This page covers the second of the two assessments: whether AI systems and search engines can find your business, verify what is true about it, cite it correctly, and let a guest book directly with you.
This assessment is about how your business appears to the outside world and to the AI systems now sitting between a guest and a booking. Whether your name, address and hours match everywhere. Whether an AI assistant asked to book you would reach your reservation system or a reseller. Whether the facts about your rooms, rates and policies are published in a form a machine can read and trust.
If you do not control your own domain and CMS, the team that does controls the website changes that affect AI visibility. The assessment still matters: take the same questions to your brand, group or agency and record who owns each capability.
01
Step
Prepare the digital session
Run this assessment as a facilitated evidence review with the people who manage the website, search presence, structured data and direct-booking path.
The questions are written to be asked aloud with follow-ups. A broad answer such as “the website is optimised” cannot be scored. A useful answer names the page or system, explains what was checked and when, identifies who can change it, and gives a real example of something working or failing.
The conversation also surfaces disagreements. If marketing believes a fact is updated everywhere on the same day but the web team says site changes are released monthly, that gap is a finding. Do not average the two accounts; keep both and resolve the evidence.
How to run it
Book one 90–120-minute session, or two shorter sessions. Bring the three or four roles that hold the evidence; an agency or corporate web counterpart may be one of them.
Work through the questions in order. Let the relevant owners join for their section rather than keeping everyone for the whole session.
Nominate one person to capture the evidence and, for each question, whether control sits with the property, corporate, an outsourced provider or a mix.
The 90–120-minute duration is provisional until the digital assessment has been tested in pilots.
XLSXReusable download
Digital readiness workbook
What it is
A structured workbook containing all 43 digital questions, the scoring key and an unscored responsibility-routing record.
When to use it
Use it whether the domain is controlled locally or by a brand, group or agency.
What you enter or decide
Record public and system evidence, the responsible party and the lower supported score band.
What it produces
A digital-readiness percentage calculated from accepted scores, evidence coverage and flagged input issues. A person interprets the findings and who must act.
The workbook does not crawl a website or independently verify a provider claim.
The workbook contains all 43 digital questions, the digital scoring key and an unscored routing record. Set the route to domain controlled; do not answer or score a separate gate.
02
Step
Bring the people who hold the evidence
This half concentrates in three or four roles rather than spreading across the property. Where a function is outsourced — and web, SEO and content very often are — bring the agency or the head-office team into the session, or send them the questions in advance.
Who
Roughly
What they will be asked about
IT, web, or whoever runs your website
~20 questions
Crawler access, page performance, structured data, HTTPS, booking links, whether anything blocks a machine from reaching your real content
Marketing
~22 questions
How the business appears online, guest stories and testimonials, hub pages, consistency of facts across channels, outside recognition
Revenue and distribution
~7 questions
Direct booking integrity, whether every room type and rate has a working direct path, whether an AI-assisted booking would be attributable
Owner, MD or GM
~6 questions
Accountability for AI visibility, who is responsible for checking what AI says about the business, named contacts and whether they are current
HR
~2 questions
Whether contact information and employee listings on external platforms are kept current
Counts overlap because many questions name two roles; they add to more than 43 for that reason. The point is who to invite, not an exact split.
Use the website, marketing, distribution and corporate roles named for each question. The person at the property need not guess what the provider’s systems do.
03
Step
Capture digital evidence
This is the part that determines whether the result is worth anything.
Record and transcribe it
Record the session and get a transcript if you plan to use the scoring prompt or GPT. A facilitator cannot reliably trace 43 answers, evidence details and disagreements while also leading the conversation.
The transcript becomes the evidence input for scoring. Whether a person or a language model scores it, use what was actually said rather than a later summary.
Tell people they are being recorded
Say at the start what the recording is for, who will see it and how it will be handled. Obtain any agreement required for the people and location involved.
Explain that the recording is there to capture the website and booking evidence accurately, not to assess individual performance.
Answer with specifics. Name the page, platform, feed, booking path or validation method; say when it was checked; describe what actually happens; and identify who can change it. General descriptions cannot be scored by a person or a machine.
