GEST in Action: How an IoT Service-Call System Cuts Hotel Response Time and Lifts Guest Satisfaction

Walk into any five-star hotel and look at the bedside table. There is almost certainly a phone with twelve buttons whose labels have faded into illegibility. There is a Do Not Disturb tag hanging from the door handle, paper, easy to miss, often left up by accident long after the guest has gone out. There is a service menu in a leather folder somewhere, three layers down. None of this is how a guest who paid 800 EUR a night wants to ask for a fresh towel.

GEST is the system we built to replace all of it. A small wireless button on the bedside, in the bathroom, by the dressing area. One press calls housekeeping. A double press flags it as priority. The request lands on a staff tablet within seconds, routed to the right department, tracked against an SLA, and logged for analytics. No phone calls, no door tags, no guesswork. This piece walks through the architecture, the deployment story, and the ROI math we have validated in luxury properties.

The Problem with the Status Quo

In-room phones were designed for a 1990s service model. They assume the front desk has time to answer, that the guest knows which button reaches housekeeping, and that the language barrier is small. None of these assumptions hold in a 200-room property with a multinational guest list and a lean overnight team. The result is measurable: average response time for a non-urgent housekeeping request in a typical luxury hotel is between 14 and 22 minutes from the moment the guest decides to ask. About 18 percent of those requests are abandoned before anyone arrives, surfacing later as complaints in the post-stay survey.

The Do Not Disturb tag is even worse. It carries a single bit of information, no timestamp, and no record. Housekeepers walk a corridor, see five tags, and have to make a manual note to come back later. Some never get checked, especially during shift changes. The system loses information at every step.

The fix is not a fancier phone. It is to make the request itself a structured digital event from the first moment. That is the design principle behind GEST and the broader category of hotel IoT service-call systems.

GEST Architecture at a Glance

The system has four layers, each chosen for a specific operational reason.

1. The button

A custom enclosure around an ESP32 module, two CR123A lithium cells, a tactile dome switch, and a small RGB LED for confirmation. The radio is configurable per deployment: Wi-Fi when the property has dense AP coverage and the IT team is comfortable joining devices to a dedicated SSID, LoRa when the building has thick stone walls or the customer wants to avoid the IT politics of new SSIDs entirely. Battery life is between 4 and 7 years on Wi-Fi (deep sleep, wake on press) and over 10 years on LoRa.

The enclosure is brushed aluminium, available in finishes that match common hotel hardware (champagne, matte black, polished nickel). The button is mounted with a single screw and a 3M VHB pad for retrofit, or recessed into millwork for new builds.

2. The gateway

For LoRa deployments, one gateway covers a typical full floor of a hotel. For Wi-Fi deployments, the existing AP infrastructure is the gateway and no extra hardware is needed. Gateways forward press events over MQTT to the GEST cloud, hosted on Azure.

3. The cloud

Azure IoT Hub ingests events. A routing engine maps button ID to room, room to floor, request type to department, and current shift to the correct staff member. The escalation timer starts immediately. Every state transition is written to an event store for analytics. Read more on the cloud platform we use across products.

4. The staff app

Native iOS and Android apps for housekeeping, F&B, and front office staff. Push notifications, room-level acknowledgment, completion timestamping, and a manager dashboard. Built on the same patterns we use for our other mobile control products.

Button Placement and UX

The hardware is the easy part. The UX is where most service-call deployments fail. Three principles drive the placement and gesture design.

• One button, layered gestures. A single short press is the default service request. A double press flags it as priority. A long press (held for two seconds) cancels a request that was triggered by accident. This is enough vocabulary for 95 percent of requests and avoids the cognitive load of a labeled keypad.
• Context per location. A button next to the bed defaults to housekeeping. A button in the bathroom defaults to towels and amenities. A button in the minibar area defaults to F&B. Same hardware, different routing rules in the cloud, configured per room type.
• Confirmation feedback. The LED pulses green when the cloud confirms receipt, amber while staff is en route, and solid green for two seconds when staff acknowledges. The guest never wonders if anything happened.

For the most premium properties, we offer a touchless variant: a capacitive sensor under a polished surface, with a haptic feedback pulse, so the guest never sees a visible button at all. The hardware cost is roughly 40 percent higher and we only recommend it for ultra-luxury where invisibility is a brand requirement.

Mapping Requests to Workflows

The cloud-side rule engine is where operational knowledge becomes code. A typical configuration for a 200-room property contains around 40 rules covering:

• Department routing by request type and time of day (housekeeping daytime team versus night turn-down)
• Load balancing across available staff (round-robin within the responsible team)
• Floor proximity (prefer the housekeeper already on the same floor)
• Guest VIP status pulled from the PMS (route VIP requests to senior staff)
• Language preference (route to a staff member who speaks the guest’s preferred language when known)

Rules are versioned and tested in a staging environment before they go live. Hotels are sensitive about changes that affect guest-facing service, and we treat rule changes the same way we treat firmware updates.

Escalation and SLA Tracking

Every request carries an SLA target. Standard housekeeping is 8 minutes. Priority is 4 minutes. F&B in-room is 12 minutes. Engineering response is 15 minutes. The system tracks four state transitions: created, acknowledged, en route, completed. If acknowledgment does not happen within 60 seconds, the request escalates to the shift supervisor. If completion does not happen within the SLA, it escalates to the duty manager and is flagged in the daily report.

The escalation rules are not punitive. They are observability. The supervisor needs to know that a request is sitting unacknowledged because the housekeeper is in a deep-clean of another room and cannot answer. That information lets the supervisor reroute, not blame.

