{
    "id": 916,
    "date": "2026-08-14T12:00:00",
    "date_gmt": "2026-08-14T05:00:00",
    "guid": {
        "rendered": "https:\/\/brandgineer.co\/?p=916"
    },
    "modified": "2026-08-20T23:21:24",
    "modified_gmt": "2026-08-20T16:21:24",
    "slug": "what-is-google-ai-overview",
    "status": "publish",
    "type": "post",
    "link": "https:\/\/brandgineer.co\/en\/what-is-google-ai-overview\/",
    "title": {
        "rendered": "What Is Google AI Overview and How to Optimize for It"
    },
    "content": {
        "rendered": "<p class=\"wp-block-paragraph\"><strong>Google AI Overview is an AI-generated summary at the top of the search results page that synthesises information from an average of around 13 sources and links to them.<\/strong> More important than the definition: how AI Overview selects those sources changed fundamentally over the past twelve months, and most advice currently circulating is based on the old mechanism.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>Key Takeaways<\/strong><\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li>In July 2025, 76% of AI Overview citations came from pages ranking in the organic top 10. By March 2026 that figure had fallen to 38%, according to an Ahrefs study of 863,000 keywords and 4 million URLs. Organic ranking is no longer the gate.<sup data-fn=\"883fec62-8e45-441e-b4c4-9b37c4bd962f\" class=\"fn\"><a id=\"883fec62-8e45-441e-b4c4-9b37c4bd962f-link\" href=\"#883fec62-8e45-441e-b4c4-9b37c4bd962f\">1<\/a><\/sup><\/li>\n\n\n\n<li>Google decomposes a query into sub-questions and finds the best source for each. A page ranking first can still lose a citation slot to a page that answers a sub-question better.<sup data-fn=\"552b32d2-095e-4086-ba51-ea40d337961c\" class=\"fn\"><a id=\"552b32d2-095e-4086-ba51-ea40d337961c-link\" href=\"#552b32d2-095e-4086-ba51-ea40d337961c\">2<\/a><\/sup><\/li>\n\n\n\n<li>AI Overview synthesises new content from multiple sources, unlike featured snippets which quote verbatim from a single page.<\/li>\n\n\n\n<li>Google states AI Overviews appear on roughly 50% of US queries, up from 6.49% in early 2025.<sup data-fn=\"20f866e9-edc7-4728-a047-0e1580bae1db\" class=\"fn\"><a id=\"20f866e9-edc7-4728-a047-0e1580bae1db-link\" href=\"#20f866e9-edc7-4728-a047-0e1580bae1db\">3<\/a><\/sup><\/li>\n\n\n\n<li>Top-10 overlap varies sharply by industry: healthcare, insurance, and education sit at 68-75%, while e-commerce stayed essentially flat. BrightEdge&#8217;s analysis puts overlap even lower, around 17%.<sup data-fn=\"f8883265-1e1f-46e9-8975-c011691e60d5\" class=\"fn\"><a id=\"f8883265-1e1f-46e9-8975-c011691e60d5-link\" href=\"#f8883265-1e1f-46e9-8975-c011691e60d5\">4<\/a><\/sup><\/li>\n\n\n\n<li>For Vietnamese-language queries AI Overview favours Vietnamese sources. ChatGPT does the opposite: 43% of its background queries run in English even when the original prompt was not. Vietnamese is an advantage on AIO but not sufficient for ChatGPT.<sup data-fn=\"dbd24f0f-f8d2-4d1b-92c6-9047cfb14499\" class=\"fn\"><a id=\"dbd24f0f-f8d2-4d1b-92c6-9047cfb14499-link\" href=\"#dbd24f0f-f8d2-4d1b-92c6-9047cfb14499\">5<\/a><\/sup><sup>,<\/sup><sup data-fn=\"59aca1c6-6097-4288-88af-770ab9af8c7c\" class=\"fn\"><a id=\"59aca1c6-6097-4288-88af-770ab9af8c7c-link\" href=\"#59aca1c6-6097-4288-88af-770ab9af8c7c\">6<\/a><\/sup><\/li>\n<\/ul>\n\n\n\n<h2 id=\"what-google-ai-overview-is\" class=\"wp-block-heading\">What Google AI Overview Is<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>Google AI Overview is an AI-generated summary that appears above traditional search results, synthesising information from multiple sources in Google&#8217;s index and linking to those sources.<\/strong><\/p>\n\n\n\n<p class=\"wp-block-paragraph\">From the user&#8217;s perspective, they get a direct answer instead of a list of links to read through. Technically, Google runs a language model over a set of sources retrieved from the index, generates a synthesised answer, and attaches citations to the sources used.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Google states AI Overviews appear on roughly 50% of US queries. Independent studies give a wider range, from 21% under conservative methodology to 65% under broader sampling, depending on how the keyword set is chosen. The exact number matters less than the direction: in early 2025 the figure was 6.49%, meaning nearly eightfold growth in about fifteen months.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Prevalence varies sharply by query type. Health and medical queries show AI Overviews around 71% of the time. Transactional and purchase queries only around 19%. In other words, AI Overview dominates the research and consideration stage and cedes ground to traditional results at the point of purchase.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">For brands, this places AI Overview squarely in the consideration stage: the user is still learning, has not decided, and the answer they read will shape their shortlist.