--- name: dataforseo-ai-answer-visibility description: "Weekly: of your top 20 non-branded Google queries, how many get an AI answer citing your domain in ChatGPT, Perplexity or AI Overviews, plus AI referral visits. Use when the user asks for this, or when it runs on a schedule." --- # AI answer visibility: DataForSEO, GA4 and Search Console Weekly: of your top 20 non-branded Google queries, how many get an AI answer citing your domain in ChatGPT, Perplexity or AI Overviews, plus AI referral visits. Ranking third on Google no longer means you are in the answer. AI Overviews, ChatGPT and Perplexity each pick a handful of sources, and a page that ranks well can be left out of all of them. Search Console cannot tell you, and the AI tools have no report for site owners. Every Monday the agent takes your top 20 non-branded queries from Search Console, asks DataForSEO for the AI Overview, the ChatGPT answer and the Perplexity answer to each, and checks whether your domain is cited. It sets that against the visits GA4 records from AI assistants, landing page by landing page, so you see which cited pages actually send people. ## Two ways to get the data Use the API or MCP route when you have an MCP server, a CLI or an API token for the app. Otherwise, if you can drive a browser the user is signed in to (Claude in Chrome, a Playwright or Chrome DevTools MCP server), use the browser route. Say which one you used. ## Route A: API or MCP server ### What you need - A DataForSEO API login and password in `DATAFORSEO_LOGIN` and `DATAFORSEO_PASSWORD`, or the DataForSEO MCP server, with a balance (each run spends about $0.30). - Read access to the Search Console property and Viewer access to the GA4 property, plus your brand terms as a regular expression. If something is missing, say what and stop; do not invent numbers. ### Steps 1. **Fix the 20 queries for the month** (Google Search Console): On the first run each month, call `searchanalytics.query` with `dimensions: ["query"]` for the last 28 days, drop queries matching your brand regular expression, and keep the top 20 by clicks with their average position and top page. Save them in `ai-visibility-queries.json` so every weekly run in the month asks the same 20. 2. **Ask the three AI surfaces** (DataForSEO): For each query: `POST /v3/serp/google/organic/live/advanced` with `keyword`, `location_code` (2840 for the US), `language_code: "en"`, `depth: 10` and `load_async_ai_overview: true`, reading the `references` (`domain`, `url`) inside the `ai_overview` item; `POST /v3/ai_optimization/chat_gpt/llm_scraper/live/advanced` with the same keyword and `force_web_search: true`, reading `sources[].domain`; and `POST /v3/ai_optimization/perplexity/llm_responses/live` with `user_prompt` and `model_name: "sonar"`, reading `items[].sections[].annotations[].url`. That is 60 paid requests a run, roughly $0.30 at list prices; sum the `cost` fields and test on `https://sandbox.dataforseo.com/v3/` first. 3. **Pull visits from AI assistants** (Google Analytics): Call `runReport` for the last 28 days with `dimensions: ["sessionSource", "landingPagePlusQueryString"]`, metrics `sessions` and `keyEvents`, and a `dimensionFilter` on `sessionSource` with `matchType: "PARTIAL_REGEXP"` and `chatgpt\.com|perplexity\.ai|gemini\.google\.com|claude\.ai|copilot\.microsoft\.com`. 4. **Join citations, rankings and visits**: A query counts as cited when any cited domain, lowercased and without `www.`, equals your domain or ends with `.yourdomain`. Anti-join: queries where Search Console has you at position 3 or better and no surface cites you. Then reduce cited URLs and GA4 landing pages to paths (no query string, which strips ChatGPT's `utm_source`, no trailing slash) and join: cited pages that get AI visits, and AI-visited pages that none of the 20 answers cite. 5. **Compare and post**: Read last week's card with `GET https://briefkin.dev/v1/cards/insight-dataforseo-ai-answer-visibility` and append this week's count to its history. Post a metric card: cited queries out of 20, the split by surface, AI-assistant sessions, and a table of the cited queries and the top-3 queries nobody cites. ## Route B: the browser ### What you need - Signed in to Search Console and GA4, and to ChatGPT and Perplexity accounts of your own. If something is missing, say what and stop; do not invent numbers. ### Steps Start at https://search.google.com/search-console/performance/search-analytics. The user is signed in already; never type passwords or codes, and ask if a sign-in page appears. 1. **Take the top 20 queries**: Pick the property, set Date to `Last 28 days`, sort Queries by Clicks, skip branded queries and note the top 20 with their average position. 2. **Ask ChatGPT and Perplexity**: In `https://chatgpt.com/` with search on, and in `https://www.perplexity.ai/`, ask each query in a new chat. Open the sources or citations panel and note whether your domain is listed. 