Track fitness class attendance and retention with n8n, Google Sheets, and AI

Webhook check-in → Google Sheets member lookup → AI engagement label & personalized message → log, email, and Discord alert → weekly retention report every Monday

Who

Fitness studio operators or small-gym admins who want automated member management

You get

Automatically log class check-ins, identify at-risk members, send personalized follow-ups, and deliver a weekly retention digest to managers

You need

n8n · google-sheets · gmail · discord

Steps

1. Prepare spreadsheets and credentials

   Create two tabs in Google Sheets: MemberRoster and AttendanceLog. Ensure columns include member_id, member_name, member_email, membership, join_date, checkin_date, status, etc. In n8n, configure the matching credentials: Google Sheets OAuth2, OpenAI API, Gmail OAuth2, and Discord Bot.

2. Configure webhook trigger and input validation

   Set up a Webhook node in n8n with path fitness/checkin and method POST. The front-desk system sends JSON with member_id, class_name, status, and optional notes. The Normalize Check-in node generates a checkin_id and standardizes fields. The Validate Check-in Input node checks that member_id and class_name are present; if missing, it triggers Discord and Gmail alerts.

   Execution prompt · Source text。Alert Invalid on Discord node content field

   Copy prompt

   ={{ 'Invalid check-in rejected: member_id=' + $json.member_id + ' class=' + $json.class_name + ' at ' + $json.received_at + '. Missing required fields.' }}

3. Match member roster

   The Read Member Roster node reads the MemberRoster sheet from Google Sheets. The Match Member by ID node uses a Code node to match the check-in member_id against the roster and extract member_name, member_email, membership, join_date. If not found, it triggers Discord and Gmail unknown-member alerts.

   Execution prompt · Source text。Alert Unknown on Discord node content field

   Copy prompt

   ={{ 'Unknown member check-in: id=' + $json.member_id + ' class=' + $json.class_name + ' at ' + $json.received_at + '. Member not in roster; no email sent.' }}

4. AI judges engagement and drafts a message

   The Build Agent Input node assembles member and check-in context for the AI Agent node. AI Judge Engagement and Write Message uses gpt-4o-mini to label engagement (at_risk/new/engaged) based on check-in status and generate a short personalized message under 50 words. Parse Agent Output parses the AI-returned JSON, extracts engagement_status and message, and flags needs_attention.

   Execution prompt · Source text。AI Judge Engagement and Write Message node text field (user input)

   Copy prompt

   ={{ 'Member: ' + $json.member_name + ' | membership: ' + $json.membership + ' | class today: ' + $json.class_name + ' | check-in status: ' + $json.status + ' | notes: ' + $json.notes + ' | joined: ' + $json.join_date + ' | new member: ' + $json.is_new + ' | received at: ' + $json.received_at }}

   Execution prompt · Source text。AI Judge Engagement and Write Message node options.systemMessage field

   Copy prompt

   You are the member experience assistant for a small fitness studio. A front desk class check-in just came in. Judge the member engagement: at_risk when the check-in status is no_show (write a warm, gentle nudge encouraging them to come back soon, never blaming them); new when the member joined within the last 30 days (write a short encouraging note that helps them build the habit); otherwise engaged (write a brief warm thank you). Write ONE short message under 50 words addressed to the member by first name. Never mention internal metrics or algorithms. Return ONLY valid JSON with double quoted keys and no code fences: {\"engagement_status\":\"at_risk|new|engaged\",\"message\":\"...\"}. No markdown, no emojis, no text outside the JSON.

5. Log check-in and send notifications

   Build Check-in Log Row formats the record; Append Check-in to Log writes it to the AttendanceLog sheet. Build Member Message assembles the email content; Email Member Message sends it via Gmail. If needs_attention is true (at_risk or no_show), Alert Manager on Discord posts a retention alert to the manager channel.

   Execution prompt · Source text。Alert Manager on Discord node content field

   Copy prompt

   ={{ 'Retention alert: ' + $('Build Check-in Log Row').item.json.member_name + ' | class ' + $('Build Check-in Log Row').item.json.class_name + ' | status ' + $('Build Check-in Log Row').item.json.status + ' | AI ' + $('Build Check-in Log Row').item.json.engagement_status + '. Follow up to keep them engaged.' }}

6. Weekly scheduled retention report

   Weekly Insights Trigger fires every Monday at 09:00 (cron: 0 9 * * 1). Compute Weekly Stats reads AttendanceLog and MemberRoster to calculate weekly check-ins, unique members, class breakdown, no-show rate, at-risk members (no check-in for 14+ days, joined >30 days ago), and new members (joined within 7 days). AI Write Weekly Insight uses gpt-4o-mini to generate a friendly digest under 120 words. Parse Weekly Insight parses the output, then Email Weekly Report sends it via Gmail and Post Weekly Report on Discord pushes it to the server.

   Execution prompt · Source text。AI Write Weekly Insight node text field (user input)

   Copy prompt

   ={{ 'Write the weekly studio report. Data: ' + $json.stats_blob }}

   Execution prompt · Source text。AI Write Weekly Insight node options.systemMessage field

   Copy prompt

   You are the studio analytics assistant for a small fitness studio. Given the weekly attendance stats, write a friendly digest under 120 words for the studio owner: total check-ins, unique members, the busiest class, no-show count and rate, any members at risk of churn, and new members for the week. End with one actionable recommendation. If the week was quiet or there was no data, say so plainly. Return ONLY valid JSON: {\"digest\":\"...\"}. No markdown, no emojis, no text outside the JSON.

   Execution prompt · Source text。Post Weekly Report on Discord node content field

   Copy prompt

   ={{ 'Weekly studio report: ' + $('Parse Weekly Insight').item.json.total_checkins + ' check-ins, ' + $('Parse Weekly Insight').item.json.unique_members + ' members, ' + $('Parse Weekly Insight').item.json.no_show_rate + '% no-show.\n' + $('Parse Weekly Insight').item.json.digest }}

You are setting up an n8n workflow to track fitness class attendance and member retention. Please confirm first: 1) Google Sheets has MemberRoster and AttendanceLog tabs with matching column headers; 2) n8n credentials are configured for Google Sheets OAuth2, OpenAI API, Gmail OAuth2, and Discord Bot; 3) the front-desk system can POST to /fitness/checkin. Then import workflow.json or configure nodes manually following the step descriptions.