<\/p>\n\n\n\n<hr class=\"wp-block-separator has-alpha-channel-opacity\"\/>\n\n\n\n<h2 id=\"how-ai-overview-differs-from-featured-snippets-and-ai-mode\" class=\"wp-block-heading\">How AI Overview Differs from Featured Snippets and AI Mode<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>These three formats are often conflated but operate on three different mechanisms, and confusing them leads to the wrong optimisation strategy.<\/strong><\/p>\n\n\n\n<figure class=\"wp-block-table\"><table class=\"has-fixed-layout\"><thead><tr><th><\/th><th><strong>Featured Snippet<\/strong><\/th><th><strong>AI Overview<\/strong><\/th><th><strong>AI Mode<\/strong><\/th><\/tr><\/thead><tbody><tr><td><strong>How content is created<\/strong><\/td><td>Verbatim quote<\/td><td>Synthesised, new text<\/td><td>Synthesised, conversational<\/td><\/tr><tr><td><strong>Number of sources<\/strong><\/td><td>1<\/td><td>~13 on average<\/td><td>Many, varies per turn<\/td><\/tr><tr><td><strong>Query type<\/strong><\/td><td>Short, well-defined<\/td><td>Research queries<\/td><td>Long, conversational<\/td><\/tr><tr><td><strong>Main condition<\/strong><\/td><td>Be the best page for the question<\/td><td>Answer one sub-question well<\/td><td>Match intent at a granular level<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>Featured snippets<\/strong> pull a verbatim passage from a single page in Google&#8217;s index. The game here is winner-takes-all: either you are the chosen page or you are not.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>AI Overview<\/strong> synthesises a new answer from an average of around 13 sources. The game changes entirely: you do not need to be the best source for the whole question, only the best source for a part of it. This is good news for smaller brands, because it turns a winner-takes-all contest into one with multiple slots.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>AI Mode<\/strong> handles long, conversational queries and selects content at a much more granular, intent-specific level than traditional search.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">There is a notable relationship between featured snippets and AI Overview: pages previously selected as featured snippets correlate strongly with pages cited in AI Overviews. That makes sense, since both reward the same thing, a direct, compact, self-contained answer.<\/p>\n\n\n\n<hr class=\"wp-block-separator has-alpha-channel-opacity\"\/>\n\n\n\n<h2 id=\"how-source-selection-changed\" class=\"wp-block-heading\">How Source Selection Changed<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>This is the most important section of the guide, and the one where most published advice is now out of date.<\/strong><\/p>\n\n\n\n<p class=\"wp-block-paragraph\">The most common advice you will read is that to appear in AI Overview you must first rank in the organic top 10. That used to be true. It no longer is.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>The data on the shift<\/strong><\/p>\n\n\n\n<p class=\"wp-block-paragraph\">In July 2025, 76% of AI Overview citations came from pages ranking in the organic top 10 for the same query. By March 2026, an Ahrefs study of 863,000 keywords and 4 million AI Overview URLs found that figure had fallen to 38%.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">In other words, <strong>62% of citations now come from pages outside the top 10<\/strong>. BrightEdge&#8217;s analysis puts the overlap even lower, around 17%. Most of the citation growth is coming from pages ranking between positions 21 and 100.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>Why this is happening<\/strong><\/p>\n\n\n\n<p class=\"wp-block-paragraph\">The mechanism lies in how Google decomposes queries. Rather than finding the single best page for the whole question, the system splits the question into sub-questions and finds the best source for each. A page ranking first for the primary query can lose a citation slot to a page that answers a specific sub-question better, even when that URL does not rank for the original query at all.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>The strategic implication, and this is the part worth acting on<\/strong><\/p>\n\n\n\n<p class=\"wp-block-paragraph\">If you are a smaller brand in Vietnam competing against long-established sites with high domain authority, this change works in your favour.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Previously, appearing in AI Overview meant winning the ranking battle, which meant overcoming years of accumulated backlinks and authority. Now you can get in by answering one specific sub-question better than everything currently ranking. That is a contest about content quality and specificity, not domain age.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Concretely: instead of writing a 3,000-word overview hoping to rank for the head keyword, write content that fully answers the sub-questions nobody is answering properly. That is where the real gap sits.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>One important caveat<\/strong><\/p>\n\n\n\n<p class=\"wp-block-paragraph\">This shift is not uniform across industries. Trust-sensitive sectors retain high overlap: healthcare, insurance, and education sit around 68-75%. For those topics Google still strongly favours content that already ranks well organically, and you cannot skip traditional SEO.