Is 60–70% guaranteed for our hotels?

No. It is a portfolio typical for messaging after Skills and after-hours automation are live. Your rate depends on property type, channel mix, integration depth, and how many flows you keep in human-in-the-loop by brand policy.

Do FAQ answers ever count as autonomous resolution?

Only when the guest’s request was informational and no operational action was required. If the guest asked for an action and received only a KB reply, it does not count.

How is this different from “containment”?

Containment means a human did not touch the thread. Autonomous resolution means the outcome was completed (or correctly refused) under policy. Containment can include FAQ deflection; resolution cannot.

Can we verify this in a paid pilot?

Yes. Pilots should report resolution by skill, channel, and property, plus escalation rate and AI vs human CSAT on the same cohort. That is the evidence package we recommend every buyer demand.

Where can I see Skills coverage vs competitors?

On each D3x vs competitor page we publish a Skills vs FAQ automation matrix with resolves / routes / answers statuses. Use that matrix with this methodology when you score RFPs.

How D3x measures autonomous resolution

METHODOLOGY · SKILLS VS FAQ AUTOMATION

Most “80% automation” is FAQ answers, not resolved work.

D3x publishes a 60–70% autonomous resolution range on guest messaging. This page defines exactly what that means, so operators, CIOs, and procurement teams can compare vendors on the same scorecard.

Public definition · Production messaging · Updated 2026

01 · THE SCORECARD

Skills resolution vs FAQ automation

Vendors often count a knowledge-base reply as “automated.” The guest still needs towels, a folio correction, a reservation change, or a maintenance fix, and that unfinished work lands in the operator inbox. D3x Skills execute in your PMS and ops stack. That is the difference between deflecting and resolving.

Answers

FAQ automation

Matches the guest question to a KB article or scripted reply
Does not write back to the PMS, housekeeping, or finance systems
Often counted as “automated” even when staff must finish the request
Looks strong on deflection charts; weak on inbox load

Guest got text. Hotel still has work.

Resolves

Skills resolution

Understands intent against live reservation and property context
Completes the outcome inside hotel systems within policy guardrails
Logs the decision and reason for audit; escalates only when required
Removes the thread from the human queue when the work is done

Guest got the outcome. Hotel stack updated.

02 · DEFINITION

What autonomous resolution means at D3x

In plain language: the guest asked for something, the agent finished it (or correctly declined under policy), and no teammate had to take over the thread. A polite FAQ answer that leaves the operational request unfinished is not autonomous resolution.

Formula

Autonomous resolution = numerator ÷ denominator

Numerator

Guest messaging threads closed without human takeover, where the agent completed the requested outcome (or a policy-safe refusal) under Skills Engine rules.

Denominator

All guest messaging threads in scope for automation in the measured period (by property and channel), excluding spam, test traffic, and threads intentionally held out of automation.

What counts

Late checkout confirmed and written to the PMS within policy
Housekeeping or maintenance ticket created with room context in the ops system
Invoice requested, corrected, or reissued through connected finance flows
Upsell accepted and written into the reservation
Check-in / access flow completed with PMS write-back
Booking created from live rates when the skill is enabled
Policy-safe refusal that fully closes the request (e.g. late checkout denied with clear reason, no staff needed)
Lost & found logged in a structured ops workflow without human takeover

What does not count

FAQ or property-knowledge reply with no completed operational outcome
“I’ve opened a ticket for the team” handoffs (that is routing, not resolution)
Bot replies that still require a human to finish the same request
Deflection to a help centre article counted as success
Threads closed by staff after the agent failed or escalated
Spam, marketing blasts, or internal test conversations

Grey areas (how we treat them)

Human-in-the-loop approval

If policy requires a human approval before an action commits, the thread is not autonomous until the action completes without further staff ownership of the guest conversation. Approvals are governance; they are not free automation credits.

Co-pilot / draft-for-staff

Drafts that a human must send do not count as autonomous resolution. The human still owns the outcome.

Multi-intent threads

If one thread asks for towels and a folio correction, both outcomes must be completed (or correctly refused) without takeover. Partial completion with staff finishing the rest does not count as autonomous.

Warm transfer

A clean handoff with full context is good product behaviour. It is not autonomous resolution for that thread.

