Azerbaijan inbound trends: Q1 2026, corrected against H1 data

Birtour is a Baku-based DMC and MICE operator, and this note is our read of the official Azerbaijan inbound series for our tour-operator and corporate partners. The State Statistical Committee put H1 2026 arrivals at 1,092,600, down 10.3% year on year, after a Q1 of 484,388, down 4.7%. India fell 68.3% in Q1. Germany rose 76.6%. Business-purpose travel was the only segment that grew. We published a rebound call in May that the bulletins do not support, and this update says where.

Why the headline arrival number misleads a DMC buyer

Azerbaijan publishes arrivals, not tourists. Of 2,570,212 arrivals in 2025, 1,804,754 (70.2%) travelled for tourism purposes, and the rest entered for work, transit or other reasons. Iran is the extreme case at 8.7% tourism-purpose. Carry the headline into a 2027 brochure plan and you overstate your addressable market by roughly 30%.

The second trap is the word Gulf. Azerbaijani releases group Iran and Iraq into "Gulf countries", and the arithmetic settles it: Iran 208,476 plus Saudi Arabia 104,364 plus UAE 43,834 plus Kuwait 26,371 plus Oman 16,233 plus Iraq 7,405 plus Bahrain 5,284 plus Qatar 2,874 comes to 414,841, matching the official 2025 bloc total of 414.8 thousand exactly. So "Gulf down 31.4%" in H1 2026 is not a GCC number. We have seen it quoted as one in two partner decks this summer.

What did Azerbaijan inbound actually do in H1 2026?

Arrivals were 1,092,600 from 189 countries in January to June 2026, down 10.3% on H1 2025, in the Committee release of 23 July 2026. Q1 closed at 484,388, down 4.7%. January was still growing at 5.3%. The break point sits in April. Read the table row by row rather than as a trend line.

Only the January, Q1, four-month and half-year rows are published. April and Q2 are subtractions of published cumulative totals. They are arithmetic, not official quarters, and we label them that way because a fair amount of secondary coverage does not.

Which source markets grew, and which collapsed?

Q1 2026 is the only period where both years publish country absolutes, so it is the strongest evidence in this note. Germany, Great Britain, Uzbekistan, China, Israel, Saudi Arabia and Türkiye all grew. India, the UAE, Oman, Kuwait and Qatar all fell hard. Russia was flat in absolute terms and still gained share, because everything around it fell faster.

India is the line that should change a 2027 product plan. It was 244 thousand arrivals in 2024, 166,944 in 2025, and 17,373 in the first quarter of 2026. It was also 13.6% of all foreign hotel nights in Azerbaijan in 2025, the single largest nationality in the country's hotel base. No official release states a cause for the fall. Azerbaijani outlets attribute it to regional conflict, which is attribution rather than measurement, and we are not going to pretend to know more than the statisticians do.

Where our May read was wrong

Two corrections, both ours. We called a tentative Indian rebound off enquiry counts. The Q1 bulletin, published on 21 April 2026, three weeks before we posted, already put Indian arrivals at 17,373 against 54,840 a year earlier. Enquiry volume on a small base was the wrong indicator, and we should have opened the bulletin first.

The second correction is on entry rules. We wrote that GCC passport holders clear an Azerbaijan e-visa in three to five working days. That stopped being true on 15 February 2026. Saudi, Omani, Kuwaiti and Bahraini ordinary-passport holders now enter visa-free, three entries in the year to 15 February 2027, up to 30 days each, with a visa needed from the fourth entry. UAE nationals hold a standing 90-day visa-free regime and Qatari nationals 30 days. The rules split by country, not by bloc, and we set out all six in the Azerbaijan entry rules for GCC groups guide.

Our desk process changed after this. Any pipeline read we publish now gets checked against the most recent Committee bulletin before it goes out, and the bulletin number wins.

Is business travel falling along with leisure?

No, and this is the number a MICE desk should carry into 2027 planning. Business-purpose arrivals rose 2.6% in Q1 2026 while tourism-purpose arrivals fell 8.1%, lifting business from 21.8% to 24.4% of tourism-purpose visitors. Business spend was 671.4 million AZN in 2025, 20.6% of all tourism expenditure, at roughly 1,692 AZN a head.

The Western long-haul markets are business-weighted, which is the same story from a second angle. In 2025, 57.7% of tourism-purpose arrivals from the USA travelled for business, 50.3% from Great Britain and 44.7% from Germany. Germany then grew 76.6% in Q1 2026. A market that is nearly half business and growing off a small base is a corporate signal, not a leisure one. Our own read of that demand sits in the GCC corporate offsite shift to the Caucasus note and, at 120 pax, in the GCC bank incentive case study.

Two observations from our own book over the same Q1 2026 window, offered as pipeline colour rather than as data. GCC private-banking, corporate offsite and incentive enquiries roughly doubled as a share of our Q1 MICE volume against Q1 2025. And across the co-quoting partnerships we run with Tbilisi and Yerevan DMCs, operators packaging Baku with Tbilisi and Yerevan were requesting room-block holds for September and October 2026 during Q1, roughly six months earlier than the equivalent 2025 requests. Both sit on a small base. The Q1 correction above is what happens when we forget to say that.

What are Baku hotels actually running at?