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">E-commerce is an exception in the other direction: overlap stayed essentially flat and AI Overview coverage actually decreased.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Before applying any strategy, check which group your industry falls into by running your own query set.<\/p>\n\n\n\n<hr class=\"wp-block-separator has-alpha-channel-opacity\"\/>\n\n\n\n<h2 id=\"five-factors-that-determine-citation\" class=\"wp-block-heading\">Five Factors That Determine Citation<\/h2>\n\n\n\n<h3 id=\"1-answer-structure-aligned-to-sub-questions\" class=\"wp-block-heading\">1. Answer structure aligned to sub-questions<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">The highest-impact factor. Each section on the page should fully answer exactly one sub-question within the first 40 to 60 words. Detail, examples, and supporting data come after.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">How to test it: cover the rest of the page and read only a section&#8217;s opening. If it answers the implied question without context, it can be cited. If not, it will be skipped.<\/p>\n\n\n\n<h3 id=\"2-topical-coverage-depth\" class=\"wp-block-heading\">2. Topical coverage depth<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">Because Google sources each sub-question separately, a page covering many related sub-questions has more chances than a page answering only the head question. This is why comprehensive guides outperform short posts on the same topic.<\/p>\n\n\n\n<h3 id=\"3-organic-ranking-at-reduced-weight\" class=\"wp-block-heading\">3. Organic ranking, at reduced weight<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">Still valuable, particularly in trust-sensitive industries. But no longer a prerequisite. If you rank between positions 20 and 50 for a query with an AI Overview, you still have a genuine chance.<\/p>\n\n\n\n<h3 id=\"4-freshness-signals\" class=\"wp-block-heading\">4. Freshness signals<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">AI Overview changes sources far more often than organic results. Recently updated content with an accurate dateModified in schema has an advantage. This is where schema markup plays a genuinely practical role.<\/p>\n\n\n\n<h3 id=\"5-machine-readability\" class=\"wp-block-heading\">5. Machine readability<\/h3>\n\n\n\n<p class=\"wp-block-paragraph\">Content must be server-side rendered. If AI crawlers cannot read it, none of the other factors matter. Test by disabling JavaScript and reloading.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">\u2192 <a href=\"\/en\/schema-markup-for-aeo\/\" data-type=\"link\" data-id=\"\/schema-markup-for-aeo\/\"><strong>Read the schema implementation and server-side rendering verification guide<\/strong><\/a><\/p>\n\n\n\n<hr class=\"wp-block-separator has-alpha-channel-opacity\"\/>\n\n\n\n<h2 id=\"the-four-step-optimisation-process\" class=\"wp-block-heading\">The Four-Step Optimisation Process<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>Step 1: Identify which queries trigger AI Overview<\/strong><\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Run 20-30 target keywords on Google and record which trigger an AI Overview. Not every query does, and optimising for queries without one wastes effort.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Also record: how long the AI Overview is, how many sources it cites, and whether it appears consistently when you re-run the query a few days later. AI Overviews are less stable than organic results.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>Step 2: Analyse the sources being cited<\/strong><\/p>\n\n\n\n<p class=\"wp-block-paragraph\">For each query with an AI Overview, open every cited source and record four things:<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li>What content format it uses: guide, listicle, definition, comparison<\/li>\n\n\n\n<li>Where on the page the cited answer sits<\/li>\n\n\n\n<li>Whether that page ranks organically for this query<\/li>\n\n\n\n<li>Which sub-question that page is answering<\/li>\n<\/ul>\n\n\n\n<p class=\"wp-block-paragraph\">The last one matters most. Once you have listed the sub-questions Google is sourcing for, you have a map of the content you need to write.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>Step 3: Structure content around sub-queries<\/strong><\/p>\n\n\n\n<p class=\"wp-block-paragraph\">For each identified sub-question, create a dedicated section on your page opening with a direct answer in 40 to 60 words.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Prioritise the sub-questions where current sources answer poorly: vague answers, missing data, or outdated figures. That is where your replacement odds are highest.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>Step 4: Monitor and iterate monthly<\/strong><\/p>\n\n\n\n<p class=\"wp-block-paragraph\">AI Overview changes sources far more often than organic results. Re-run your query set monthly, recording new sources appearing and sources dropping out.