Orsina ValeChatGPT legibility
Hotel legibility

ChatGPT can only explain the hotel it can clearly read.

A free mini-course for independent Italian boutique hotels and guesthouses that want to understand how ChatGPT describes their property. I teach owners and managers to read AI answers with care: the name, the category, the location, the source trail, and the claims that sound confident but have no stable evidence behind them.

ChatGPT answerCase · Casa Livia
Guest question

“Is Casa Livia on Lake Garda a good place for a quiet weekend?”

Generated answer

Casa Livia is a small family-run place near the lake — a wellness-focused boutique hotel, with relaxing dining options.

Room · Name
exact one consistent entity
Room · Category
blurred guesthouse read as “boutique hotel”
Room · Place
precise near the lake, right town
Room · Promise
overstated “wellness” with no evidence
§ 01 — What the course examines

What the course examines

Across 15 lectures with short self-check tests, we look at the places where an Italian hospitality business becomes readable or confused: the official name, address, type of property, room and service descriptions, nearby landmarks, review context, booking listings, old pages, and Italian-English phrasing. The course is free and carries no obligation. It is written for people who already know their own property well and now need a disciplined way to see how an AI assistant may assemble that knowledge from public fragments.

  • 15 lectures
  • Free tuition
  • 15 tests

View curriculum

§ 02 — Six tracks

The six tracks of the reading

Mark how the model arrives at the property — or where it confuses it.

01Identity Name, address and entity consistency
02Category How the property is classified
03Sources Pages, directories and review evidence
04Language Italian and English retrieval differences
05Place Location and service-area signals
06Upkeep Audits, recency and corrections
§ 03 — Read the lecture library

Read the lecture library

The lecture index is the working shelf for the course. You can follow the programme in order or return to one lecture when a specific problem appears, such as a wrong category, a stale address, or a promise that ChatGPT has invented from thin evidence.

All lectures

Make the property easier for ChatGPT to recognise before asking it to describe you well.

Start with the four rooms: name, category, place, and promise.