Retention & Churn Owner (AI-first, B2B SaaS)
TimelinesAI is hiring a Retention & Churn Owner: one AI-first operator who owns churn of paying customers and runs the loop every day. This is a full-time, fully remote role.
🚀 About TimelinesAI
TimelinesAI is a B2B SaaS that connects WhatsApp to CRMs (HubSpot, Pipedrive, Zoho and others) and to AI tools through MCP. About 2,800 paying SMB customers worldwide, most of them self-serve (PLG).
🎯 The problem
Churn is the biggest lever on our growth right now. A large share of it is failed payments, and our recovery flows broke twice this year before anyone noticed. We already have the data and the agents. What's missing is one person who owns the number and runs the loop every day.
🛠️ What's already built (you don't start from zero)
- An internal agentic CRM. It puts billing, product usage, support and call data on one card per customer, and it has an MCP server you can query from Claude Code.
- Hourly at-risk lists with the money at stake: cancellation booked, payment failed, WhatsApp disconnected, stopped using the app.
- A churn agent that looks into each account, gives a verdict and drafts a letter. A second agent fact-checks the letter, and a person decides whether to send it.
- A 10% control group that never gets contacted, so every action is measured against doing nothing.
- All rules live in GitHub. You change a rule through a pull request.
📌 What you'll own
- Failed payments, first. Make sure every failed invoice reaches the recovery flow (Stripe + Churnkey), and check every week that the campaigns are actually firing. Build a same-day track for bigger accounts across email, SMS, in-app chat and WhatsApp.
- Booked cancellations. Every one gets a decision within 24 hours, and you reach out to the ones worth saving.
- Talking to customers. Get on calls and WhatsApp chats with at-risk and recently churned customers, find out what really happened, and save the ones that can be saved. Large accounts go to sales with full context.
- Win-back and reconnect. Bring back customers who left, plus customers who still pay but whose WhatsApp went dark.
- Better agents. When a letter is wrong, fix the rule behind it (prompt, template or threshold), not just that one letter. The aim is to safely move small accounts to automatic sending.
- Root causes. Each month, report the top 3 churn reasons in dollars to the product team, with customer quotes.
- A weekly written update with your numbers.
📈 Targets for your first 3 months
- Churned MRR down about 20%
- Failed-payment recovery back to at least our 2025 level and above the control group
- 100% of booked cancellations decided within 24 hours
- Every at-risk account paying $300+/month has a named person within 1 working day
🧠 You'll fit if you
- Run AI workflows yourself every day (Claude Code or Codex, MCP, agents, n8n/Make, scripts) and can show us real examples.
- Have owned retention, churn or customer success in B2B SaaS, ideally self-serve with SMB customers, and know your numbers: churned MRR, recovery rate, save rate.
- Enjoy talking to people. An unhappy customer on a call doesn't scare you, and you'd rather pick up the phone than send a fourth email.
- Are comfortable with data: cohorts, SQL/PostHog-level questions, control groups. You check a number before you trust it.
- Write clearly and briefly in English.
💡 Nice to have: Portuguese or Spanish; Stripe dunning, Churnkey, Intercom or Customer.io; working in GitHub with pull requests.
Not a fit: agency backgrounds; people who mainly want to manage others; strategy decks without hands-on execution.
🧪 How we hire
- Screening questions: a real AI workflow and a real customer conversation, in your own words.
- Short call.
- Paid pilot of about 10 hours on a real slice of our at-risk queue.
🌎 How we work
- Full-time, fully remote, async-first; some overlap with CET working hours.
- You report to the founder and work daily with support, sales and product, with a weekly 1-hour sync.
- AI-native team: every function here is being rebuilt around agents.