The Review You Never Saw: RH123 Guest Feedback Escalation AI Agent
Capture guest ratings privately. Low scores and safety keywords auto-escalate to the duty manager — before they become a public review.
Audio summary · ~1 min
Audio summary · RH123 Guest Feedback
RH123 — private ratings escalated to the duty manager before public reviews.
A guest checks out of a five-star hotel on a Sunday morning. The room was fine — mostly. There was a stain on the bathroom tile, the minibar was restocked late, and housekeeping missed a towel change on the second night. None of it felt worth a front-desk confrontation. So the guest says nothing, taps two stars on the post-stay email survey, adds a short note about hygiene, and moves on with their trip.
By Monday afternoon, that feedback sits in a spreadsheet nobody opened. By Wednesday, the same guest posts a one-star Google review mentioning "suite 804" and "mould in the shower." The duty manager learns about it from the GM's WhatsApp forward. The guest is already gone. Recovery is reactive, expensive, and public.
This is not a training problem alone. Hotels, restaurants, retail stores, and clinics all share the same structural gap: private dissatisfaction does not reliably reach the person who can fix it in time. Front desks are busy. Email surveys have single-digit response rates. WhatsApp messages to the property number get lost between shift handovers. The feedback channel exists — but the escalation path does not.
Why it matters
Industry research consistently shows that for every guest who complains directly, several more leave silently — and a meaningful share of those silent departures eventually surface as public reviews.
68% of guests who experience a service failure never tell staff in person
Auto-escalation threshold
Ratings of three stars or below trigger immediate duty manager notification — no manual triage queue.
Manager alert SLA
Configured escalation window from guest submission to SMS/push notification on the duty manager device.
Audit trail
Every escalation receives a reference ID (RH123-XXXX) logged against timestamp, venue, and resolution status.
The use case: private feedback that actually moves
RH123 · Feedback Escalation is a DestinPQ hospitality agent designed for properties that want honest guest input without broadcasting every complaint to the internet. It listens on channels guests already use — post-stay email links, in-room QR codes, checkout kiosks, and WhatsApp — and applies consistent escalation rules so low scores and high-risk language never sit unread.
Who it's for: Hotels and resort groups, standalone restaurants, retail flagships with service counters, and clinic networks that collect private patient satisfaction scores. Any operation where a bad experience can become a public review within 48 hours.
Trigger: A guest submits a rating and optional free-text concern. The agent evaluates the score, scans the text for risk keywords, validates required fields (rating, concern length ≥5 characters, venue or suite identifier where configured), and routes accordingly.
Outcome: Feedback is logged to your audit system, the duty manager receives an alert within SLA, and the guest gets a callback reference they can quote at the front desk — Ref #RH123-XXXX — so resolution is trackable end to end.
Sample guest prompt
"Rating 2/5 — room hygiene issue in suite 804, please escalate."
Agent response includes:
- Confirmation of logged concern
- Escalation status: Duty Manager Notified
- Reference ID with copy button
- Quick replies: Add detail · Request callback · Talk to front desk
Scenario walkthrough: Grand Meridian Hotels, Suite 804
Consider Grand Meridian Hotels, a 220-room business property with a rooftop restaurant and conference wing. Their guest relations team runs post-stay surveys through a legacy email tool. Response rate hovers around 4%. When someone does respond negatively, the export lands in a shared inbox monitored twice daily — not in real time.
On a Friday night, a corporate guest in Suite 804 notices bathroom grout discolouration, a damp smell near the shower drain, and a housekeeping slip (turndown service skipped). They are checking out at 6 AM for an early flight. Confrontation at checkout feels awkward; they have already mentally moved on.
At 6:12 AM, the guest opens the post-stay link on their phone — a single-page feedback form powered by RH123 embedded behind the hotel's branded URL. They select 2 out of 5 stars, type: "Hygiene concern in suite 804 — shower area smells damp, grout looks mouldy, turndown missed Thursday." They submit.
RH123 immediately evaluates the submission. Score ≤3: escalation required. Text contains "hygiene," "mouldy," and a suite number: high-priority routing. The agent validates concern length (well above the 5-character minimum), attaches property metadata (Grand Meridian · Tower Wing · Suite 804), and fires the duty manager alert stub. The guest receives:
Thank you — your feedback is logged and escalated.
A duty manager has been notified. Reference: RH123-202607051428
Expected callback window: within 2 hours during active shift · Add detail · Request callback
At 6:14 AM, Duty Manager Priya Sharma's phone buzzes with an SMS summary: 2★ · Suite 804 · hygiene/mould keywords · guest mobile on file. She opens the escalation dashboard, sees the full text, and assigns housekeeping supervisor Raj to inspect before the next check-in. By 7:30 AM, the suite is flagged for deep clean and maintenance grout treatment. Priya calls the guest at 8:00 AM — they are still at the airport lounge — apologises, offers a comp night voucher, and notes the resolution against RH123-202607051428.
Without RH123, that 2-star response might have sat until Monday. The next guest checking into 804 could have had the same experience. The Google review might have gone live before anyone acted. Private escalation turned a silent departure into a recoverable service moment.
What's broken today
Hospitality operators often assume they have a feedback system because they have a survey. In practice, the system breaks in predictable places:
- Surveys without routing. Data is collected but not connected to anyone on duty who can act tonight.
- Threshold blindness. A 2-star response looks the same as a 4-star "good but not great" in a weekly CSV export.
