{"id":17919,"date":"2025-06-12T14:05:00","date_gmt":"2025-06-12T12:05:00","guid":{"rendered":"https:\/\/www.sortlist.com\/blog\/?p=17919"},"modified":"2026-06-18T15:27:50","modified_gmt":"2026-06-18T13:27:50","slug":"how-chatgpt-gets-information","status":"publish","type":"post","link":"https:\/\/www.sortlist.com\/blog\/how-chatgpt-gets-information\/","title":{"rendered":"Inside dark search: 7,000 sources reveal how AI recommends your product"},"content":{"rendered":"\n

By 2026, the search funnel will have changed beyond all recognition. AI will control the information online, and the influence on most buying decisions.<\/p>\n\n\n\n

Google still dominates – but influence has moved behind AI-generated answers. And dark search<\/strong> is what happens when your customers get answers from AI (Google’s AI Mode or AI Overviews, ChatGPT or Perplexity) – without ever clicking through to your website.<\/p>\n\n\n\n

No clicks, no analytics. The same influence, just harder to track.<\/p>\n\n\n\n

It sounds extreme. But we analysed 7,000 AI-chosen sources – and the numbers say the same:<\/p>\n\n\n\n

\n
AI now influences decisions upstream:<\/strong> 90% of informational queries are answered by AI, not clicked through.<\/li>\n\n\n\n
Google \u2260 the full picture<\/strong>: There\u2019s just a 13\u201315% overlap between Google\u2019s top results and ChatGPT<\/li>\n\n\n\n
RAG is your entry point<\/strong>: Real-time search (Retrieval-Augmented Generation) pulls fresh, structured, and authoritative content from the web.<\/li>\n\n\n\n
Two formats dominate<\/strong>: \u201cBest of\u201d lists and \u201cX vs Y\u201d comparison pages are the most cited AI sources.<\/li>\n\n\n\n
Blogs and directories rule<\/strong>: 70% of AI-cited content comes from blogs (51%) and directories (19%): not homepages or product pages.<\/li>\n<\/ul>\n\n\n\n
Welcome to the era of \u201cdark search\u201d<\/strong>. <\/p>\n\n\n\n
This research unpacks the AI black box – with data on how it chooses what to recommend, exactly where it finds its answers, and how to target its sources so your brand appears first. <\/p>\n\n\n\n
ChatGPT, Google, and the invisible funnel<\/h2>\n\n\n\n
If you thought ChatGPT\u2019s launch (100 million monthly users in two months) was impressive – look at its growth in 2024.<\/p>\n\n\n\n
$\"ChatGPT$
Source: Similarweb<\/a><\/figcaption><\/figure>\n\n\n\n
Yet Google is still the dominant force in search, with 16.4 billion daily <\/strong>searches.<\/p>\n\n\n\n
But the real story is how it has changed search habits.<\/p>\n\n\n\n
Google quickly reacted, releasing AI Overviews in May 2024: ChatGPT-style answers above traditional search results. Almost 20% of searches get an AI Overview.<\/p>\n\n\n\n
$\"how$ Source: SE Ranking<\/a><\/figcaption><\/figure>\n\n\n\n
Today, 90% of AI-generated searches<\/strong> are informational, although they haven’t drastically changed the volume of informational queries (users actually ask more, because AI delivers instant, better answers). <\/p>\n\n\n\n
$\"how$ Source: Semrush<\/a><\/figcaption><\/figure>\n\n\n\n
But these queries no longer drive traffic to your site. AI Overview (AIO) queries send 8 million clicks per day through to the sources it uses. But pre-AIO, this figure would have been around 100 million. <\/p>\n\n\n\n
The gap: 92 million lost clicks per day<\/strong>, and widening.<\/p>\n\n\n\n
$\"how$
Source: Google\/Semrush<\/figcaption><\/figure>\n\n\n\n
And in May 2025 Google flipped the switch on AI Mode, a full chat box inside the search results. No blue links, just a cited, ChatGPT-style answer that users rarely click past.<\/p>\n\n\n\n
$\"AI$ <\/figure>\n\n\n\n
<\/p>\n\n\n\n
\n\n\n\n
<\/p>\n\n\n\n
AI already controls information online<\/h3>\n\n\n\n
So while Google is still the gatekeeper of clicks, it\u2019s no longer the gatekeeper of influence<\/em>.<\/p>\n\n\n\n
AI (in all its forms) is now what influences decisions pre-purchase.<\/p>\n\n\n\n
ChatGPT, Perplexity, AI Mode and Google\u2019s AI Overviews <\/strong>surface information, make recommendations, and even name products. But most of the time, the user doesn\u2019t click (and you don\u2019t see their traffic in analytics). <\/p>\n\n\n\n
Users get their information (via AI), make their decision, and they purchase direct. <\/p>\n\n\n\n
That\u2019s the search funnel in the era of dark search<\/strong>: <\/p>\n\n\n\n
\n
The 53% of Informational questions are fulfilled by AI<\/strong>, and drive little to no traffic.<\/li>\n\n\n\n
The other 47% of traffic remains<\/strong> – concentrated on navigational, commercial and transactional queries – demand that has been influenced by AI answers.<\/li>\n<\/ul>\n\n\n\n $\"how$
Source: Sparktoro<\/a><\/figcaption><\/figure>\n\n\n\n
There\u2019s no Google top 10 to crack, no keyword volume to chase. That\u2019s why we call it dark search<\/strong>: influence without visibility, recommendations without rankings.<\/p>\n\n\n\n
But to get AI recommending your product, your all-important transactional traffic will remain.<\/p>\n\n\n\n
Can you influence AI\u2019s recommendations?<\/p>\n\n\n\n
Yes.<\/strong> And we\u2019ve done the research to prove exactly how.<\/p>\n\n\n\n
<\/p>\n\n\n\n
\n\n\n\n
<\/p>\n\n\n\n
Dark search: how AI chooses winners<\/h2>\n\n\n\n
70% of searches on ChatGPT can\u2019t be classified<\/strong> in the \u201ctraditional\u201d search buckets: informational, navigational, commercial, and transactional.<\/p>\n\n\n\n
<\/p>\n\n\n\n
$\"search$
Source: Semrush<\/figcaption><\/figure>\n\n\n\n
<\/p>\n\n\n\n
And that\u2019s because people don\u2019t search on ChatGPT like they do on Google. When talking with AI, users ask full questions. They mix commands with comparisons and include context.<\/p>\n\n\n\n