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">What to watch for: if you appear and then disappear, it is usually not because your content got worse, but because another source answered that sub-question better. Compare their content against yours on that specific sub-question.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">\u2192 <a href=\"\/en\/measure-brand-mentions-ai-search-2026\/\" data-type=\"link\" data-id=\"\/measure-brand-mentions-ai-search-2026\/\"><strong>See how to set up a full AI visibility tracking system<\/strong><\/a><\/p>\n\n\n\n<hr class=\"wp-block-separator has-alpha-channel-opacity\"\/>\n\n\n\n<h2 id=\"ai-overview-and-vietnamese-language-queries\" class=\"wp-block-heading\">AI Overview and Vietnamese-Language Queries<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>For Vietnamese-language queries, Google AI Overview favours Vietnamese sources. But this holds for AI Overview only, not for ChatGPT, and telling those two cases apart determines how you allocate budget.<\/strong><\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Most Vietnamese-language content on this topic says one of two things, and both are imprecise: either &#8220;AI favours English so Vietnam brands are disadvantaged,&#8221; or &#8220;AI favours Vietnamese so just write in Vietnamese.&#8221; The reality is that the two engines work on opposite mechanisms.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>Google AI Overview: same-language sources are favoured<\/strong><\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Profound&#8217;s analysis of 3.25 billion citations found that in research on the Mexican market, 96% of AI Overview citations came from Spanish-language sources. Where a local-language option exists, English sources are pushed out of the top five positions.<sup data-fn=\"588c60f6-a159-454d-87e0-5652d6930c5f\" class=\"fn\"><a id=\"588c60f6-a159-454d-87e0-5652d6930c5f-link\" href=\"#588c60f6-a159-454d-87e0-5652d6930c5f\">7<\/a><\/sup> For AIO, quality well-structured Vietnamese content is in a favourable position.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>ChatGPT: the query is translated to English before sources are found<\/strong><\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Peec AI&#8217;s study of 10 million prompts and 20 million fan-out queries found 43% of background searches ran in English even when the original prompt was in another language, and 78% of non-English sessions included at least one English background query. A concrete example from the research: Polish users asking about local auction platforms got eBay back from ChatGPT&#8217;s English fan-out queries instead of Allegro, the dominant platform in Poland.<br><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>These two findings do not contradict each other<\/strong><\/p>\n\n\n\n<p class=\"wp-block-paragraph\">They measure two different stages of the same process. Peec measured <strong>retrieval<\/strong>, meaning which sources enter the candidate pool. Profound measured <strong>citation<\/strong>, meaning which sources within the pool get selected.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Local sources can be eliminated at stage one, before the language-preference logic at stage two ever applies. That is exactly what happened in the Allegro example: the AI did not choose eBay because it preferred English content, it chose eBay because Allegro never entered the pool to be considered.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>What this means in practice for Vietnam brands<\/strong><\/p>\n\n\n\n<figure class=\"wp-block-table\"><table class=\"has-fixed-layout\"><thead><tr><th><\/th><th><strong>Google AI Overview<\/strong><\/th><th><strong>ChatGPT<\/strong><\/th><\/tr><\/thead><tbody><tr><td>Well-structured Vietnamese content<\/td><td>Primary lever<\/td><td>Necessary but not sufficient<\/td><\/tr><tr><td>Presence on English-language sources<\/td><td>Less critical<\/td><td>Condition for entering the pool<\/td><\/tr><tr><td>What to do first<\/td><td>Sub-query structure, topical coverage<\/td><td>LinkedIn, Clutch, G2, international trade press<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n<p class=\"wp-block-paragraph\">Combined with the drop in top-10 overlap, the AI Overview opportunity is fairly specific: you do not need to win the domain authority battle, you do not need to compete with English-language content, you only need to answer sub-questions in Vietnamese better than what currently exists. For most specialist topics in Vietnam, the bar to clear is currently low.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">On ChatGPT the problem is different: you need your brand to appear on the English-language sources a fan-out query might reach, even if your customers do not read English.