Routing rules.
Routing decides what happens to each review after the AI classifies it. The defaults are conservative — you can adjust everything in Settings → Routing Rules.
The decision flow
For each review, the AI computes a sentiment score (1–10) and checks for legal keywords. Then:
- Legal language? → Routed to legal escalation. No AI reply is drafted. Email + SMS alert is sent to your legal escalation contact.
- 1–3★ rating? → Routed to human queue. You review and reply manually.
- High score (4+) and positive? → Auto-post, subject to your delay window.
- Anything else → Human queue.
Routing toggles
Three toggles in Settings → Routing Rules:
- Auto-post 4–5★ responses — when off, every reply routes to the human queue regardless of rating.
- Route 1–3★ to human queue — always on by default; 1–3★ reviews are never auto-posted.
- (Third toggle — business-hours-only, see below)
Auto-post delay
Choose a buffer between when the AI generates a reply and when it actually posts. Common settings:
| Delay | Use case |
|---|---|
| 2 hours | Default — gives you time to catch a draft you don’t like |
| 4 hours / 8 hours | More flexibility to review before posting |
| 24 hours | Conservative — lets you check yesterday’s replies each morning |
Business-hours-only
Toggle this on and the AI will post replies only during business hours (Mon–Fri, 09:00–17:00 in your account time zone). If a reply would otherwise post at 11 pm Friday, it snaps forward to 09:00 Monday. Customers don’t see midnight replies that look automated.
Legal escalation
If a review contains words like lawyer, sue, fraud, scam, lawsuit, court, damages, false advertising, the system stops, does not draft a reply, and immediately:
- Sends an email to your legal escalation contact
- Sends an SMS to your legal escalation phone
Legal-flag alerts route to your Contact email and Contact phone in Settings (top nav) → Business contact info — the same fields that drive the CASL sender footer on outbound emails. Set both to people who can actually act on a legal-language review (your lawyer, founder, or ops lead — not a shared inbox nobody watches). If the fields are blank, the alert falls back to Tenlo support, not you.
The flagged review appears in the Reviews tab under the Legal filter. If it’s a false positive, click Dismiss escalation — the flag clears, status resets and the AI generates a normal draft. If it’s a real legal threat, click Reject review and handle the matter outside the platform.