3. **Check Google's AI Overview**: Search each query on `https://www.google.com/`. When an AI Overview appears, expand it and check whether one of its source links points at your domain. 4. **Read AI-assistant visits**: In GA4, open Reports > Acquisition > Traffic acquisition for the last 28 days, set the primary dimension to Session source, and filter for `chatgpt.com`, `perplexity.ai`, `gemini.google.com`, `claude.ai` and `copilot.microsoft.com`. Add Landing page + query string as a secondary dimension and note sessions per page. 5. **Count and compare**: Count queries cited on any surface, list top-3 queries cited nowhere, and match cited pages with the AI-visited landing pages. Note: Answers in your own signed-in accounts are personalized by your history, memory and location, so they can differ from what DataForSEO's clean sessions return; the browser route costs nothing but takes about an hour for 60 asks. Google shows AI Overviews only for some queries and some signed-in users. ## The report - **Cited queries** out of your top 20 non-branded queries, and the change on last week. - **By surface**: AI Overviews, ChatGPT and Perplexity. - **Top-3 queries never cited**: where you win on Google and lose in AI answers. - **AI-assistant visits** by landing page, with the cited pages that send them. - The run's cost from the `cost` fields, and the first page to rework for AI answers. ## Post it to Briefkin When `BRIEFKIN_TOKEN` is set, or `~/.config/briefkin/env` exists (the installer from https://briefkin.dev/docs writes it), also send the result to the user's Briefkin board as one card. Keep the `item_key` the same on every run so the card updates in place, use a new `event_id` per run, and put today's real numbers in place of the example ones. If the request fails, say so once and carry on. ```bash [ -n "${BRIEFKIN_TOKEN:-}" ] || . ~/.config/briefkin/env curl -sS -X POST "${BRIEFKIN_URL:-https://briefkin.dev}/v1/events" \ -H "authorization: Bearer $BRIEFKIN_TOKEN" \ -H "content-type: application/json" \ -d @- <<'JSON' { "v": 1, "event_id": "dataforseo-ai-answer-visibility/", "item_key": "dataforseo-ai-answer-visibility", "project": "example.com", "source": "AI answer visibility: DataForSEO, GA4 and Search Console skill", "kind": "metric", "title": "Cited in 4 of 20 top queries; 6 top-3 queries never cited", "metric": { "label": "Top 20 queries citing you", "value": 4, "period_label": "This week", "better": "higher", "period_start": "", "period_end": "" }, "glance": [ { "type": "Grid", "props": { "columns": 3 }, "children": [ { "type": "Stat", "props": { "label": "AI Overviews", "value": "2 of 20" } }, { "type": "Stat", "props": { "label": "ChatGPT", "value": "3 of 20" } }, { "type": "Stat", "props": { "label": "Perplexity", "value": "2 of 20" } } ] }, { "type": "Badges", "props": { "items": [ { "label": "6 top-3 queries never cited", "tone": "warn" }, { "label": "412 AI sessions, 28 days", "tone": "info" } ] } } ], "widget": [ { "type": "Table", "props": { "columns": [ "Query", "Google position", "AI Overview", "ChatGPT", "Perplexity" ], "align": [ "left", "right", "left", "left", "left" ], "rows": [ [ "free invoice template", 2.4, "Cited", "Cited", "No" ], [ "net 30 payment terms", 3.9, "Cited", "No", "Cited" ], [ "invoice vs receipt", 6.2, "No", "Cited", "Cited" ], [ "late payment email template", 8.4, "No", "Cited", "No" ], [ "invoice number format", 1.4, "No", "No", "No" ], [ "invoice generator", 1.8, "No", "No", "No" ], [ "proforma invoice", 2.2, "No", "No", "No" ], [ "recurring invoice", 2.6, "No", "No", "No" ], [ "how to invoice as a freelancer", 2.9, "No", "No", "No" ], [ "invoice due date", 3, "No", "No", "No" ] ] } }, { "type": "Callout", "props": { "text": "251 of the 412 AI-assistant sessions landed on /blog/invoice-vs-receipt, cited by ChatGPT and Perplexity. Start with /blog/invoice-number-format: position 1.4, cited nowhere. Run cost: 60 requests, $0.31.", "tone": "info" } } ], "size": "md" } JSON ``` Widgets come from the Briefkin catalog (https://briefkin.dev/v1/catalog). Without a token, just give the user the report. Set `period_start` and `period_end` to the window the number covers: the tile names it from them ("Last 7 days", "Today", "This month"). For a count right now (open items, a balance), send `"period_label": ""` so no period shows; for another name, send it as `period_label` (at most 24 characters). ## Run it on a schedule Suggested: Mondays at 8:00 (`0 8 * * 1`). From cron, with the agent signed in on that machine: ```bash 0 8 * * 1 claude -p "Run the dataforseo-ai-answer-visibility skill" # Codex: codex exec, Gemini CLI: gemini -p, Copilot CLI: copilot -p, OpenCode: opencode run ``` Or use the agent's own scheduler (Claude Code routines, Codex automations, Claude or ChatGPT scheduled tasks): https://briefkin.dev/skills/dataforseo-ai-answer-visibility#schedule Source: https://briefkin.dev/skills/dataforseo-ai-answer-visibility