Not scoreable
“Does every room type and rate plan link to a live direct booking path?”
“Yes, our booking engine is integrated.”
Scoreable
“Does every room type and rate plan link to a live direct booking path?”
“We tested all twelve room and rate links last week. Ten opened the live booking engine; two package links redirected to stale inventory. The revenue manager logged both for the web agency.”
The second answer identifies the test, date, result, failure and responsible party. The first does not contain enough evidence to place against a band.
A few rules the scoring follows
Score what is true today, not what is planned. “We're planning to” describes the current state.
The weakest evidenced part of the answer sets the score. If the homepage is complete but room pages are not, the assessment reflects the gap.
Adjectives are not evidence. “Optimised”, “seamless” and “fully integrated” carry no information. Named pages, systems, checks, dates and results do.
Where an answer sits between two levels, use the lower band unless stronger evidence supports the higher one.
Record who controls it
For every question, note whether the relevant website or channel capability is controlled by the property, decided at corporate level, run by an outsourced provider, or shared.
A low score on something you control is a task. The same score on something controlled by a brand or agency is a conversation with that party. The report should never tell a property to change something it cannot access.
If you cannot answer something
Leave it unscored and identify the role that can supply the evidence. Unanswered questions are removed from the denominator, never counted as zero.
If a question genuinely does not apply to the business model, mark it not applicable. Do not guess to fill a blank.
All 43 digital questions
Trustworthiness · 3 questions
DIG-01
If someone searched your business name right now, would they find one page with your address, phone number and hours — all matching exactly across your site, Google, and social profiles?
Asked of: Marketing / Front Office
Reports to: Trustworthiness · Digital readiness
DIG-02
Does your website run on a secure, up-to-date connection (the padlock / HTTPS) with no browser warnings on any key page?
Asked of: IT / Web
Reports to: Trustworthiness · Digital readiness
DIG-41
Is that contact information actually kept current — verified and updated when someone leaves or changes roles — or could it still point to someone who hasn't worked there in years?
Asked of: Leadership / HR
Reports to: Trustworthiness · Digital readiness
Trustworthiness / AI Readiness (AEO) · 1 questions
DIG-03
If a guest asked an AI assistant to book directly with you, would they complete that booking with you — or get routed to a third-party site instead?
Asked of: Revenue / Marketing
Reports to: Trustworthiness / AI Readiness (AEO) · Digital readiness
Expertise · 2 questions
DIG-04
Can you point to a page that clearly names who founded or leads the business, with real photos and backgrounds — not just a stock 'Our Team' graphic?
Asked of: Marketing / Leadership
Reports to: Expertise · Digital readiness
DIG-05
Does your site explain, in your own words, why a guest should trust your expertise over a competitor's — not just list amenities?
Asked of: Marketing
Reports to: Expertise · Digital readiness
Experience · 2 questions
DIG-06
Do you have guest stories or testimonials online that name a real, checkable guest or event — not just a star rating?
Asked of: Marketing
Reports to: Experience · Digital readiness
DIG-07
If an AI assistant were asked 'has anyone actually stayed here, and what was it like,' could it find a specific, detailed account on your own site to answer from?
Asked of: Marketing
Reports to: Experience · Digital readiness
Authoritativeness · 2 questions
DIG-08
Has your business, or anyone on your leadership team, been mentioned or quoted by a reputable outside source — press, industry publication, podcast — in the last year?
Asked of: Leadership / Marketing
Reports to: Authoritativeness · Digital readiness
DIG-09
Do other reputable websites (not directories you paid for) link to or reference yours?
Asked of: Marketing
Reports to: Authoritativeness · Digital readiness
AI Readiness (AEO) · 7 questions
DIG-10
If an AI assistant answered a question about your business today — address, amenities, pricing, policies — are you confident it would be accurate, or could it be pulling outdated or third-party information?
Asked of: Marketing / Leadership
Reports to: AI Readiness (AEO) · Digital readiness
DIG-16
Does your website use structured data (schema markup) at all, or is it relying purely on visible page text for AI and search to read?
