Step 2: Connect Google Sheets for Client and SEO Data

Load client details and the SEO sheet from the URL passed by the webhook.

1. In Fetch Client Details, set Sheet Name to Client information and Document ID to ={{ $('Chat Webhook Trigger').item.json.chatInput }}.
2. Credential Required: Connect your googleSheetsOAuth2Api credentials in Fetch Client Details.
3. In Retrieve SEO Sheet, set Sheet Name to SEO and Document ID to ={{ $('Chat Webhook Trigger').item.json.chatInput }}.
4. Credential Required: Connect your googleSheetsOAuth2Api credentials in Retrieve SEO Sheet.
5. Verify the execution flow: Chat Webhook Trigger → Fetch Client Details → Retrieve SEO Sheet.

Step 3: Filter and Batch Keywords for Processing

Filter eligible keyword rows and iterate through them in batches for SERP analysis.

1. In Filter Keyword Rows, keep the conditions that check ={{ $json['Keyword'] }} is not empty and ={{ $json['<h1>'] }} is empty.
2. In Validate Keyword Presence, ensure the condition checks that ={{ $json['Keyword'] }} exists before continuing.
3. Use Batch Through Keywords to iterate the filtered rows; this will later loop back from Update Spreadsheet Row to process the next keyword.
4. Confirm the flow: Retrieve SEO Sheet → Filter Keyword Rows → Validate Keyword Presence → Batch Through Keywords.

Step 4: Configure SERP Search and Competitor Crawling

Search the SERP and crawl competitor pages, then extract headings for AI analysis.

1. In Search SERP via API, set URL to https://api.apify.com/v2/acts/nFJndFXA5zjCTuudP/run-sync-get-dataset-items and keep Method as POST.
2. Set JSON Body to the provided payload and keep the query line "queries": "{{ $json['Keyword'] }}" intact.
3. Credential Required: Connect your httpHeaderAuth credentials in Search SERP via API.
4. Connect Firecrawl credentials to all five crawl nodes: Crawl Result One, Crawl Result Two, Crawl Result Three, Crawl Result Four, and Crawl Result Five.
5. Confirm each crawl node uses the proper URL expression, e.g., Crawl Result One → URL ={{ $('Search SERP via API').item.json.organicResults[0].url }}.
6. Keep the sequential extraction chain intact: Crawl Result One → Extract Headings One → Crawl Result Two → Extract Headings Two → Crawl Result Three → Extract Headings Three → Crawl Result Four → Extract Headings Four → Crawl Result Five → Extract Headings Five.

Step 5: Set Up AI Generation and Parsing

Generate SEO metadata and content briefs using Anthropic models, Supabase retrieval, and OpenAI embeddings.

1. In Generate Meta & H1, keep the full prompt text and confirm the expressions referencing keywords, client details, and competitor headings are intact.
2. Attach Anthropic Model Primary as the language model for Generate Meta & H1. Credential Required: Connect your anthropicApi credentials in Anthropic Model Primary.
3. Attach Supabase Retriever Primary as a tool to Generate Meta & H1 and keep Tool Name as BaseDeDonnees with Table Name ={{ $('Fetch Client Details').item.json['Supabase database'] }}_documents. Credential Required: Connect your supabaseApi credentials in Supabase Retriever Primary.
4. OpenAI Embeddings Primary supplies embeddings to Supabase Retriever Primary. Credential Required: Connect your openAiApi credentials in OpenAI Embeddings Primary. Add credentials to the parent nodes, not the sub-node connection.
5. Attach Structured JSON Parser as the output parser for Generate Meta & H1, and keep the schema example as provided. Add parser settings to the parent node, not the parser sub-node.
6. In Compose Content Brief, keep the prompt text that references {{$('Generate Meta & H1').item.json.output.h1}} and competitor headings from all extraction nodes.
7. Attach Anthropic Model Secondary as the language model for Compose Content Brief. Credential Required: Connect your anthropicApi credentials in Anthropic Model Secondary.
8. Attach Supabase Retriever Secondary with the same table expression ={{ $('Fetch Client Details').item.json['Supabase database'] }}_documents and connect OpenAI Embeddings Secondary for embeddings. Credential Required: Connect your supabaseApi and openAiApi credentials in those nodes.

Step 6: Assemble and Update the SEO Sheet

Merge AI outputs with original rows and update the Google Sheet.

1. In Assemble Sheet Update, keep the code that maps AI outputs to <title>, <meta-desc>, <h1>, and brief.
2. In Update Spreadsheet Row, set Operation to update and Sheet Name to FR.
3. Set Document ID to ={{ $('Chat Webhook Trigger').item.json.chatInput }}.
4. Ensure Matching Columns includes mots clés for the update match.
5. Credential Required: Connect your googleSheetsOAuth2Api credentials in Update Spreadsheet Row.
6. Confirm the loop: Assemble Sheet Update → Update Spreadsheet Row → Batch Through Keywords to process the next row.

Step 7: Test & Activate Your Workflow

Validate end-to-end execution and enable the workflow for production use.

1. Click Execute Workflow and send a test request to Chat Webhook Trigger with a valid Google Sheets URL in chatInput.
2. Verify that Search SERP via API returns organicResults and that all Crawl Result and Extract Headings nodes produce heading outputs.
3. Confirm Generate Meta & H1 outputs structured fields matching meta_title, meta_description, and h1 from Structured JSON Parser.
4. Check the Google Sheet to ensure <title>, <meta-desc>, <h1>, and brief are updated in the FR sheet.
5. Once verified, toggle the workflow to Active to run in production.