Review-Response Agent (Automation Blueprint)
A build-it-once automation blueprint that connects to YOUR own Google Business Profile, Trustpilot or Facebook feed, watches for new reviews, and drafts an on-brand reply to every one — star-rating aware, de-escalating on the bad ones — then parks each draft for your one-click approval. It never auto-posts. You get the full build spec for BOTH n8n and Make.com, the exact reply prompts, a field-mapping table and a test checklist, so you can stand it up on your own accounts and keys in an afternoon.
There are two ways to stop reviews going unanswered: write every reply yourself, forever, or build something that writes the first draft and waits for you. This blueprint is the second option — A$79 one-off for a written specification of an agent that watches your Google Business Profile, Trustpilot or Facebook feed, drafts an on-brand reply to each new review, and holds that draft for your approval. It is a plan rather than a running service. The file arrives on your order page as a Markdown download in about 15 minutes, and the flow itself gets built on your accounts, by you.
Nothing reaches the public without your click
The agent never posts. Every reply it writes stops at an approval step, and the blueprint maps that path node by node: a new review arrives, the AI step drafts a response, the draft lands in your approval inbox, and it only goes live to Google, Trustpilot or Facebook when you release it. You also nominate an escalation contact at intake, so the spec is written knowing which human the difficult ones should reach.
The honest catch is that you are the one wiring this together. The draft-for-approval rule holds because the specification tells you to build it that way — connect the AI step straight to the publish action instead and the safeguard is gone. Follow the flow as written and you stay the publisher of every word.
A one-star review and a five-star review need different words
The pack includes a paste-ready prompt set keyed to star rating, with separate prompts for the one-star through five-star scenarios. The low-rating prompts are written to de-escalate rather than argue, and how far they go on your behalf depends on the answer you give at intake about handling negative reviews. The five-star prompts cover the replies that are easy to write and easiest to never get around to.
These are prompts, not a sentiment engine. They instruct the AI step inside your own automation, so what comes out reflects the model you connect and the tone brief you set.
n8n or Make.com, running on your own keys
You get the full build spec for both platforms — nodes and modules, step by step — so you can pick whichever you already use rather than being locked to one vendor. The flow runs on your logins and your API keys throughout. There is no ECTD hosting layer, no monthly fee to us, and nothing to cancel later.
The field-mapping table shows which review data feeds which step, and the five-minute test checklist plus troubleshooting FAQ cover the part people get stuck on: proving it fires before you trust it with live reviews. Three free AI revisions are included, which is usually enough to re-cut the tone brief once you have read the spec properly.
What you get
- Full build spec for n8n AND Make.com (nodes/modules step by step)
- Field-mapping table: review source → AI step → approval inbox
- Star-rating-keyed reply prompt set (1★ to 5★ scenarios), paste-ready
- "Draft for approval, never auto-post" safety rule baked into the flow
- "Test it in 5 minutes" checklist + troubleshooting FAQ
- Runs entirely on your own accounts and API keys — you own it outright
- 3 free AI revisions included
Turnaround: ~15 minutes. Delivery: Automation Blueprint — Markdown download (n8n + Make.com specs).
Frequently asked questions
My reviews aren't on Google — will this still work?
The blueprint is written for Google Business Profile, Trustpilot or Facebook, and you nominate one of them at intake. If your reviews sit somewhere else, this is not the right product as specified. The flow depends on the review source exposing a feed your automation platform can read, and we cannot promise that for every review tool on the market.
What lands on my order page — a working automation or a document?
A document. The Markdown file contains the build spec for n8n and Make.com, the star-rating reply prompts, the field-mapping table and the test checklist. Nothing is connected on your behalf, nothing runs on our infrastructure, and no part of this is a managed service.
Beyond the A$79, what will this cost me to run?
Whatever your n8n or Make.com plan costs and whatever your AI model usage costs, both billed to you directly by those providers. The A$79 is one-off and there is no subscription to us.
How much building am I actually signing up for?
Enough to need a comfortable afternoon in the n8n or Make.com editor connecting your own accounts. The spec walks the nodes and modules step by step and the troubleshooting FAQ covers the common snags, but this is not a one-click install.
The tone comes back wrong for my business — then what?
Three free AI revisions are included. The usual fix is re-cutting the tone brief or the negative-review instruction from your intake and having the spec and prompts rewritten against it.
Included in Tradie AI Pack — save with a bundle.