Tulum's 3 series have diverged, and each one points to a different plan
MEXICO · HOSPITALITY
An InsightIA note · September 21, 2026 · corrected reissue of September 23, 2026 · reissue of October 12, 2026
Year to date through July, tourist arrivals in Tulum are down 8.6% and traffic at its airport, 33.5%. July occupancy stands at 48.7%.
What changed
The July 2026 "¿Cómo vamos en turismo?" (How are we doing in tourism?) report, the most recent that the Secretaría de Turismo de Quintana Roo (Quintana Roo Tourism Secretariat) publishes, closes the January–July cumulative period [V]. Tourist arrivals at the Tulum destination went from 962 thousand to 879 thousand, 8.6% fewer [D]. Arrivals in the Mexican Caribbean fell 3.6% [V]. In the same series, with the same universe and the same method, Tulum is falling 2.4 times faster than its region [D]. Tulum's airport, meanwhile, handled 530 thousand passengers against 797 thousand, 33.5% fewer, and its international passengers fell 43% [V].
What it reveals
Tulum's 2 declines come from the same report and the same cumulative period, and they measure different units: passenger movements on one side, estimated tourists on the other [V]. They also come from 2 methods, which the report itself states: the airport series is a count by the airport operator, and the destination series is an estimate by the Secretariat [V]. The report labels the air series "total de pasajeros" (total passengers) and does not define whether a passenger counts the arrival, the departure, or both [D]. Depending on how it is read, the airport itself therefore accounts for between 30% and 60% of the destination's tourists [D]. This is consistent with visitors still arriving through another gateway, and substitution remains unproven [H]: Cancún's airport also fell, by 5.2% [V].
What does not depend on that definition is the comparison of Tulum against its region, and there the destination is losing ground 2.4 times faster [D]. The registered inventory totals 239 hotels and almost 12 thousand rooms [V], and the report does not publish how many operated in July [D]. July occupancy was 48.7%, the second consecutive month below 50% [D], with Cancún at 70.6%, Costa Mujeres at 74.8% and the Riviera Maya at 58.4% [V].
Our hypothesis is that current practice anchors the asset plan to the series on which the asset was underwritten, and revises the plan when that series moves [H]. What this way of deciding fails to see is that the 3 series in this report have diverged: 33.5%, 8.6% and occupancy almost 22 points below Cancún [D]. Structured as a Decision System, Execution on Decision Systems would track the plan against all 3 series at once and reopen it when they diverge, rather than when a single one falls [H]. A plan that is already reviewed against all 3 and that reacted to the divergence before July would refute the hypothesis.
Which decision is affected
Who decides: the owner or the committee evaluating a hotel asset in Tulum
What they decide: whether to reposition it, and toward which source market
What changed in that decision: their criterion, namely which of the 3 series — airport traffic, destination arrivals, and occupancy — governs it
Whoever underwrote the asset on the strength of the airport reads 33.5% and concludes that the asset lost its access. Whoever looks at arrivals reads 8.6% and concludes that it is holding up. Against its own region, Tulum is not holding up [D], and neither of the 2 readings explains why July occupancy was almost 22 points below Cancún's [D]. That gap is resolved through positioning, price, and source market [H]. This week, the owner or the committee can review which of the 3 series is anchoring the asset plan and how often that series is checked again.
The question
Which of the 3 series should anchor the plan for your asset?
Sources
Secretaría de Turismo del Estado de Quintana Roo (Quintana Roo State Tourism Secretariat), "¿Cómo vamos en turismo?" (How are we doing in tourism?), July 2026, in the Sistema de Información Turística de Quintana Roo (Quintana Roo Tourism Information System) · https://siturq.gob.mx/como-vamos
Within that report: passenger movements at the airports, with the declared source being Grupo Aeroportuario del Sureste and the Secretaría de la Defensa Nacional (Ministry of National Defense); tourist arrivals, sourced to the Secretariat's own estimates; hotel occupancy and supply, sourced to the hotel associations, the Consejo de Promoción Turística (Tourism Promotion Council) and the municipal tourism offices
Evidence markers: [V] verified against a primary source · [D] derived · [H] hypothesis. The supporting evidence book, with 1 row per unit of claim, is kept in this piece's file.
The report prints the 33.5% for Tulum's airport, the 43% for its international passengers, the 5.2% for Cancún and the 3.6% for the Mexican Caribbean. The 8.6% for arrivals in Tulum is our own calculation from the absolute figures that the same report publishes, because the arrivals table prints the change only for the Mexican Caribbean. The 2.4 ratio is the quotient of 8.6% divided by 3.6%, which yields 2.38. The report marks the 2026 figures as preliminary.
The report labels the airport series "total de pasajeros" and does not define whether a passenger counts the arrival, the departure, or both. Counting both, the airport itself accounts for at most 30% of the destination's tourists; counting the arrival alone, 60%. The thesis therefore rests on the change in each series, measured year over year within its own definition, and on the regional comparison, which does not need that definition. The 30% to 60% proportion appears only in movement 2, with its condition stated.
Occupancy corresponds to the Tulum destination as the report defines it, and the report's supply table groups that row under the municipality of Solidaridad. The note does not compare July 2026 occupancy with July 2025 occupancy, because the report publishes the 2026 series and not the previous year's.
"The most recent that it publishes" refers to the reading date of September 23, 2026, and the report does not print its publication date.
This note reissues the one published on September 23, 2026. It retains the event, the figures, the sources, and the supporting evidence book, and adds the Decision Systems function to which the decision belongs and the improvement hypothesis that follows from it. The headline and the standfirst were rewritten to state the thesis.