Integration with PMS and Housekeeping Software

GEST is not an island. It integrates with the Property Management System (Opera, Mews, Protel, Apaleon are the most common) to enrich every request with guest context: name, language, VIP status, special preferences, stay history. It integrates with housekeeping management platforms (HotSOS, Knowcross, Quore) so completion events feed back into the room-status workflow.

Integration patterns we have shipped:

• OAuth-based REST polling for PMS systems with a documented API
• HL7-style flat-file drops for older PMS systems that only export overnight reports
• Webhook out to housekeeping software on completion events
• Direct PMS room status updates over OPERA’s OXI interface

For more on how we approach this layer, see our overview of third-party integrations.

Analytics: Heatmaps and Patterns

Once a hotel has been on GEST for sixty days, the analytics start to show patterns nobody knew were there. The dashboards we ship as standard include:

• Request type by floor and time — surfaces the fact that floor 7 generates 3x more towel requests than other floors (the gym is on the same elevator core)
• Response time distribution by shift — reveals that the 22:00 to 02:00 shift consistently misses SLA on the high-floor suites because of elevator wait times
• Repeat requests by room — identifies rooms with maintenance issues (a leaking shower generates daily towel requests)
• Guest satisfaction correlation — correlates response time with post-stay survey scores at the individual stay level

This is the same analytics platform we use across our products, described in more detail under analytics dashboards. The data also flows into our broader AI workstreams for anomaly detection and demand forecasting.

ROI: A Hypothetical 200-Room Luxury Hotel

Numbers below are conservative averages from comparable properties. Treat them as a planning model, not a guarantee.

Inputs

• 200 rooms, 80 percent average occupancy, 365 nights/year → 58,400 occupied room-nights
• Average 1.4 service requests per occupied room-night → 81,760 requests/year
• Pre-GEST average response time: 17 minutes
• Post-GEST average response time: 4.5 minutes (validated across our deployments)
• Average ADR: 520 EUR
• Pre-GEST guest satisfaction score (NPS proxy): 58

Hard savings

Staff time saved per request: roughly 6 minutes (eliminated phone handling, walk-back-and-forth from a partial request, supervisor escalation). Across 81,760 requests that is 8,176 staff-hours per year. At a fully loaded rate of 22 EUR/hour for housekeeping staff, that is 179,872 EUR per year.

Soft revenue lift

NPS lift in our deployments averages +14 points within six months. Industry research correlates each 10-point NPS increase with roughly 1.2 percent revenue lift in luxury hospitality through repeat bookings and referrals. Applied to 200 rooms at 520 EUR ADR and 80 percent occupancy: 30,368,000 EUR annual room revenue, 1.68 percent lift = 510,182 EUR. Even discounted heavily for attribution uncertainty, the soft revenue is the larger line item.

Investment

Hardware and installation for 200 rooms with 3 buttons per room: 90,000 EUR one-time. Cloud and software subscription: 24,000 EUR per year. Year-one total: 114,000 EUR.

Payback

Hard savings alone pay back the investment in around 7.6 months. With the soft revenue lift included, payback drops below 3 months. From year two onward, the operating cost of the system is roughly 13 percent of the hard savings it generates.

Security and Privacy

Hotels are high-value cyber targets and any in-room device gets scrutinized. The GEST design starts from three commitments:

• The button captures no personal data. It transmits a button ID, a press type, and a timestamp. The mapping to a guest happens in the cloud against the PMS, never on the device.
• All transport is TLS. Wi-Fi devices use WPA2-Enterprise on a dedicated SSID. LoRa devices use AES-128 with per-device keys provisioned at manufacture.
• Cloud is GDPR-compliant by design. Personal data is purged 30 days after checkout. Aggregated analytics are retained indefinitely.

For properties that also deploy smart access control (Nuvlock or third-party), GEST integrates so that a do-not-disturb gesture at the bedside also temporarily blocks housekeeping access at the door, eliminating the corridor-tag failure mode entirely.

Deployment Timeline for a 200-Room Property

A typical deployment runs 8 to 10 weeks from contract to go-live.

1. Weeks 1-2: Site survey, RF assessment, room-type mapping, finish selection, PMS integration scoping.
2. Weeks 3-5: Hardware production and pre-provisioning. Cloud tenant setup, rule engine configuration, staff app branding, PMS API testing in sandbox.
3. Weeks 6-7: Pilot floor installation (typically a 20-room floor), staff training, two-week supervised pilot.
4. Weeks 8-9: Full property rollout, two rooms per installer per hour. Property-wide commissioning over a long weekend with no guest disruption.
5. Week 10: Go-live, daily review for the first 14 days, handover to operations.

The pilot floor is non-negotiable. It catches the 5 to 10 percent of rules that look right on paper but produce surprising routing in practice (the night manager who actually handles all engineering escalations after midnight, the F&B team that does not start until 06:30 but inherits 05:00 requests).

Where GEST Goes Next

The roadmap items most often requested by hotel customers are voice activation (a discreet always-on microphone with on-device wake-word detection, no audio leaving the room), proactive offers triggered by inferred guest behavior (a coffee request at 06:55 every morning becomes a one-tap recurring schedule), and tighter integration with in-room A/V control. The voice and behavioral pieces are where our AI capabilities earn their keep, and we use the same guest-data hygiene principles end to end.

Make the First Move on Guest Experience

Hotels compete on a thousand small details. A 12-minute wait for a fresh towel is a detail every guest remembers, and one that no operations manual can fix without the right plumbing underneath. GEST provides the plumbing: the buttons, the routing, the dashboards, the integration with the systems your team already uses. The ROI is honest, the deployment is non-disruptive, and the operational lift is visible in the first week.

FSS designs and operates GEST end to end, from the PCB to the staff app. Talk to us about a pilot on one floor of your property and we will share the data from comparable luxury deployments under NDA. Start with our IoT solutions overview or go straight to the connected devices service page.