<\/p>\n\n\n\n<h2 id=\"best-practices-from-ai-visibility-measurement-platforms\" class=\"wp-block-heading\">Best Practices from AI Visibility Measurement Platforms<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>Four platforms specialising in AI visibility measurement have published their methodologies. These are the principles worth applying, even if you track manually.<\/strong><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>AI output is probabilistic, so one measurement is not one result.<\/strong> Evertune samples each prompt 100 times per model to reach statistical significance rather than running it once and drawing a conclusion.<sup data-fn=\"a8a6a2be-121c-4ad9-94ff-f169d8d0f9c5\" class=\"fn\"><a id=\"a8a6a2be-121c-4ad9-94ff-f169d8d0f9c5-link\" href=\"#a8a6a2be-121c-4ad9-94ff-f169d8d0f9c5\">8<\/a><\/sup> This is the most important methodological point and the most commonly ignored: the same prompt entered twice can produce two different answers. Run it once and you cannot tell whether you are looking at a trend or at noise.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Applying this manually: run each prompt at least 3 to 5 times in the same session, and record the proportion of runs in which your brand appeared rather than a yes or no.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>Distinguish what the AI already knows from what it just found.<\/strong> Evertune measures both: the baseline knowledge inside the model from training, and content retrieved in real time. These require different strategies. Baseline knowledge only shifts when the model is retrained, while retrieved content responds quickly to new publishing. For AI Overview the retrieval component carries more weight, which is where new content has the fastest effect.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>Use unaided prompts to measure genuine awareness.<\/strong> If you type &#8220;is Brandgineer any good,&#8221; the AI is forced to talk about Brandgineer and you learn nothing. A genuine awareness prompt looks like &#8220;which AEO agencies are good in Ho Chi Minh City,&#8221; so you can see whether the brand surfaces on its own.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>Model perceptions shift on a 3 to 6 month cycle.<\/strong> That is the appropriate window for evaluating an AEO programme, not 30 days. Combined with AI Overview changing sources more frequently than organic results, the right cadence is to measure monthly but draw conclusions quarterly.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>Always measure alongside competitors.<\/strong> All four platforms (Profound, Peec AI, Evertune, Scrunch) put competitive comparison at the centre rather than tracking one brand&#8217;s absolute numbers.<sup data-fn=\"6a503fb5-cf07-4a91-a7fb-3c7aa13829fa\" class=\"fn\"><a id=\"6a503fb5-cf07-4a91-a7fb-3c7aa13829fa-link\" href=\"#6a503fb5-cf07-4a91-a7fb-3c7aa13829fa\">9<\/a><\/sup> The reason is simple: a 5% citation rate may be excellent or terrible depending on where competitors sit.<\/p>\n\n\n\n<h2 id=\"frequently-asked-questions\" class=\"wp-block-heading\">Frequently Asked Questions<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>What is Google AI Overview?<\/strong> Google AI Overview is an AI-generated summary at the top of the search results page that synthesises information from multiple sources and links to them. Unlike a featured snippet, which quotes verbatim from a single page, AI Overview generates new text by synthesis and cites around 13 sources on average.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>How does AI Overview differ from a featured snippet?<\/strong> Featured snippets quote verbatim from one page. AI Overview synthesises new content from an average of around 13 sources. The practical consequence: with featured snippets you need to be the best page for the whole question, while with AI Overview you only need to answer part of it well.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>Do I need to rank in the organic top 10 to appear in AI Overview?<\/strong> Not anymore. In July 2025, 76% of citations came from top-10 pages. By March 2026 that had fallen to 38%, per an Ahrefs study of 863,000 keywords. Today 62% of citations come from pages outside the top 10, mostly from positions 21 to 100.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>What is a sub-query and why does it matter?<\/strong> Google decomposes a query into sub-questions and finds the best source for each. A page ranking first can still lose a citation slot to a page that answers a sub-question better, even when that page does not rank for the original query.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>How does optimising for AI Overview differ from optimising for ChatGPT?<\/strong> AI Overview is anchored to Google&#8217;s index and selects sources per sub-question, so the lever is content structure and topical coverage. ChatGPT leans more on training data and entity signals accumulated from external sources, so the lever is entity building and citations from authoritative sources.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>Is Vietnamese-language content at a disadvantage in AI Overview?