Home / Sample Findings

A sample finding, the way leadership reads it.

Below is an illustrative sample of how SignalTide presents a finding. Each one records what an AI assistant could read about the property, what it said as a result, the evidence behind it, and the next step — grouped by the four pillars.

Illustrative sample. Not a client result. SignalTide has no hotel or spa clients and no published client results yet — proof is pending.
ILLUSTRATIVE NUMBERS · NOT CLIENT DATA · NOT A PERFORMANCE CLAIM

What the monthly report makes visible.

This invented example shows the reporting format, not a client result. A real report uses the property’s own prompt set, its own named competitor set, and its own direct-versus-OTA share.

Fictional example: one month, one prompt set, five assistants
PromptYou appearedNamed instead
boutique hotel in South Beach with a spa2 of 5Two flagged resorts, one independent
where to stay in Mid-Beach, quiet rooms0 of 5Three flagged resorts
small hotel near Ocean Drive, walkable3 of 5One flagged resort, one independent
Miami Beach hotel with a good breakfast1 of 5Two flagged resorts
where to book a massage in Miami Beach0 of 5Two hotel spas, one day spa
independent hotel in Brickell for a work trip4 of 5One flagged resort
Prompts where the property appeared3 of 6Movement shown against last month
  • Every prompt and result here is invented to show the reporting format. This is not client data and not a forecast.
  • Each prompt is run five times per assistant, because a single AI answer is not a stable result.
  • A prompt where the property does not appear is reported as zero rather than omitted.
  • The names that appear instead are recorded by name, so movement month over month is measured against someone real.
  • The report also carries the property's direct-versus-OTA share beside the placement, as a separate measure.

How to read it. Placement is reported as how often the property appeared across repeated runs, not as a rank position. No search engine or AI assistant sells a position, so none is promised here. The direct-versus-OTA share is shown beside it because findability that still ends in an OTA booking has not solved the owner’s problem.

Explore Shortlist Management
FINDING 01 · FOUNDOWNER: Marketing

No Schema.org markup for the property or its rooms

WHAT THE MACHINE READPage source: no Hotel, LodgingBusiness or Accommodation node
WHAT IT LED TOOnly a generic Organization tag; address, amenities and room types unreadable
EVIDENCE

An assistant looking for a Miami Beach hotel with a spa has nothing structured to read on the property's own site. The address, amenity list and room categories exist as images and prose, not as markup an assistant can parse and quote.

FINDING 02 · FOUNDOWNER: Marketing

Check accessible property information for missing or inconsistent facts

WHAT THE MACHINE READ/llms.txt returns 404
WHAT IT LED TONo structured summary of the property for an assistant to read
EVIDENCE

Assistants may use accessible site content and third-party listings; conflicting or stale facts warrant verification.

FINDING 03 · FOUNDOWNER: Web / Hosting

AI crawlers are blocked before they reach the page

WHAT THE MACHINE READrobots.txt: Disallow for GPTBot, ClaudeBot and PerplexityBot
WHAT IT LED TOAn edge rule also returns 403 to AI user agents
EVIDENCE

The site was locked down for SEO scrapers years ago and never revisited. Both the robots.txt directives and a firewall rule refuse the crawlers assistants use, so the property cannot be read at all, whatever the page says.

FINDING 04 · FOUNDOWNER: Web vendor

The page renders its content only with JavaScript

WHAT THE MACHINE READPage source: 41 words of body copy
WHAT IT LED TORooms, amenities and policies appear only after scripts run
EVIDENCE

A browser shows a full page. A fetcher that does not execute JavaScript sees an almost empty document, so the property's own detail never reaches the assistant that is trying to describe it.

FINDING 05 · DESCRIBEDOWNER: Marketing / F&B

An assistant describes a restaurant that closed two years ago

WHAT THE MACHINE READAssistant answer: “on-site restaurant, open for dinner”
WHAT IT LED TOThe restaurant closed in 2024; the site still lists it in prose and in an old listing
EVIDENCE

The closure was announced on social media but never removed from the website, a directory listing and an old press page. The assistant repeats the most consistent version it can read, and that version is wrong.

FINDING 06 · DESCRIBEDOWNER: Marketing

“Steps from the beach” is repeated as fact

WHAT THE MACHINE READSite copy: “steps from the beach”
WHAT IT LED TOThe walk is twelve minutes across Collins Avenue
EVIDENCE

The phrase is pleasant and imprecise. An assistant repeats it verbatim, so a guest who asked for beachfront arrives expecting something the property cannot deliver, and the review reflects the gap rather than the stay.

FINDING 07 · CHOSENOWNER: Marketing

The Google Business Profile carries three categories and no recent photographs

WHAT THE MACHINE READProfile: 3 categories, 6 photographs, most recent from 2022
WHAT IT LED TOCompetitors in the same answer carry fuller categories, current photographs and weekly posts
EVIDENCE

Assistants quote the profile and the reviews attached to it. A profile that has not been worked in years gives an assistant little to say, and what it says is thin compared with the properties named above the hotel.

FINDING 08 · BOOKED DIRECTOWNER: Revenue / Marketing

The answer links to an OTA rather than the property's own site

WHAT THE MACHINE READAssistant answer: two OTA links and one directory link
WHAT IT LED TONo link to the property's own booking page in the answer
EVIDENCE

The property is named, which is the first battle, but the links an assistant chooses point to a reseller. The booking arrives, and the commission goes with it. The direct-versus-OTA share in the monthly report is where this shows up.

What a real month adds to this. Your own prompt set, your own named competitor set, and the movement since last month — measured, not illustrated. The findings above show the shape of the report, not a client result.

Shortlist Management — $3,000 a month. One property. Six-month minimum. The $1,750 Shortlist Audit credits in full if you start within 30 days. No percentage of your spend, your bookings or your commission. The Shortlist Test is free.

Run my free Shortlist Test

AI assistants name a handful of hotels. Shortlist Management measures whether yours is on it, works on what keeps it off, and reports the movement every month.

FLORIDA-REGISTERED · PROFESSIONALLY INSURED · FOUNDER-LED · FIXED FEE, NO PERCENTAGE OF SPEND
MIAMI 4:07 AM© 2026 SignalTide LLC