National capacity utilisation was 19.3% in H1 2026 against 21.4% in H1 2025, with Baku at 29.4%. Hotel overnights fell 8.3% and hotel revenue fell 3.8%, so revenue per occupied night rose about 5%. Operators are trading volume for rate. Azerbaijan's occupancy metric is bed-capacity based and is not STR-comparable room occupancy.

The composition explains the pressure. India was 13.6% of foreign hotel nights in 2025, Saudi Arabia 8.3%, UAE 5.1%. India and the Gulf together were about 27% of foreign hotel nights, and those are precisely the markets that fell in 2026. Supply kept growing into it: 859 hotel and hotel-type facilities at the start of 2025, up 6.2%, and roughly 31,750 rooms, up 10.7%. Falling demand against rising room count is why the rate conversation on group blocks is easier this year than last.

Why is the airport holding up when arrivals are not?

Heydar Aliyev International handled 3,334,000 passengers in H1 2026 against 3,452,000 in H1 2025, down 3.4%, while foreign arrivals fell 10.3%. Transfer passengers passed 275,000, up more than 40%. The airport served 76 destinations with 40 carriers, against 69 destinations a year earlier. Baku is converting into a connecting hub faster than it is losing origin-destination demand.

The route file kept expanding through the downturn. Air Serbia opened Belgrade to Baku on 3 May 2026, twice weekly. FLYONE Asia began scheduled service to Baku and Ganja and launched Baku to Tashkent. FLYONE opened Baku to Chisinau on 16 July 2026. AZAL launched Shymkent and put Aktobe, Atyrau and Brussels on sale. Wizz Air starts Baku to Bratislava in October 2026. Roughly 20 of the carriers at the airport are low-cost or hybrid. For a circuit operator, the flight side is improving while the demand side falls, which is an unusual combination and a good moment to contract.

What we would not read into these numbers

Four things we will not do with this dataset, named here rather than left for a partner to discover after quoting us. Azerbaijan's occupancy measure is not internationally comparable, no reliable Baku ADR series exists, MICE is not a published category at all, and several aggregators reported the 2025 headline with the sign reversed.

• We will not benchmark this occupancy internationally. Azerbaijan publishes bed-capacity utilisation. Setting 19.3% next to an STR city figure compares two different measurements.
• We will not publish an ADR for Baku. No STR, CoStar, Colliers or JLL series was reachable for this market. The only rate-shaped figures we found were short-term-rental derived, which is not the hotel base.
• We will not quote a MICE arrivals number. There is no MICE category in the Committee's output. Business purpose is the only proxy, and it bundles every kind of corporate travel.
• We will not source these figures from aggregators. Several outlets reported 2025 as a 2.1% rise and the CIS line as up 8.2%. The primary release says both fell.

What should tour operator partners do with this?

Three moves for a 2027 product plan. Re-weight away from single-market dependency, since India carried 13.6% of foreign hotel nights and then fell 68.3% in a quarter. Price the Caucasus loop as one product rather than Azerbaijan alone. And treat corporate as its own line, because it is the only segment in the official series that grew.

The operational playbook for the loop is in our Caucasus circuit guide for tour operators, and the leisure anchor most partners pair with it is the 5-day Baku itinerary. For the corporate line, the scope we deliver against sits on the MICE Azerbaijan page and the cost shape on the 50-pax corporate offsite breakdown. Net-rate conversations run through the Azerbaijan DMC desk.

What we are watching for the rest of 2026

Five things, and each has a dated release behind it rather than a hunch. The Jan-to-Sep bulletin due in October is the first read on whether Q3 stabilised. Saudi Arabia's Q2 shape, the business share of tourism-purpose arrivals, the 15 February 2027 visa-free expiry, and the still-unpublished H1 2026 statistical bulletin are the other four.

• Whether April was the floor. April 2026 was the sharpest single month at roughly 23.5% down. The Jan-to-Sep release, due in October, is the first read on whether Q3 stabilised.
• Saudi Arabia. Up 13.6% in Q1 2026 but only 1.8% of H1 arrivals, which implies a sharp Q2 fall. Saudi is our largest GCC corporate origin, so this line matters more to us than the bloc total.
• The business share. 24.4% of tourism-purpose arrivals in Q1 2026, up from 21.8%. If it holds through the Q3 bulletin, the corporate case for Baku is stronger than the headline suggests.
• The visa-free pilot clock. The Saudi, Omani, Kuwaiti and Bahraini concession runs to 15 February 2027. Anything contracted near or beyond that date needs re-checking against the Ministry register.
• The H1 2026 statistical bulletin. Not published as of 10 August 2026. Country-level absolutes for H1 exist only as share arithmetic until it lands.

Sources and method

Arrival, spend and hotel figures come from the State Statistical Committee: the 2025 inbound bulletin, the January to March 2026 bulletin, and the releases of 21 April, 20 May, 23 July and 27 July 2026. Airport figures come from the operator's own H1 releases of 10 July 2025 and 15 July 2026. Entry rules come from the Ministry of Foreign Affairs register.

Everything labelled derived is arithmetic on published cumulative totals, and it is flagged where it appears rather than in a footnote. Birtour pipeline observations are our own enquiry book, normalised year on year; they are directional notes for partners, not statistical claims, and the correction above is what happens when we let the two blur. We rebuild this note each quarter against the newest bulletin.

For operator partners re-weighting a 2027 Caucasus programme against these numbers, the working playbook is the Caucasus circuit guide. For corporate buyers pricing a Baku offsite against the business-travel line above, start with the 50-pax corporate offsite breakdown.