The review request engine: more Google reviews without the reputation-suite bill
Ask every customer at the right moment, get a draft reply for every review — on a box you own
Paste it into your AI assistant — ChatGPT, Claude, Gemini, whichever you use. It first asks what tools you already have, then rebuilds this blueprint to fit them.
- A list of finished jobs with an email or mobile number, in something n8n can read — spreadsheet, booking tool or invoice app
- Your Google Business Profile review link, and acceptance that you ask every customer — filtering by expected rating is against Google's rules
- Approved Google Business Profile API access if you also want new reviews pulled in for drafting
- A mailbox or SMTP relay that can send as your domain, plus an SMS provider account if you want texts
Version history
Ask every customer for a Google review at the right moment, watch new reviews land in your inbox with a draft reply ready — for the price of the server you already run, not the €300/month a reputation suite charges.
The architecture
the same flow as text
or a "Done" button on your phone
▼
[ n8n (self-hosted) — the review engine ]
├─ same evening: thank-you email/SMS with your Google review link
├─ +7 days: one gentle nudge — skipped if they already reviewed
└─ watches the mailbox where Google's "new review" alerts arrive
▼
[ Ollama (local AI, optional) drafts a reply in your voice ]
▼
Draft lands in your inbox → you approve, paste on Google, done.
Nothing is ever posted to Google automatically.
Two deliberate design choices. First, review detection uses Google's own new-review notification emails instead of the Business Profile API — the API requires an access-request form, approval takes 3–10 business days, and your quota starts at zero until you ask for more. Fine for an agency managing fifty listings; pointless friction for one shop. Second, replies are drafted, never sent. Your name is on every public reply — the approval step is the feature, not a limitation.
Paste it into your AI assistant — ChatGPT, Claude, Gemini, whichever you use. It first asks what tools you already have, then rebuilds this blueprint to fit them.
The problem
Your Google rating is the first thing a new customer sees, and it is built by the small minority who bother to write something — which skews toward the annoyed. Happy customers leave reviews too, but almost only when asked, and asking by hand never survives a busy week. The software category that fixes this is called reputation management, and its 2026 price list is brutal for a single-location business: NiceJob from $75/month, Birdeye from $299/month on an annual contract, Podium from $399/month on an annual contract. Strip away the multi-location dashboards and the sales gloss and the core loop is: send a message with your review link after each job, nudge once, alert you when a review arrives, help you reply. That loop is an afternoon of n8n — a tool half the blueprints in this library already use — plus your free Google review link.
Tool choices — and why
Setup outline
1. Get your review link: Google Business Profile → Ask for reviews. Test it on your own phone — it should open the review box in one tap.
2. Deploy n8n (official Docker image) or reuse your existing instance.
3. Wire the trigger that matches how a job "ends" for you: webhook from your booking tool, paid-invoice event, or the phone-shortcut route.
4. Write ONE thank-you message. Short, human, from you — "Thanks for coming in today. If you have 30 seconds, a review helps us more than you'd think: [link]". Send it the same evening; goodwill decays fast.
5. Add the +7-day nudge with an "already left one? Thank you — ignore this" line, and suppress it when a review from that name has been detected.
6. Point an n8n mailbox watcher at the address that receives Google's new-review notifications; parse out rating and text.
7. Optional: pipe the review into Ollama with a short prompt describing your tone; the draft reply lands in your inbox for approval.
8. Log every ask in a simple sheet — same customer never gets asked twice, and you can see your ask-to-review rate (5–15% is normal; same-day asks do better).
Pitfalls — the real ones
Verified repos
n8n — Sustainable Use License (free self-hosted internal use), ~189k stars, active (checked 10/07/2026)
Ollama — MIT, 175k stars, active (checked 10/07/2026)
Get the next blueprint as it publishes — free, practical, verified tooling only.
Subscribe · Want this audited or extended for your business? DM @RXed_EU
What this costs to run
Priced as Draft a reply — incoming message plus tone/policy context; output is a full email. a drafted reply per incoming review. Adjust the volume to yours; the bill is a range because the assumptions are ranges.
| Model | $/day | $/month | Reasoning |
|---|
Cheapest eight shown — straight per-token
arithmetic on list prices: no caching, no batch discount, thinking tokens bill as output.
All models + every assumption in the full explorer →
Capability & cost data: Artificial Analysis