# 48% of hoteliers don't trust their data for AI

Canonical-URL: https://notitur.com/en/digest/hotel-data-ready-ai-actabl
Text-Version: https://notitur.com/en/digest/hotel-data-ready-ai-actabl.txt
Markdown-Version: https://notitur.com/en/digest/hotel-data-ready-ai-actabl.md
Language: en
Published: 2026-11-22T12:31:28.574Z
Modified: 2026-11-22T12:31:28.574Z
Author: Joan Sanz
Topics: Artificial Intelligence
Summary: Nearly half of hotel executives doubt their information is ready for artificial intelligence. These are the four tests that separate useful data from noise.

## Article

Nearly half of hotel company leaders, **48%**, are not confident their data is ready for artificial intelligence. That is the finding of the **State of Hotel AI 2026** research from Destination AI, cited in [Hotel Operations' analysis of how to get hotel data ready for AI](https://hoteloperations.com/hotel-data-ready-for-ai/?utm_source=rss&#038, utm_medium=rss&#038, utm_campaign=how-to-get-your-hotel-data-ready-for-ai). The number matters: almost half the sector would let a model decide on figures it does not control.

The underlying problem is familiar to anyone who has managed a hotel. **Joseph Benjamin**, CTO at **Actabl**, puts it plainly: if you sit down with an owner to explain why the asset missed budget by **USD 200,000**, an approximate answer is useless. And **Stephen German**, who leads product at the same company, adds the other symptom: too many systems, different questions and different answers for the same metric. That lack of a single source of truth is where AI fails before it even starts.

Actabl's proposal rests on four concrete tests every hotel data set should pass before feeding a model:

- Be **connected and normalised**, so the same indicator means the same thing across systems.
- Undergo **daily accuracy checks**, not an annual clean-up.
- Have **clear ownership by the hotel company**, not by whichever vendor is in charge.
- Be **governed** so guest and financial data stay protected.

The reading is clear: AI investment is eating the budget while the underlying data remains a collage. And that is an opportunity many are missing, because sorting out data is not a technology project, it is a revenue project and a trust project with the owner. Before buying another shiny assistant, test the existing one on a single real decision, a budget, a forecast, a rate, and see if the number balances to the cent.

The door opening is wide for independent hotels and small management companies: whoever has clean, governed data will negotiate better with OTAs, sharpen forecasts and defend the P&L with arguments rather than intuition. The risk is staying with the model's promise without having touched the source spreadsheet.

Watch two things in the coming months: whether vendors start competing on data ownership and governance rather than interface alone, and whether management companies build data quality into their contracts. That is where it will become clear who is serious about AI and who just wants a demo.

## Sources

- Hotel Operations: https://hoteloperations.com/hotel-data-ready-for-ai/?utm_source=rss&#038;utm_medium=rss&#038;utm_campaign=how-to-get-your-hotel-data-ready-for-ai

## Citation

48% of hoteliers don't trust their data for AI. Notitur. 2026-11-22. https://notitur.com/en/digest/hotel-data-ready-ai-actabl