Asked of: IT / Web
Reports to: AI Readiness (AEO) · Digital readiness
DIG-17
Is that structured data on every page that needs it — rooms, offers, policies, reviews — or only on the homepage?
Asked of: IT / Web
Reports to: AI Readiness (AEO) · Digital readiness
DIG-18
Does your structured data cover the full set of facts a guest or an AI assistant would actually need — pricing, amenities, cancellation policy, accessibility — or just the basics like name and address?
Asked of: IT / Web / Marketing
Reports to: AI Readiness (AEO) · Digital readiness
DIG-24
If a guest books after an AI assistant recommended you, would you know that's why they booked — or would it just look like ordinary website traffic?
Asked of: Revenue / Marketing
Reports to: AI Readiness (AEO) · Digital readiness
DIG-25
Do you have a distinctive offering — a special floor, a signature program, a members-only amenity — that only exists in a PDF, an old press release, or a scanned brochure, rather than as real text on a live page?
Asked of: Marketing / Web
Reports to: AI Readiness (AEO) · Digital readiness
DIG-26
If you handed someone every PDF and press release on your website, would they find facts about your business that aren't published anywhere else — facts AI can't reliably read out of a PDF?
Asked of: Marketing / Web
Reports to: AI Readiness (AEO) · Digital readiness
AI Readiness (AEO) / Sovereignty · 1 questions
DIG-11
Does your business publish its own facts in a form AI systems can read directly — ideally in one place covering every core entity (rooms, offers, policies, locations) — or does AI have to guess by scraping your web pages?
Asked of: IT / Marketing
Reports to: AI Readiness (AEO) / Sovereignty · Digital readiness
Sovereignty / Consistency · 5 questions
DIG-12
If two different AI assistants were asked to describe your business today, do you know whether they'd say the same thing?
Asked of: Marketing
Reports to: Sovereignty / Consistency · Digital readiness
DIG-21
If your business facts change tomorrow — a price, a policy, an amenity — how long before that change is reflected everywhere AI and search look?
Asked of: Marketing / Revenue
Reports to: Sovereignty / Consistency · Digital readiness
DIG-33
If you compared every page on your site that mentions the same fact — a rate, a policy, a fee — would you be confident they all say exactly the same thing?
Asked of: Marketing / Web
Reports to: Sovereignty / Consistency · Digital readiness
DIG-42
Have you done a full review of your highest-priority facts — name, address, phone — across every channel they appear, not just your own website and main social profiles?
Asked of: Marketing
Reports to: Sovereignty / Consistency · Digital readiness
DIG-43
Does that review include LinkedIn — both current employees whose profiles list your business, and former employees whose profiles still do?
Asked of: Marketing / HR
Reports to: Sovereignty / Consistency · Digital readiness
Governance / AI Readiness (AEO) · 1 questions
DIG-13
Is there anyone responsible for checking what AI tools say about your business, and is that a one-time check or something done on a regular schedule?
Asked of: Leadership
Reports to: Governance / AI Readiness (AEO) · Digital readiness
Technical Foundations · 4 questions
DIG-14
When did you last confirm that AI crawlers — not just Google's search bot — can actually access and read your website?
Asked of: IT / Web
Reports to: Technical Foundations · Digital readiness
DIG-15
Do you know how fast your key pages load, and would that pass a basic technical health check?
Asked of: IT / Web
Reports to: Technical Foundations · Digital readiness
DIG-22
Do you know which AI bots and search crawlers have actually visited your site in the last 30 days?
Asked of: IT / Web
Reports to: Technical Foundations · Digital readiness
DIG-23
Could a language selector, age gate, location picker, or cookie banner be stopping a crawler from ever reaching your real content?
Asked of: IT / Web
Reports to: Technical Foundations · Digital readiness
Content Architecture · 2 questions
DIG-19
Do you have dedicated 'hub' pages for each core part of your business — for example Accommodations, Amenities, Dining, Meetings & Events, Offers, Pet-Friendly, Local Area — or is that information scattered across the site?
Asked of: Marketing / Web
Reports to: Content Architecture · Digital readiness
DIG-20
Do your related content pages actually link back to those hub pages, so a reader (or an AI crawler) can find the full picture from any entry point?