<\/strong> No, on AI Overview Vietnamese is an advantage. Profound&#8217;s research on the Mexican market found 96% of AI Overview citations came from local-language sources. ChatGPT is different: Peec AI found 43% of ChatGPT&#8217;s background queries run in English even when the original prompt was in another language, so local sources can be eliminated at the retrieval stage.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>How many times should I run each prompt when checking AI Overview?<\/strong> At least 3 to 5 times in the same session. AI output is probabilistic and the same prompt can return different results. Evertune samples each prompt 100 times to reach statistical significance. Run it once and you cannot separate signal from noise.<\/p>\n\n\n\n<hr class=\"wp-block-separator has-alpha-channel-opacity\"\/>\n\n\n\n<h2 id=\"data-sources\" class=\"wp-block-heading\">Data Sources<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\"><em>Written by Abbie, AEO Expert at Brandgineer | brandgineer.co<\/em> <em>Optimised for AI visibility across ChatGPT, Google AI Overviews, and Perplexity.<\/em><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>Want to know which queries in your industry have AI Overviews and who is being cited?<\/strong> <a href=\"\/en\/contact\/\">Book a consultation with Brandgineer \u2192<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><strong>Read next:<\/strong> <a href=\"\/en\/increase-brand-visibility-ai-search-vietnam-2026\/\" data-type=\"link\" data-id=\"\/increase-brand-visibility-ai-search-vietnam-2026\/\">The 5-step guide to increasing AI visibility<\/a> | <a href=\"\/en\/schema-markup-for-aeo\/\">Schema Markup for AEO<\/a> | <a href=\"\/en\/measure-brand-mentions-ai-search-2026\/\">How to measure AI visibility<\/a><\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><\/p>\n\n\n<ol class=\"wp-block-footnotes\"><li id=\"883fec62-8e45-441e-b4c4-9b37c4bd962f\">Ahrefs Study: Only 38% of Google AI Overview Citations Come From Top 10 Pages (863,000 keywords, 4 million URLs). <a href=\"https:\/\/ahrefs.com\/blog\/ai-overview-citations-top-10\/\" target=\"_blank\" rel=\"noreferrer noopener\">https:\/\/ahrefs.com\/blog\/ai-overview-citations-top-10\/ <\/a> <a href=\"#883fec62-8e45-441e-b4c4-9b37c4bd962f-link\" aria-label=\"Jump to footnote reference 1\">\u21a9\ufe0e<\/a><\/li><li id=\"552b32d2-095e-4086-ba51-ea40d337961c\">How Google AI Mode and AI Overviews Select Sources. First Motion. <a href=\"https:\/\/firstmotion.com\/insights\/how-google-ai-mode-and-ai-overviews-select-sources\" target=\"_blank\" rel=\"noopener\">https:\/\/firstmotion.com\/insights\/how-google-ai-mode-and-ai-overviews-select-sources<\/a> <a href=\"#552b32d2-095e-4086-ba51-ea40d337961c-link\" aria-label=\"Jump to footnote reference 2\">\u21a9\ufe0e<\/a><\/li><li id=\"20f866e9-edc7-4728-a047-0e1580bae1db\">Google AI Overviews Statistics 2026. DemandSage. <a href=\"https:\/\/www.demandsage.com\/ai-overviews-statistics\/\" target=\"_blank\" rel=\"noopener\">https:\/\/www.demandsage.com\/ai-overviews-statistics\/<\/a> <a href=\"#20f866e9-edc7-4728-a047-0e1580bae1db-link\" aria-label=\"Jump to footnote reference 3\">\u21a9\ufe0e<\/a><\/li><li id=\"f8883265-1e1f-46e9-8975-c011691e60d5\">AI Overview Citations and Rank Overlap. BrightEdge. <a href=\"https:\/\/www.brightedge.com\/resources\/weekly-ai-search-insights\/rank-overlap-after-16-months-of-aio\" target=\"_blank\" rel=\"noopener\">https:\/\/www.brightedge.com\/resources\/weekly-ai-search-insights\/rank-overlap-after-16-months-of-aio<\/a> <a href=\"#f8883265-1e1f-46e9-8975-c011691e60d5-link\" aria-label=\"Jump to footnote reference 4\">\u21a9\ufe0e<\/a><\/li><li id=\"dbd24f0f-f8d2-4d1b-92c6-9047cfb14499\">How Query Language Reshapes AI Citations. Profound, 2026. <a href=\"https:\/\/www.tryprofound.com\/blog\/how-query-language-reshapes-ai-citations\" target=\"_blank\" rel=\"noopener\">https:\/\/www.tryprofound.com\/blog\/how-query-language-reshapes-ai-citations<\/a> <a href=\"#dbd24f0f-f8d2-4d1b-92c6-9047cfb14499-link\" aria-label=\"Jump to footnote reference 5\">\u21a9\ufe0e<\/a><\/li><li id=\"59aca1c6-6097-4288-88af-770ab9af8c7c\">ChatGPT Searches in English, Even When You Don&#8217;t. Peec AI, 2026 (10 million prompts, 20 million fan-out queries). <a href=\"https:\/\/peec.ai\/blog\/chatgpt-searches-in-english-even-when-you-don-t\" target=\"_blank\" rel=\"noopener\">https:\/\/peec.ai\/blog\/chatgpt-searches-in-english-even-when-you-don-t<\/a> <a href=\"#59aca1c6-6097-4288-88af-770ab9af8c7c-link\" aria-label=\"Jump to footnote reference 6\">\u21a9\ufe0e<\/a><\/li><li id=\"588c60f6-a159-454d-87e0-5652d6930c5f\">How Query Language Reshapes AI Citations. Profound, 2026. <a href=\"https:\/\/www.tryprofound.com\/blog\/how-query-language-reshapes-ai-citations\" target=\"_blank\" rel=\"noopener\">https:\/\/www.tryprofound.com\/blog\/how-query-language-reshapes-ai-citations<\/a> <a href=\"#588c60f6-a159-454d-87e0-5652d6930c5f-link\" aria-label=\"Jump to footnote reference 7\">\u21a9\ufe0e<\/a><\/li><li id=\"a8a6a2be-121c-4ad9-94ff-f169d8d0f9c5\">How Evertune Measures AI Visibility: Methodology. Evertune. <a