The two-day tax as a readiness measure
Are Morch puts hotel AI readiness on a time measure. His column for Hotel Online, published on 6 October 2026, counts the hours that commercial teams spend exporting, copying, reconciling and rebuilding reports by hand. He calls the total the two-day tax.
He defines it as the share of every working week that revenue, sales and commercial teams spend assembling reports by hand. He writes that nobody chose to pay it and that it is collected before strategic work can begin.
What the column counts each week
The column reports that more than 80 per cent of hotel commercial teams spend one to two days each week on manual reporting and analysis. It reports that fewer than 10 per cent cut manual effort by more than 30 per cent, and that 13 per cent of hotel chains report measurable returns from AI so far.
Morch names sequencing as his own principle. A hotel fixes its data and its workflows first, and the tool purchase follows.
He marks one example as illustrative. A commercial lead at a 120-room independent hotel who spends one and a half days of a five-day week on reporting gives up 30 per cent of that person's capacity, or about 78 working days a year.
The chain-side numbers from h2c
h2c published its AI Opportunity Study 2026 on 1 October 2026. The study draws on 122 responses from 113 hotel chains: 101 online survey responses and 21 executive interviews. The participating groups cover Europe, the Middle East and Africa, Asia Pacific and the Americas, represent more than 8,200 properties and about 1.3 million rooms, and average 73 properties a chain.
Ninety-one per cent of the chains use AI. A further 8 per cent plan to adopt it within 12 to 24 months. Twenty-eight per cent carry a company-wide AI strategy led by senior leadership, and 13 per cent report measurable return on investment. The chains rate the contribution of AI to overall business performance at 5.6 out of 10 and their own internal AI knowledge at 3.4 out of 10.
The barriers rank this way on the study's own list: a lack of AI expertise, training or skills at 56 per cent; integration with existing systems at 38 per cent; no clear AI strategy or roadmap at 32 per cent; data security and privacy at 29 per cent; data governance and access management at 28 per cent; and cost at 22 per cent.
Daily use splits three ways. External tools such as ChatGPT carry 60 per cent of reported use, AI embedded in vendor systems carries 27 per cent, and internally developed or hosted tools carry 12 per cent. Thirty-one per cent of the chains now run primarily on vendor-provided AI, against 21 per cent in 2025.
Michaela Papenhoff, h2c founder and managing director, states that the industry has stopped asking whether to use AI, and that the question now is how chains turn broad experimentation into connected, governed and scalable systems.
The fragmentation behind the hours
Otelier's 2026 Hotel Operations Index, published on 28 January 2026, carries the operational side of the same problem. Eleven per cent of its respondents report a fully integrated technology stack. Ninety-one per cent still rely on some level of manual reporting, including inside workflows they have already automated. Fifteen per cent are very confident in the accuracy and timeliness of their operational data.
Twenty-seven per cent of respondents run more than seven technology platforms, and 27 per cent spend more than 11 hours a week consolidating or reconciling data. Fifty-two per cent describe industry progress as "slowly but steadily".
Twenty-five per cent of the index respondents consider themselves ready to adopt AI, and 40 per cent consider themselves not ready at all. The two use cases they rank highest are predictive demand modelling and cross-department data collaboration.
How a team counts its own two-day tax
Morch sets out a measurement a hotel can run inside one week. He asks revenue, sales, front office leadership and finance to log every block of time spent exporting, copying, reconciling or rebuilding a report, together with the system the data came from and where it went next. Each recurring report then gets the specific decision it supports written beside it. A report that supports no decision stops.
The remaining hours sort into three groups: work the systems the hotel already runs can automate, work that needs an integration between systems, and work where AI helps once the data is clean.
He reports that a meaningful share of manual reporting turns out to be work the existing systems can already do. He places the count first in the sequence, with no software purchase in it.
The question Morch puts to vendors
He puts one question to every vendor before a purchase: which of our manual hours does this remove, and how will we measure that? He writes that a vague answer points to another platform to reconcile.
The day-use lines none of the three documents carries
The three documents measure commercial reporting hours, chain AI use and operational fragmentation. None carries a day-use line. None states an hourly or daytime rate, a day-use occupancy figure, a daytime window or a revenue figure for a part-day slot.
Dayuse's United States search returned 2,574 day hotels on 14 October 2026. The first page of 50 cards ran from US$54.00 to US$199.00, with 24 four-star cards, 24 three-star cards and two at two stars. The window of 8:00 to 14:00 is the most common on the page, at 18 of the 50 cards. Every price reads in US dollars.
A day-use operator reading Morch's count starts from its own hourly inventory, which none of the three documents measures.
What the column leaves open
The column places 69 per cent of hotels keying guest preferences in by hand, and 19 per cent using AI to extract them, inside its State of Distribution paragraph. The State of Distribution 2026 report does not carry that pair; PhocusWire attributed it to h2c on 1 October 2026, and the h2c release read for this piece does not carry it either.
The State of Distribution site reports fully manual reporting workflows at 50 per cent of distribution specialists, and rising technology budgets at 62 per cent of hotels. The one-to-two-days-a-week figure the column leads on is not stated on that site.
h2c publishes no cost figure, no licence price and no headcount for the teams behind its 113 chains. The Otelier index states no sample size on the page read here.