Asked of: Marketing / Web
Reports to: Content Architecture · Digital readiness
Direct Booking Integrity · 5 questions
DIG-27
Does every room type and rate plan on your site link straight through to a live, bookable reservation — not a 'call to book' notice or a redirect to a third-party site?
Asked of: Revenue / IT
Reports to: Direct Booking Integrity · Digital readiness
DIG-28
If your inventory or pricing changes in your reservation system today, how long before your website's booking links reflect that — same day, or whenever someone remembers to update the site?
Asked of: Revenue / IT
Reports to: Direct Booking Integrity · Digital readiness
DIG-30
Does every bookable entity on your property — every room type, package, and experience — have its own working direct booking path, or only your flagship rooms?
Asked of: Revenue
Reports to: Direct Booking Integrity · Digital readiness
DIG-31
If someone clicked 'book now' on every single page of your site today, would every one of those links actually lead to a real, current reservation — not a dead link or stale inventory?
Asked of: IT / Web
Reports to: Direct Booking Integrity · Digital readiness
DIG-32
Of the bookings you count as 'direct' today, are you 100% certain none of them actually originated from an OTA or metasearch click — not just landed on your site, but started there?
Asked of: Revenue / Finance
Reports to: Direct Booking Integrity · Digital readiness
Trustworthiness / Direct Booking Integrity · 1 questions
DIG-29
If you searched your own hotel's name plus 'book now' online, are you confident the top results lead guests to your real reservation system — not a lookalike site or a reseller charging a markup?
Asked of: IT / Marketing
Reports to: Trustworthiness / Direct Booking Integrity · Digital readiness
Freshness · 1 questions
DIG-34
Does anything near your prices, hours, or policies tell a reader when that information was last confirmed true — or is everything undated?
Asked of: Marketing / Web
Reports to: Freshness · Digital readiness
Completeness (Negative Space) · 1 questions
DIG-35
For the questions guests ask most often that you'd have to answer 'no' to — no pool, no pets over a certain size, no early check-in — is that 'no' actually published somewhere, or does your site just stay silent on it?
Asked of: Marketing / Front Office
Reports to: Completeness (Negative Space) · Digital readiness
Third-Party Corroboration · 1 questions
DIG-36
For the facts that matter most about your business, would an independent source — press, a booking platform, a government listing — say the same thing you say about yourself?
Asked of: Marketing
Reports to: Third-Party Corroboration · Digital readiness
Structured / Prose Consistency · 1 questions
DIG-37
If you checked your structured data against your visible page text for the same fact — check-in time, price, an amenity — would the two actually agree with each other?
Asked of: IT / Web
Reports to: Structured / Prose Consistency · Digital readiness
Information Gain · 1 questions
DIG-38
Is there anything on your site that only your business could truthfully say — a specific number, a named partnership, a dated result — or could every sentence be copy-pasted onto a competitor's site unchanged?
Asked of: Marketing
Reports to: Information Gain · Digital readiness
Accountability · 1 questions
DIG-39
For the facts a guest could be genuinely harmed or disputed by — allergen handling, accessibility, cancellation exceptions — is there a specific person or policy behind that claim, or is it just an anonymous 'we'?
Asked of: Leadership / Marketing
Reports to: Accountability · Digital readiness
Trustworthiness / Expertise · 1 questions
DIG-40
Can a guest find real names and direct contact information for the people behind your business — ownership, service leads, customer support — not just a generic 'info@' address?
Asked of: Leadership / Marketing
Reports to: Trustworthiness / Expertise · Digital readiness
04
Step
Score only what the evidence supports
Every question is scored from 0 to 10 against the same five levels, so the number reflects what is actually happening rather than anyone's impression of it.
Level
What it means in practice
Not in place
It does not exist, and it has not been considered.
Early or ad hoc
One person does it informally. Nothing is written down.
Developing
Some structure exists, but it is not consistent, connected or measured.
Established
Structured, consistent, and delivering results. Room to optimise.
Leading
Best in class: measured, largely automated, improving continuously.
