See llms.txt for all machine-readable content.

Back to Templates

Send weekly AI visibility reports with Apify, OpenAI and SMTP email

Created by

Created by: Bilal || babel42
Bilal

Last update

Last update 8 hours ago

Categories

Quick overview

Every Monday this workflow runs the Apify AI Visibility Tracker for each client to measure how ChatGPT, Perplexity, Gemini and Google AI Overviews mention their brand. It compares results with last week, writes a plain-English summary with OpenAI, and sends each client a report by email, Slack or both.

How it works

  1. Runs every Monday at 8am on a schedule trigger.
  2. Loads your agency settings and client list. Each client can paste their own buyer questions in bulk, one per line. Otherwise the Actor writes 10 questions on the first run and the same set is reused every week.
  3. Starts an Apify run of babel42/ai-visibility-tracker for each client, checks it once a minute until it finishes, then fetches its RUN_SUMMARY. Long question lists are fine.
  4. Compares this week's metrics with last week's saved results and stores the new ones. Changing a client's questions starts a new baseline. Failed runs are flagged.
  5. If the run succeeded, OpenAI writes a 3 to 5 sentence summary from the week-on-week brief.
  6. Builds an HTML email and a shorter Slack post, then sends whichever the client has set up: email, Slack or both. If a run failed, your agency gets an alert by email and/or Slack instead.

Setup

  1. Apify: create an HTTP Header Auth credential with name Authorization and value Bearer <YOUR_APIFY_TOKEN>, then select it in Start the run, Check the run and Get the results.
  2. OpenAI: add an OpenAI credential in OpenAI Chat Model, or replace it with another chat model.
  3. Email and Slack: add an SMTP credential in Send the email (or swap in Gmail or Outlook) and a Slack credential in Post to Slack, and invite the Slack app to each channel. Delete the branch you don't use.
  4. Clients: in Your clients, enter your agency details and one entry per client: brand, domain, category, competitors, market, a reportTo email and/or slackChannel, and optionally their own questions pasted one per line.
  5. Activate the workflow. Last week's results are saved only on scheduled runs, so a manual test run always looks like a first report.

Requirements

  • An Apify account and API token (about $0.045 per question per AI assistant, so about $1.80 per client per week for 10 questions on 4 assistants)
  • An OpenAI API key, or any other chat model n8n supports
  • An SMTP, Gmail or Outlook account for email, and/or a Slack workspace with an app that can post to your channels

Customization

  • Keep clients in a Google Sheet: replace Your clients with a Google Sheets read that outputs the same fields. A questions cell with one question per line works as it is.
  • Change the AI assistants, number of questions or spending cap per client with the assistants, maxQueries and maxBudgetUsd fields.
  • Change the schedule in Every Monday at 8am, or edit the summary prompt in Write the summary.

Additional info

Each client's question set stays the same every week so the trend is a fair comparison. All clients' runs start at once and each uses 1 GB of Apify memory, so check your plan's memory limit if you have many clients. A failed run alerts your agency with the run's status and a link to it, and the client gets nothing that week.