href=\"https:\/\/www.evertune.ai\/platform\/methodology\" target=\"_blank\" rel=\"noopener\">https:\/\/www.evertune.ai\/platform\/methodology<\/a> <a href=\"#a8a6a2be-121c-4ad9-94ff-f169d8d0f9c5-link\" aria-label=\"Jump to footnote reference 8\">\u21a9\ufe0e<\/a><\/li><li id=\"6a503fb5-cf07-4a91-a7fb-3c7aa13829fa\">Evertune vs Scrunch: GEO Platform Comparison. Evertune. <a href=\"https:\/\/www.evertune.ai\/resources\/insights-on-ai\/evertune-vs-scrunch-which-geo-platform-delivers-the-best-results\" target=\"_blank\" rel=\"noopener\">https:\/\/www.evertune.ai\/resources\/insights-on-ai\/evertune-vs-scrunch-which-geo-platform-delivers-the-best-results<\/a> <a href=\"#6a503fb5-cf07-4a91-a7fb-3c7aa13829fa-link\" aria-label=\"Jump to footnote reference 9\">\u21a9\ufe0e<\/a><\/li><\/ol>\n\n\n<p class=\"wp-block-paragraph\"><\/p>",
        "protected": false
    },
    "excerpt": {
        "rendered": "<p>Google AI Overview is an AI-generated summary at the top of the search results page that synthesises information from an average of around 13 sources and links to them. More important than the definition: how AI Overview selects those sources changed fundamentally over the past twelve months, and most advice currently circulating is based on [&hellip;]<\/p>",
        "protected": false
    },
    "author": 3,
    "featured_media": 986,
    "comment_status": "open",
    "ping_status": "open",
    "sticky": false,
    "template": "",
    "format": "standard",
    "meta": {
        "_acf_changed": false,
        "_gspb_post_css": "",
        "footnotes": "[{\"content\":\"Ahrefs Study: Only 38% of Google AI Overview Citations Come From Top 10 Pages (863,000 keywords, 4 million URLs). <a href=\/en\/\\\"https:\/\/ahrefs.com\/blog\/ai-overview-citations-top-10\/\\\"\/ target=\\\"_blank\\\" rel=\\\"noreferrer noopener\\\">https:\/\/ahrefs.com\/blog\/ai-overview-citations-top-10\/ <\/a>\",\"id\":\"883fec62-8e45-441e-b4c4-9b37c4bd962f\"},{\"content\":\"How Google AI Mode and AI Overviews Select Sources. First Motion. <a href=\/en\/\\\"https:\/\/firstmotion.com\/insights\/how-google-ai-mode-and-ai-overviews-select-sources\\\"\/>https:\/\/firstmotion.com\/insights\/how-google-ai-mode-and-ai-overviews-select-sources<\/a>\",\"id\":\"552b32d2-095e-4086-ba51-ea40d337961c\"},{\"content\":\"Google AI Overviews Statistics 2026. DemandSage. <a href=\/en\/\\\"https:\/\/www.demandsage.com\/ai-overviews-statistics\/\\\"\/>https:\/\/www.demandsage.com\/ai-overviews-statistics\/<\/a>\",\"id\":\"20f866e9-edc7-4728-a047-0e1580bae1db\"},{\"content\":\"AI Overview Citations and Rank Overlap. BrightEdge. <a href=\/en\/\\\"https:\/\/www.brightedge.com\/resources\/weekly-ai-search-insights\/rank-overlap-after-16-months-of-aio\\\"\/>https:\/\/www.brightedge.com\/resources\/weekly-ai-search-insights\/rank-overlap-after-16-months-of-aio<\/a>\",\"id\":\"f8883265-1e1f-46e9-8975-c011691e60d5\"},{\"content\":\"How Query Language Reshapes AI Citations. Profound, 2026. <a href=\/en\/\\\"https:\/\/www.tryprofound.com\/blog\/how-query-language-reshapes-ai-citations\\\"\/>https:\/\/www.tryprofound.com\/blog\/how-query-language-reshapes-ai-citations<\/a>\",\"id\":\"dbd24f0f-f8d2-4d1b-92c6-9047cfb14499\"},{\"content\":\"ChatGPT Searches in English, Even When You Don't. Peec AI, 2026 (10 million prompts, 20 million fan-out queries). <a href=\/en\/\\\"https:\/\/peec.ai\/blog\/chatgpt-searches-in-english-even-when-you-don-t\\\"\/>https:\/\/peec.ai\/blog\/chatgpt-searches-in-english-even-when-you-don-t<\/a>\",\"id\":\"59aca1c6-6097-4288-88af-770ab9af8c7c\"},{\"content\":\"How Query Language Reshapes AI Citations. Profound, 2026. <a href=\/en\/\\\"https:\/\/www.tryprofound.com\/blog\/how-query-language-reshapes-ai-citations\\\"\/>https:\/\/www.tryprofound.com\/blog\/how-query-language-reshapes-ai-citations<\/a>\",\"id\":\"588c60f6-a159-454d-87e0-5652d6930c5f\"},{\"content\":\"How Evertune Measures AI Visibility: Methodology. Evertune. <a href=\/en\/\\\"https:\/\/www.evertune.ai\/platform\/methodology\\\"\/>https:\/\/www.evertune.ai\/platform\/methodology<\/a>\",\"id\":\"a8a6a2be-121c-4ad9-94ff-f169d8d0f9c5\"},{\"content\":\"Evertune vs Scrunch: GEO Platform Comparison. Evertune. <a href=\/en\/\\\"https:\/\/www.evertune.ai\/resources\/insights-on-ai\/evertune-vs-scrunch-which-geo-platform-delivers-the-best-results\\\"\/>https:\/\/www.evertune.ai\/resources\/insights-on-ai\/evertune-vs-scrunch-which-geo-platform-delivers-the-best-results<\/a>\",\"id\":\"6a503fb5-cf07-4a91-a7fb-3c7aa13829fa\"}]"
    },
    "categories": [
        6
    ],
    "tags": [],
    "class_list": [
        "post-916",
        "post",
        "type-post",
        "status-publish",
        "format-standard",
        "has-post-thumbnail",
        "hentry",
        "category-aeo-geo-how-to"
    ],
    "blocksy_meta": {
        "has_hero_section": "enabled",
        "hero_elements": [
            {
                "id": "custom_title",
                "enabled": true,
                "heading_tag": "h1",
                "title": "Home",
                "__id": "6vpNanA2EL-thdAiYSJ_L"
            },
            {
                "id": "custom_description",
                "enabled": true,
