results drop by nearly 60% Rosh Media, citing SparkToro and Similarweb 2026, 2026-07-01