Your hotel probably isn’t on the shortlist.
When travelers ask AI where to stay in Miami Beach, a few names come back. SignalTide measures whether yours is one of them, and works on what keeps it off. $3,000 a month, founder-led.
Shortlist Management
Shortlist Management costs $3,000 USD a month for one independent Miami hotel or boutique spa, on a six-month minimum. The Shortlist Audit is $1,750 one time and credits in full against the engagement if you start within 30 days. The Shortlist Test is free, needs no call, and is four booking intents across two engines — indicative, not the full protocol.
When a traveler asks an AI assistant where to stay in South Beach, a handful of names come back. Michael Brady measures whether your property is one of them, fixes the schema and crawler access that decide whether it can be read at all, builds the citations assistants quote from, and reports placement and movement against a named competitor set every month. No ranking position is promised and no number of bookings is promised.
A channel, not a novelty
Independent hotels already watch two discovery channels closely: what Google shows, and what the OTAs do with their rates and listings. A third now sits beside them. This replaces nothing a property already does — SignalTide does not set rates, manage inventory, buy media or touch revenue management. Your channel manager shows OTA versus direct. Search Console shows clicks from Google. Neither tells you whether an assistant named you, or what it said about your resort fee. The engagement covers three channels: AI discovery, OTA placement and direct booking. Being named by an assistant is not the same as being bookable, so both are worked, along with the OTA listing the property is compared against.
The Miami Shortlist Index, Edition 1
Published 23 October 2026. Six AI tools — Google AI Mode, Gemini, ChatGPT, Grok, Perplexity and Copilot — were asked three verbatim traveler search phrases, logged out, against 31 independently owned South Beach hotels between 1st and 23rd Street.
15 of the 31 hotels were never named by any of the six tools, for any of the three phrases. 16 were named at least once. Exactly one, The Betsy, was named by all six. No single tool named even half the market: Grok reached 11 of 31, Google AI Mode 10, ChatGPT 9, Gemini 8, Copilot 8 and Perplexity 4 — so optimizing for one assistant is not optimizing for the others.
The tools also disagreed with each other, and with hotels’ own published information, about resort fees, beach proximity and whether a pool exists, and one tool could not verify basic facts for eleven properties at all. Limits are published with the findings: one date, three phrases, searched from Michigan rather than Miami, hotels only, and no claim that placement produced or lost a booking. Read the Miami Shortlist Index .
The Booking Map
68% of all US Google searches ended without a click across all search topics in early 2026 (SparkToro), and click-through for the top unpaid result fell between 47% and 60% when an AI summary appeared (Ahrefs measured 47%, SparkToro close to 60%). 63.4% of bookings in Cloudbeds’ 2025 global independent-hotel sample went through an online travel agency, against 36.6% direct. SignalTide models OTA commission at 18% against roughly 4% direct, a 14-point gap — those two are model assumptions for illustration, not measured figures for any property. See the Booking Map .