Monthly tourism indicators — 3-5 star hotel guests, occupancy, revenues

Schema

Latest rows

Use the API

curl https://omandata.dev/v1/tourism/latest.json

Downloads

• CSV
• Parquet
• Changelog & raw snapshots

Official source

NCSI via data.gov.om (Tourism, dataset dedblxg)

Data license: Open Government License - Sultanate of Oman

Source notes

Monthly hotel-sector indicators for the Sultanate of Oman from NCSI's Tourism dataset on data.gov.om: number of hotel guests, room occupancy rate (percent), and hotel revenues in millions of Omani Riyals. Series run from 2012-01 to the latest published month; months missing any of the three indicators are dropped rather than published with gaps, which removes all of 2013 (the source carries occupancy and revenues for that year but no guest counts). Totals only — the source's breakdowns by guest nationality, hotel class and governorate are out of scope for v1, and the query pins those three dimensions to their total members because the unfiltered cube returns 828 series instead of 3. Occupancy is the source's "Room Occupancy Ratio"; its other percentage member, "Occupancy Rate (%)", is the same figure rounded to one decimal, starts 2014-01 and exists only under the "HOTELS CLASSIFIED (3 - 5 ) STARS" hotel class, so it is not used. Despite the source's "total" hotel-class label, the monthly figures are NCSI's classified three- to five-star hotel universe, not all accommodation — which is why the titles say so: from 2015-01 the total and the 3-5-star series are identical month for month, and the Jan-May 2026 figures published here reproduce NCSI's press release for 3-5-star hotels exactly (856,929 guests, 47.9% average occupancy). NCSI revises recent months as late hotel returns arrive — guests for Jan-Feb 2026 moved from 439,287 to 447,943 between the February and May releases — so figures for the newest months should be read as provisional. Two caveats come with the portal itself. Its REST API is called with the public frontend client_id that data.gov.om embeds in its own homepage HTML: a published site key rather than a credential, and one that can rotate, so a 403 makes the client re-scrape it from the homepage and retry once instead of hard-coding a key that will one day expire. And its unit-scale metadata is not to be trusted here: revenues arrive as plain riyals under an empty unit label and scale 1, so the riyals-to-millions factor is pinned in the parser, guarded by the source's declared scale and by a plausibility test. Regional and hotel-class series in the source stop at 2014-12 and are another reason the national totals are the only v1 output.