By Hussam Sufan · May 27, 2026 · 10 min
Until recently, searching Google was an act of exploration: type keywords, scan 3–4 results, go back, refine. Today the dynamic is different. A user types a long, context-rich question into ChatGPT and gets a complete answer without leaving the interface. No click. No visit. No GA4 session.
Traditional search works like a librarian: it hands you a list of books to investigate yourself. Answer Engines act like a graduate-trained assistant that, besides identifying the shoes you want, recommends the best option for your case, budget and size.
Adoption is growing fast: LLM-referred traffic is still small in absolute terms (0.02–0.03% of total sessions in LATAM client data), but growing 5x year over year. ChatGPT concentrates 95% of that traffic.
Not all queries are equal in the face of AI. Informational questions ("what is a checking account", "symptoms of diabetes") are the ones LLMs resolve most easily — and the ones that stop sending visits to your site.
Most exposed sectors: health, education and news. E-commerce and financial services sit in the middle: research phases get resolved inside the chatbot, but the transaction still requires visiting a site.
Avinash Kaushik, a digital analytics authority, published in late 2025 a loss-and-recovery forecasting model for the Answer Engine era. His premise: companies will lose up to 30% organic traffic and 15% paid traffic during 2026 if they don't take AEO action.
| Query type | Example | Risk | Potential loss |
|---|---|---|---|
| Informational | "what is a checking account" | High | 60% |
| Commercial | "best SMB bank in Chile" | Medium-high | 35% |
| Transactional non-brand | "open online checking" | Moderate | 15% |
| Brand / Navigational | "Tenpo digital account" | Low | 5% |
The strategic lesson: brand marketing has never been more important.
You are an expert in SEO and search intent analysis. I will attach a CSV/Excel file exported from Google Search Console with a list of keywords (queries) and their clicks. Your task is to classify EACH keyword into exactly one of these 4 search intent categories: 1. INFORMATIONAL — The user wants to learn or understand something. Includes "what is", "how does it work", "difference between", "guide to", "tutorial", definitions, concepts, symptoms, explanations. 2. COMMERCIAL — The user is researching options before deciding. Includes "best", "top", "comparison", "vs", "reviews of", "recommendations", "alternatives to". 3. TRANSACTIONAL NON-BRAND — The user wants to take an action (buy, hire, download) but does NOT mention a specific brand. Includes "buy", "price of", "hire", "download", "quote", "book", "open account". 4. BRAND / NAVIGATIONAL — The user mentions a specific brand or wants to go to a particular site. Includes our brand name, our product names, or variations of our URL. INSTRUCTIONS: - Return a table with 3 columns: keyword | clicks | category - Do NOT change the keyword text or click counts - If a keyword is ambiguous, classify it by MOST LIKELY intent - At the end, include a summary with total clicks per category and percentage of each MY BUSINESS CONTEXT (fill in as needed): - Industry: [FILL IN] - Brand name: [FILL IN] - Main products/services: [FILL IN] The attached file contains the exported queries. Classify them all.
Copy this prompt, paste it into ChatGPT (or Claude, or Gemini) and attach the CSV exported from Search Console. In under a minute you'll have keywords classified to feed the calculator.
Enter the data you got from Search Console and GA4 (or what the prompt above produced). The calculator applies Kaushik's model and shows your projected loss in sessions and revenue, plus recovery potential with AEO actions.
Based on Avinash Kaushik's loss model, adapted for global markets
Clicks by category (Search Console)"what is", "how does it work", "difference between"
"best", "comparison", "which to choose"
"buy", "hire", "price of"
"your brand + product", direct searches
Organic channel, sessions report
Revenue attributed to organic channel
Calculating the loss is only half the exercise. The same model proposes 6 recovery actions and 6 growth tactics, each with an estimated gain percentage.
Adapt existing content to work as an AI answer: questions in H2s, direct answers in the first paragraph of each section. Implement JSON-LD schemas (product, FAQ, HowTo, organization, review, aggregate rating) — ~4% recovery. Build topical authority with author bios and structured org data — ~3%. Optimize brand queries, create comparative content where your brand wins, and produce practical guides with use cases.
Topic clusters. Use-case content. Third-party authority and editorial link building. Video and audio (LLMs index YouTube transcripts). Reviews as first-party CMS content. AI-shopping-ready product feeds.
Tools like SearchBrand query multiple LLMs in real time, show whether your brand appears, identify which sources are cited and monitor your visibility score over time and against competitors.
See also: Best AEO Tools 2026.
According to Kaushik's model, companies can lose up to 30% of organic traffic during 2026. Informational queries lose up to 60%; brand queries only 5%.
Informational queries answer questions ("what is"). Commercial compare options ("best X for Y"). Transactional non-brand seek action without a brand ("buy"). Brand queries include your name. Use the included prompt to automate the classification.
Partially. Structured schemas, topical authority and comparative content can recover 10–15% of lost traffic. The key is getting Answer Engines to cite you as a source.
Clicks by intent category from Search Console, total organic sessions from GA4 and organic revenue — all from the same period.
A framework that classifies organic traffic by intent and assigns potential loss percentages, with recovery tactics and estimated gain percentages.