How short of staff are hotels and restaurants, in which countries, and what does a hospitality worker cost?

How tight the market actually is

Official vacancy and shortage measures, each against its own economy-wide benchmark. The US rate is a monthly job-openings rate; the Japanese diffusion index is the share of establishments reporting a shortage minus the share reporting a surplus, not a vacancy rate. The two cannot be compared with each other — only each against its own all-industry row.

Turnover: the number that actually separates hospitality from other sectors

Japan publishes annual accession and separation rates by industry on one methodology, which makes the comparison with the whole economy clean. US JOLTS hires are a monthly rate, not an annual one.

What a hospitality worker costs the employer

Eurostat labour costs cover wages plus employer social contributions, per person employed per year, for accommodation only (NACE I55). The US figures are gross earnings paid to the worker in leisure and hospitality (NAICS 70) and are not comparable with the EU employer-cost figure.

Value added per person employed, accommodation (NACE I55), thousand EUR

One Eurostat indicator, one year, all size classes. This is the ceiling on what a hotel can pay and on what software can plausibly charge per employee: a Belgian hotel employee produces three times the value added of an Estonian one.

Value added per hour worked by employees, accommodation (NACE I55), EUR

The same dataset on an hourly basis, which removes differences in part-time intensity. The spread stays wide: an hour of Belgian hotel work carries more than three times the value added of a Lithuanian one.

Persons employed per accommodation enterprise (NACE I55)

This is the single most important row set for anyone pricing hospitality software per seat or per property. The average European accommodation business is a handful of people, and the differences between countries are larger than the differences between hotel categories inside one country.

Who the workers are: the Dutch case as a worked example

Eurostat publishes employment in accommodation and food service by age for every member state; the Netherlands is shown as a full series because it has the starkest age profile in the set we hold. A workforce that is majority under 25 explains the turnover numbers above and dictates how onboarding and scheduling products must work.

What hospitality operators buy, and what stops them

Eurostat asks every EU enterprise the same AI questions, which makes accommodation and food service comparable with other sectors. The blocker shares are taken among enterprises that ever considered using AI, not among all enterprises — a narrower denominator. The AHLA rows come from one survey of 246 hotel owners conducted in late February 2026; it is a multiple-response question, so the shares do not sum to 100.

Methodology

Every number on this page is a fact node in the God of Startups Atlas — a knowledge graph where each metric carries its source URL, reporting period, evidence tier and the verbatim quote it was taken from. Tier 1 means official statistics (Eurostat, US Bureau of Labor Statistics, Japan's Ministry of Health, Labour and Welfare); tier 2 means industry research with a disclosed sample. Values are shown in the unit and currency the source reports; nothing is converted or modelled. Two figures are arithmetic on two published values and are labelled as such in the row: the twelve-month change in US weekly earnings and the youth share of Dutch hospitality employment.

Related benchmarks

Analysis built on this data

Market cells behind this benchmark

Fact tables from the same knowledge graph — one market, every number with its source, period and evidence tier.