ITNº 19Hospitality Tech

90% of Hotel Leaders See AI Gains in Routine Tasks, Only 15% on the Frontline

A 107-leader AHLA-Actabl survey reports routine-task AI gains at 90%, but academic analysis of 202 hotel tasks finds 90.78% face automation bottlenecks.

· 3 min at the table

AI in Hotel Operations: Productivity Gains and Frontline Challenges
AI in Hotel Operations: Productivity Gains and Frontline ChallengesLUM3N / pixabay

Menu du jour

  1. First course

    AHLA-Actabl Destination AI survey of 107 hotel leaders: 90% report AI improves routine tasks, only 15% see frontline gains.

  2. Second course

    Study in Annals of Tourism Research: 90.78% of 202 hotel tasks in 12 Spanish hotel occupations face moderate-to-severe automation constraints.

  3. Third course

    Labor hours per occupied room fell 3.1% at full-service and 3.5% at select-service hotels in H1 2026.

  4. Fourth course

    Only 24% of surveyed leaders believe AI investments have paid for themselves; 42% trust their data readiness.

  5. Fifth course

    Housekeepers face constraints in 94.59% of tasks, waiters in 93.55%, receptionists in 85.71%.

Ninety percent of hotel-company leaders say artificial intelligence improves time spent on routine tasks, yet only 15% report any improvement in frontline operations. Those two figures, published in a November 2026 report by the American Hotel & Lodging Association (AHLA) and Actabl, frame the central tension in hotel AI: administrative productivity is moving; guest-facing work is not.

The report draws on a Destination AI survey of 107 hotel-company leaders. Among respondents, 84% said AI improved their own work. Only 6% of respondents held operational roles, which means the findings largely reflect management perception rather than frontline experience — a caveat the report itself does not resolve.

What does the labor data actually show?

Participating full-service hotels recorded a 3.1% decline in labor hours per occupied room during the first half of 2026; select-service properties recorded 3.5%. The numbers indicate improved labor productivity. They do not establish AI as the cause. No controlled comparison separates technology effects from occupancy mix, staffing changes or pricing.

A separate study, published in the November 2026 edition of Annals of Tourism Research, quantifies why the frontline gap persists. In "Job Automation Constraints in Hotel Work: Task-Level Bottlenecks in Four- and Five-Star Sun-and-Beach Hotels," researcher Santiago Melián-González analyzed 202 tasks across 12 hotel occupational categories in Spain. Three AI systems assessed each task against engineering constraints: physical manipulation, manual dexterity, social perceptiveness and negotiation.

The result: 90.78% of tasks encountered at least one moderate-to-severe bottleneck. Applying a severe-only threshold, the figure falls to 40.48%. The percentages describe technical obstacles, not job losses. Housekeepers faced moderate-to-severe constraints in 94.59% of tasks, waiters in 93.55% and receptionists in 85.71%.

The study has limits. It weighted all tasks equally regardless of frequency or duration, and it assessed feasibility on paper rather than in operating hotels. It cannot quantify labor savings.

Where PMS automation works, and where it stalls

Property management systems hold reservations, guest profiles, availability, occupancy and room status. AI assistance in retrieving and cross-checking this structured data is where the 90% administrative gains concentrate. A receptionist handling an early arrival can get faster answers on alternative rooms and housekeeping status.

Resolution, however, still requires explaining options to the guest and exercising judgment — exactly the social-perceptiveness constraints the academic study measured. Data quality also sets the ceiling: wrong room statuses or incomplete reservations produce unsuitable recommendations, adding correction work rather than removing it. Hotels can measure results through reservation processing times, room-assignment accuracy, unresolved requests and manual correction rates.

POS and housekeeping: partial, not wholesale

Hotel POS systems already handle order processing, payments and room-charge posting; AI could extend into demand forecasting and transaction analysis. But the waiter data is instructive: 93.55% of waiter tasks face moderate-to-severe constraints, while only 35.48% face severe ones. That gap defines the realistic opportunity — automate individual activities such as payment, keep the service role intact.

Housekeeping shows the same split. PMS occupancy data and departure forecasts can improve scheduling and room prioritization, measured through turnaround times and scheduling accuracy. Physical cleaning and inspection remain manual work under the study's dexterity constraints.

What does this change for operators?

The investment picture is cautious. Only 24% of surveyed leaders believed their organization's AI investments had paid for themselves overall; 28% said it was too early to tell. Just 42% expressed confidence in their data readiness for AI.

The practical answer is workflow-level measurement rather than adoption-level claims. Operators should benchmark before and after implementation — reservation retrieval times, billing corrections, order accuracy — while accounting for occupancy and staffing changes. Otherwise, as the survey's own 15% frontline figure suggests, faster processing can create work elsewhere in the hotel.

Shiji Group, whose insights publications informed the analysis, serves over 91,000 hotels worldwide through its PMS, POS, distribution and payments portfolio. Expect operators to keep buying AI for the back office while the 202-task constraint data keeps full frontline automation off the near-term procurement list.

via insights.shijigroup.com (Original)

Share this article:

More from Grace Kim

Grace Kim

Show full bio

Market editor covering companies, deals and strategy in European hospitality industry at Eurotables Report.

137 articles

Other tables tonight

« Previous tableNext table »