You asked it to return "hot" or "cold". Monday it returned "hot". Friday: "I'd say this one looks fairly hot." Your IF node doesn't match, and the lead silently goes nowhere. The build (n8n, ~10 min): → Every AI node feeding a decision returns a fixed shape, not a sentence: {"tag":"hot","reason":"..."} → List the allowed values in the prompt, and add one more: "unknown" if it isn't sure → Switch the node to JSON / structured output mode if it has one → Next node checks the tag is on your list. Not on it → route to a review branch, never the default path → Log every "unknown" in a Sheet The review branch is the whole build. A default ELSE path treats a broken answer as a real one — that's how a hot lead gets tagged cold and nobody ever finds out. Same logic behind the 25 free automations + 30 prompts in the Classroom. Which of your AI nodes is trusted without a check? 👇 Deploy, don't study.

The Vault just hit all 25 automations. For the next Build With Me — Live, I'll build whichever one YOU vote for, from scratch, end to end. Vote below and drop a comment with the client type you'd use it for 👇

The Vault now holds all 25 automations. I just dropped the final three inside The Vault (25 Automations) in the Classroom: 23 · Appointment reminder + no-show reducer — 24h + 2h SMS/email reminders with one-tap confirm or reschedule. Kills no-shows for any business that books time. 24 · Auto invoice payment chaser — polite, escalating reminders (Day 3/7/14) until the invoice is paid, then it stops and says thanks. Get paid without the awkward chase. 25 · Monthly client ROI report (auto) — on the 1st, it pulls each client's numbers and emails a clean one-page "here's what we did for you" report. This is the one that makes clients renew. Each comes with the full flow, build steps, message templates and the variables to swap per client — same format as the rest of the Vault. Deploy, don't study. Pick one, ship it to a client this week, and drop a comment when it's live. 👇

Contracts end. Access doesn't. Three months later their Twilio number, their inbox and their API key are still live in your workflows — and you're still paying for the tools. The build (n8n, ~20 min): → One Sheet: every client, every credential you were ever given — API keys, phone numbers, inbox access, Drive folders, paid seats → Client marked "ended" → n8n builds a dated offboarding checklist and pings you → Same run deactivates their workflows and flags which tools to cancel → 7 days later: one email to the client — their data export, and what's been switched off → Nothing ticked after 14 days → it nags you again The export email is the whole build. Handing back a clean export is what gets you the referral. Going quiet is what gets you the bad review. Same logic behind the 25 free automations + 30 prompts in the Classroom. Which ex-client's credentials are still live in your account right now? 👇 Deploy, don't study.

Freelancers don't have a pricing problem. They have a 47-days-late problem — and chasing feels rude, so it never happens. The build (n8n, ~20 min): → One Sheet: invoice number, client, amount, due date, status → Daily trigger checks anything past due → Day +3: one short reminder, friendly, invoice attached — assume it slipped → Day +10: same facts, no apology, payment link on its own line → Day +21: work pauses — stated as a policy, not a threat → Marked paid → the sequence stops immediately The escalation ladder is the whole build. One flat tone for every reminder either annoys a good client or gets ignored by a bad one. And send it from your own address, never noreply@. A reminder that looks automated gets treated like marketing. Same logic behind the 25 free automations + 30 prompts in the Classroom. Which invoice is sitting unpaid right now? 👇 Deploy, don't study.