Those roll up into an overall position — Initial, Emerging, Developing, Established or Leading — plus a breakdown by area. The breakdown is the part worth reading. A single headline figure is useful for getting attention in a leadership meeting; the detail underneath tells you what to do on Monday.
One rule specific to this half
Some of these questions are straight yes-or-no facts — whether HTTPS exists, whether you are certain about something. Those score 0 or 10 with nothing in between, because there is no partial version of them. The middle of the scale is reserved for questions that genuinely describe a spectrum.
Question bands: Not in place (0–2), Early or ad hoc (3–4), Developing (5–6), Established (7–8), Leading (9–10). Use the lower band unless stronger evidence supports the higher one.
Overall positions are Initial (0–20%), Emerging (21–40%), Developing (41–60%), Established (61–80%) and Leading (81–100%). The percentage is points earned divided by ten times the number of scored questions. Questions without sufficient evidence are excluded from that denominator and reported as gaps, never as zero.
Calculate scores in the workbook
A person chooses each score using the rubric. In Question set, enter the supporting evidence and score, then select the Evidence state. Open Results to see the percentage and coverage. The formulas check entries and calculate accepted scores; they do not judge the free-text answer.
A supported zero counts as a scored answer. Clear the score for insufficient-detail, not-applicable and not-covered answers.
Untouched rows remain visible as not-covered. Resolve flagged input issues before using a result.
Results show points earned, eligible points and coverage. No scored answers produces no percentage; fewer than half scored means evidence is too thin for a reliable maturity result.
Only the 43 digital questions enter the result. The routing record is unscored, and composite category labels are kept intact.
You can also enter human-reviewed AI scores to cross-check the arithmetic. Keep the scoring method identified and review differences before accepting a report.
Generate a full report
Before using a prompt or the linked GPT: follow the recording and data-handling checklist ↑. Upload only a safely prepared copy to an approved service and account. If sharing is inappropriate, use the manual workbook route instead.
Upload the transcript and only the workbook for the half you are scoring to your chosen language model, then paste Prompt 1. Keep names and sensitive guest information out of what you share. Review the cited evidence and disagreements before accepting a score. An AI-scored run may differ from a self-scored run; always label the method.
The report has two parts: a short management summary explaining the priority and first move, followed by the evidence and handoff details for the team responsible for investigating it.
Or use this tool
Hospitality AI Readiness Assessor
Use the GPT with your completed workbook or session transcript to turn the evidence into a readable readiness summary and next-step outline. Review the evidence before sharing or acting on the result.
A position and a breakdown by category — visibility, trustworthiness, booking integrity, technical foundations and the rest.
An audit of your website. This is a questionnaire; nothing is crawled or scanned.
A named priority category and one first move.
A rewritten site, a content plan, or an SEO strategy.
A clear statement of which findings sit with you and which sit with your brand or your agency.
A recommendation to buy any particular product or work with any particular firm.
This is a self-assessment, so nobody independent verifies the answers. Teams often describe the website state they expect rather than the page, feed, crawler path or booking journey a guest would reach today.
Use current examples, checks and dates wherever possible. An inflated score does not improve visibility; it only hides where evidence or control is missing.
A low answer is a starting position, not a judgment on the person reporting it. It identifies a capability that needs evidence, ownership or improvement.
When the result names your weakest area, there are three legitimate routes and the Alliance ranks none of them: work the breakdown internally as a checklist, engage a vendor, or engage a consultant. This assessment names no company and sells nothing.
The complete worked band anchors are not yet available. Treat the result as an introductory self-assessment, not a validated certification. The 90–120-minute digital session estimate is provisional until tested in pilots.
06
Step
Generate a 12-month roadmap
Use the separate advisory prompt with the scored assessment. It sequences work by dependencies, not simply by the lowest score, and does not make the roadmap an AIHA assessment result.
The roadmap keeps the same two-part format: a brief management summary of the proposed sequence, followed by a supporting table with evidence, responsibilities, dependencies and completion checks for the same items.
Apply the same data-handling checklist ↑ to the scored report and any supporting evidence before sharing them with an AI service.