                "description_visibility": {
                    "desktop": true,
                    "tablet": true,
                    "mobile": false
                },
                "__id": "-EQRFbHB-k3P--mBEU-Rk"
            },
            {
                "id": "custom_meta",
                "enabled": true,
                "meta_elements": [
                    {
                        "id": "author",
                        "enabled": true,
                        "label": "By",
                        "has_author_avatar": "yes",
                        "avatar_size": 25,
                        "__id": "Q2hqTsCmH_mznzPReu3Aj"
                    },
                    {
                        "id": "post_date",
                        "enabled": true,
                        "label": "On",
                        "date_format_source": "default",
                        "date_format": "M j, Y",
                        "__id": "SZXJO-yGuVN0Q_1EmY-S3"
                    },
                    {
                        "id": "updated_date",
                        "enabled": false,
                        "label": "On",
                        "date_format_source": "default",
                        "date_format": "M j, Y",
                        "__id": "hjLUJVCFmm1N62SjFT-s7"
                    },
                    {
                        "id": "categories",
                        "enabled": false,
                        "label": "In",
                        "style": "simple",
                        "__id": "wZYxeryYDhs6_ziCcW-b2"
                    },
                    {
                        "id": "comments",
                        "enabled": true,
                        "__id": "7Isf6KIxMnJ5SIrCTFVqg"
                    }
                ],
                "page_meta_elements": {
                    "joined": true,
                    "articles_count": true,
                    "comments": true
                },
                "__id": "aaBBlzTZQoIcMBgx6XPOQ"
            },
            {
                "id": "breadcrumbs",
                "enabled": false,
                "__id": "0velMIYmgHz4sWiUYtQz2"
            },
            {
                "id": "content-block",
                "enabled": false,
                "__id": "fI7VM2T-ukBDDR-YZAj96"
            }
        ],
        "styles_descriptor": {
            "styles": {
                "desktop": "[data-prefix=\"single_blog_post\"] .entry-header .page-title {--theme-font-size:30px;} [data-prefix=\"single_blog_post\"] .entry-header .entry-meta {--theme-font-weight:600;--theme-text-transform:uppercase;--theme-font-size:12px;--theme-line-height:1.3;}",
                "tablet": "",
                "mobile": ""
            },
            "google_fonts": [],
            "version": 6
        }
    },
    "acf": [],
    "_links": {
        "self": [
            {
                "href": "https:\/\/brandgineer.co\/en\/wp-json\/wp\/v2\/posts\/916",
                "targetHints": {
                    "allow": [
                        "GET"
                    ]
                }
            }
        ],
        "collection": [
            {
                "href": "https:\/\/brandgineer.co\/en\/wp-json\/wp\/v2\/posts"
            }
        ],
        "about": [
            {
                "href": "https:\/\/brandgineer.co\/en\/wp-json\/wp\/v2\/types\/post"
            }
        ],
        "author": [
            {
                "embeddable": true,
                "href": "https:\/\/brandgineer.co\/en\/wp-json\/wp\/v2\/users\/3"
            }
        ],
        "replies": [
            {
                "embeddable": true,
                "href": "https:\/\/brandgineer.co\/en\/wp-json\/wp\/v2\/comments?post=916"
            }
        ],
        "version-history": [
            {
                "count": 11,
                "href": "https:\/\/brandgineer.co\/en\/wp-json\/wp\/v2\/posts\/916\/revisions"
            }
        ],
        "predecessor-version": [
            {
                "id": 976,
                "href": "https:\/\/brandgineer.co\/en\/wp-json\/wp\/v2\/posts\/916\/revisions\/976"
            }
        ],
        "wp:featuredmedia": [
            {
                "embeddable": true,
                "href": "https:\/\/brandgineer.co\/en\/wp-json\/wp\/v2\/media\/986"
            }
        ],
        "wp:attachment": [
            {
                "href": "https:\/\/brandgineer.co\/en\/wp-json\/wp\/v2\/media?parent=916"
            }
        ],
        "wp:term": [
            {
                "taxonomy": "category",
                "embeddable": true,
                "href": "https:\/\/brandgineer.co\/en\/wp-json\/wp\/v2\/categories?post=916"
            },
            {
                "taxonomy": "post_tag",
                "embeddable": true,
                "href": "https:\/\/brandgineer.co\/en\/wp-json\/wp\/v2\/tags?post=916"
            }
        ],
        "curies": [
            {
                "name": "wp",
                "href": "https:\/\/api.w.org\/{rel}",
                "templated": true
            }
        ]
    }
}