03 · WORKED EXAMPLE

Same guest request. Two very different “automation” outcomes.

In-stay guest on WhatsApp: “Can I have late checkout until 2pm? I have a train at 3.” Property policy allows late checkout for €35 on weekdays when occupancy allows.

Path A

FAQ automation path

01Bot matches “late checkout” to a KB article
02Replies with generic late-checkout hours and a fee range
03Asks the guest to contact reception to confirm
04Thread sits in the inbox for a human to check occupancy, charge the fee, and update the PMS

Counted as “automated” by many vendors. Work still open.

Path B

Skills resolution path

01Agent reads the live reservation and occupancy rules
02Offers the €35 late checkout within policy
03Guest accepts in-thread
04PMS updated, fee applied, confirmation sent, decision logged with reason

Counts as autonomous resolution. Inbox empty for this request.

04 · THE 60–70% RANGE

Where the published range comes from

Portfolio typicals on guest messaging run 60–70% autonomous resolution once Skills are enabled and after-hours automation is live. The range is not a guarantee for every property on day one. It reflects production portfolios after rollout, with high-stakes flows kept in co-pilot or human-in-the-loop where brand policy requires it.

60–70%

Autonomous resolution

Portfolio typical · messaging

250K+

Messages / month

Production volume across live groups

4.58

AI CSAT

vs 4.62 human · 90-day sample

68%

McDreams automation

Published case · ~2 of 3 messages

What moves a property inside the range

Property type and ask mix (hostels and aparthotels often resolve more routine volume than ultra-luxury with tight VIP policies)
Channel mix (WhatsApp and web chat usually outperform messy shared email until email skills are fully enabled)
How aggressively after-hours and overflow automation is turned on
Integration depth: PMS + housekeeping + invoice write-backs raise true resolution; FAQ-only configs suppress it
Policy posture: more human-in-the-loop approvals lower autonomous rate by design, which can be correct for the brand

05 · BUYER AUDIT

Five questions every hotel AI vendor should answer

Bring this scorecard to every demo. If a vendor cannot answer with production evidence, treat their “80% automation” claim as FAQ deflection until proven otherwise.

01
Do you write back to our PMS and ops systems, or only answer from a knowledge base?
Write-backs are the line between resolution and chat.
02
Is a FAQ reply counted as resolved in your automation rate?
If yes, you are comparing different units. Ask for a Skills / stack-resolution rate separately.
03
Can we see resolution broken down by skill, channel, and property?
Portfolio averages hide FAQ-heavy channels and weak ops skills.
04
What are escalation rate and AI CSAT on the same cohort as the automation number?
High automation with collapsing CSAT is not a win.
05
Is every autonomous action logged with reason for audit?
Enterprise groups need decision trails, not black-box deflection stats.

06 · GLOSSARY

Words vendors overload, defined tightly

Autonomous resolution: Thread closed without human takeover because the requested outcome was completed (or correctly refused) under policy, with stack updates where required.
Automation rate: Vendor-dependent. Often includes FAQ deflection. Always ask what sits in the numerator.
Deflection / containment: Guest did not reach a human. Useful, but not the same as work finished in hotel systems.
Routing: Agent creates a ticket or hands the thread to staff. Good triage; not resolution.
FAQ / answers: Knowledge reply only. Counts as service quality only when no operational action was requested.
Skills Engine: D3x’s codified hospitality logic above the LLM: policies, PMS quirks, and executable skills that turn intent into stack actions.

07 · LIMITS & HONESTY

What this methodology does not claim

Trust beats a vanity percentage. We would rather publish a lower autonomous rate on a tightly governed luxury portfolio than inflate the number with FAQ deflection.

The 60–70% range is a messaging portfolio typical, not a contractual SLA for every site in week one
Voice and email can differ from messaging until those channels run the same skills at the same depth
Luxury and VIP-heavy policies often suppress autonomous rates on purpose
Early rollouts start lower while skills, integrations, and after-hours rules are tuned
Reservation-change automation that is still on the product roadmap is not counted as live resolution today

08 · FAQ

FAQ

Book demo See competitor scorecards

SEE IT ON YOUR VOLUME

Bring your last 90 days of inbox volume.

We will map which requests become Skills resolutions on your PMS and ops stack, and which should stay human by policy.

Book demo