\ud83d\udc49 \u201cWhich are the best CRMs for pre-seed startups in 2025\u201d
\ud83d\udc49 \u201cCompare Notion and Trello for content planning\u201d
\ud83d\udc49 \u201cShow me alternatives to Mailchimp with better support. I don\u2019t need fancy designs\u201d<\/p>\n\n\n\n
These queries are:<\/p>\n\n\n\n
\n
Conversational<\/li>\n\n\n\n
Mixed-intent<\/li>\n\n\n\n
Often specific to use case or user role
<\/li>\n<\/ul>\n\n\n\n
To tackle this complex kind of query AI normally pulls answers from two places – static retrieval or web search. One hard to influence, the other much easier.<\/p>\n\n\n\n
<\/p>\n\n\n\n
Static retrieval: AI\u2019s trained memory<\/strong><\/h3>\n\n\n\n
This is everything ChatGPT read before<\/em> its last update. Books, websites, articles, academic journals… It\u2019s used for questions where the answer doesn\u2019t change over time.<\/p>\n\n\n\n
Think:<\/p>\n\n\n\n
\ud83d\udc49 \u201cWhen did the Roman Empire fall?\u201d
\ud83d\udc49 \u201cHow does influencer marketing differ for industrial brands?\u201d
<\/p>\n\n\n\n
You can\u2019t easily influence these answers. They’re baked into the model\u2019s memory. Unless you control a massive portion of the public internet, you\u2019re unlikely to change its view.<\/p>\n\n\n\n
<\/p>\n\n\n\n
RAG: Real-Time Search<\/strong><\/h3>\n\n\n\n
RAG stands for Retrieval-Augmented Generation<\/strong>. This is when ChatGPT actively looks online: hitting APIs, scraping structured content, checking documentation.<\/p>\n\n\n\n
Think:<\/p>\n\n\n\n
\ud83d\udc49 \u201cWhat\u2019s the cheapest CRM for startups in 2025?\u201d
\ud83d\udc49 \u201cCompare Monday.com vs ClickUp\u201d
\ud83d\udc49 \u201cBest noise-cancelling headphones under \u20ac200\u201d
<\/p>\n\n\n\n
These are your real opportunities<\/strong>. If your content is recent, structured, and authoritative, AI might pull directly from you – and being present in them is the easiest way to get recommended by AI.
<\/p>\n\n\n\n
The search funnel in 2026<\/h2>\n\n\n\n
This new user behaviour requires a new concept of the search funnel. It needs to: <\/p>\n\n\n\n
\n
Take into account the 70% of queries that don’t fit into the traditional search buckets.<\/li>\n\n\n\n
Understand the extent of real-time (RAG) information retrieval: the simplest way to influence results. <\/li>\n<\/ul>\n\n\n\n
So we analysed hundreds of AI queries and grouped them by intent, stage, and how likely they are to trigger a real-time (RAG) response. The pattern is clear:<\/p>\n\n\n\n
\n