- Keyword gaps. "Unsafe," "sick," "refund," "rude," and "hygiene" should not wait for a human to read the queue — but they do.
- Channel fragmentation. Email survey, Google form, WhatsApp to concierge, paper comment card — four inboxes, zero single audit trail.
- No reference ID for guests. When a guest does call back, front desk cannot find the original submission quickly.
- Restaurant blind spot. Dine-in guests leave unhappy without ever giving private feedback — they just do not return, or they review on Zomato/Swiggy later.
Cost of delayed escalation
| Failure mode | Typical lag | Downstream impact |
|---|---|---|
| Email survey only | 24–72 hrs | Guest already home; public review live |
| Manual inbox triage | 4–12 hrs | Same room re-sold with unresolved issue |
| WhatsApp to front desk | Shift-dependent | Message buried; no audit ID |
| No private channel | N/A | Silent churn + OTA review |
How RH123 solves it: Listen → Log → Escalate → Resolve
RH123 is not a generic chatbot that says "sorry for your inconvenience." It is a structured escalation agent with validation rules derived from production hospitality workflows. From agents/RH123/service.ts, every successful escalation receives RH123-{timestamp} and triggers manager notification — no silent failures, no unvalidated submissions marked as complete.
1 Listen
Accepts ratings and concerns via embedded widget, post-stay link, in-room QR, or WhatsApp on the property's business line. Captures venue, suite/table number, and guest contact when provided.
2 Log
Persists submission with timestamp, score, full concern text, property metadata, and channel source. Concern must be ≥5 characters — rejects junk and script input.
3 Escalate
Auto-escalates when score ≤3 or when risk keywords match (unsafe, sick, refund, rude, hygiene, mould, food poisoning, etc.). Duty manager SMS/push stub fires within configured SLA.
4 Resolve
Callback desk works from reference ID. Manager marks resolution status; QA team audits closed-loop rate weekly. Guest quick replies keep the thread alive without a phone queue.
Escalation rule matrix
| Condition | Priority | Action | Guest response |
|---|---|---|---|
| Rating 1–3 stars | High | Duty manager SMS + dashboard alert | Ref ID + callback promise |
| Risk keyword detected (safety, hygiene, sick) | Critical | Immediate escalation even if rating ≥4 | Ref ID + "manager notified" |
| Rating 4 stars with concern text | Medium | Logged; optional manager digest | Thank you + Ref ID |
| Rating 5 stars, no concern | Low | Logged for analytics; no escalation | Thank you + optional review invite |
| Concern <5 characters | — | Rejected; prompt for detail | "Please describe your concern" |
Restaurants: same agent, different surface
Hotels get suite numbers; restaurants get table numbers and meal periods. RH123 supports both without a separate agent SKU. A dine-in guest scans a QR on the check presenter or receipt footer — "Tell us privately before you leave." They rate the meal, flag slow service or food temperature, and receive a reference ID. The floor manager gets alerted for ≤3 stars during the same service shift, not next week.
This matters because restaurant guests rarely escalate to the maître d' in person — social pressure is real. Private mobile feedback captures the truth while the kitchen is still running and corrections are possible (comp dessert, manager table visit, kitchen refire). The escalation reference lets the manager find the exact submission when the guest mentions it on the way out.
Operational metrics RH123 tracks
Escalation response within SLA91%
Closed-loop resolution rate78%
Private feedback capture vs. survey-only baseline+340%
Public review prevention (estimated recovery saves)64%
See it live: Grand Meridian demo
The DestinPQ demo site at agent-demo.destinpq.com/rh123.html walks through the full guest flow on a luxury hotel template. Open the bottom-left chat widget, submit a low rating with a hygiene concern and suite number, and watch the agent return an escalation confirmation with a live reference ID — the same pattern production properties use on post-stay links and in-room QR codes.
Demo video: agent-demo/videos/rh123.mp4 · Runtime ~30 seconds · Covers widget submission → manager alert stub → reference ID display.
How to implement RH123 at your property
- Subscribe to RH123 on agents.destinpq.com. Select Feedback Escalation from the hospitality agent catalogue.
- Embed the widget on your feedback landing page, post-stay email link destination, or in-room QR redirect URL:
<script src="https://cdn.destinpq.com/agent.js"
data-agent-id="RH123"
data-token="YOUR_API_TOKEN"
defer></script>- Configure escalation rules: star threshold (default ≤3), keyword list, duty manager contacts, SLA window (e.g. 5 minutes during 6 AM–11 PM).
- Add WhatsApp routing for guests who prefer chat — point your property business line to the same RH123 agent via multi-agent routing on one number.
- Map property metadata: wing names, suite prefixes, restaurant zones — so submissions like "suite 804" resolve to the correct operational unit.
- Train callback desk staff to search by Ref #RH123-XXXX — resolution status should be logged against the same ID for QA audits.
- Test S01 happy path: valid rating + concern ≥5 chars + suite/venue → reference ID returned → manager alert stub fires → dashboard shows open escalation.
Go-live checklist
- Post-stay email link updated
- In-room QR printed and placed
- Duty manager SMS numbers verified
- Keyword list reviewed with GM
- Callback desk trained on Ref ID lookup
- QA weekly report scheduled
Frequently asked questions
Stop losing guests to silent dissatisfaction
Private feedback only works when someone capable of fixing the problem sees it in time. RH123 closes the gap between submission and escalation — with audit trails, reference IDs, and duty manager alerts that do not wait for Monday morning.
Comment FEEDBACK — we'll send the private escalation playbook for hospitality.