Tourism marketing strategy: find the audience behind a quiet month

The charts use an index to put different-sized destinations on the same readable scale. An index of 100 means that nights per day match the annual average for that region and visitor group. An index of 80 means 20% below that average; 120 means 20% above it. Each group has its own baseline.

“Domestic” means living in the destination country. “Foreign” means living outside it. These categories describe residence, not nationality, language or distance traveled. A domestic visitor to Zagreb could come from anywhere in Croatia, not only from nearby.^[2][3]

We looked for a substantial contrast recurring in the same month for the same group in all three years: a total index of 90 or less and a group index of 110 or more. These are practical screening thresholds, not a statistical significance test or a definition of the best month to advertise.

In May 2025, Asturias recorded 453,350 accommodation nights, including 121,092 by visitors living outside Spain. Total nights per day were 18.2% below the region’s annual average. Foreign nights per day were 16.5% above their own annual average. The same contrast passed our thresholds in May 2023 and May 2024.^[1]

Foreign visitors accounted for 26.7% of May’s nights, compared with 18.8% across the full year. They were still a minority. The finding is that their presence was more pronounced in this month, not that businesses should stop serving domestic visitors.

This is a seasonal pattern, not a growth claim. May’s foreign-night count fell from 129,009 in 2023 to 121,092 in 2025. Even so, foreign nights per day in May remained above that group’s annual daily average in each year. A business can therefore see a recurring audience pattern without assuming that the number of stays is rising. The annual baseline is recalculated for each year, rather than held at its 2023 level.

For a hotel, attraction or destination creator, this is a reason to inspect the practical information accompanying May content. Are arrival instructions clear? Are opening dates, reservation requirements and the location of the experience easy to find? A beautiful image can start the story; useful detail can answer the next question.

The residence split alone cannot choose the language of that content. A business would need its own customer questions or a more detailed origin breakdown before deciding which translations to prioritize. Nor does a May stay reveal when someone researched or booked it. Use the pattern to choose an audience to investigate, rather than treating May as a proven campaign launch date.

Zagreb recorded 190,420 nights in November 2025, including 59,166 by residents of Croatia. Total nights per day were 14.9% below the city’s annual average, while domestic nights per day were 31.4% above their own average. Domestic visitors supplied 31.1% of November’s nights, versus 20.1% for the year.^[1][3]

For a concrete reading of the index, Zagreb’s 59,166 domestic nights over November’s 30 days equal about 1,972 nights per day. Its annual total of 547,728 domestic nights over 365 days averaged about 1,501 per day. Dividing the first by the second and multiplying by 100 gives an index of 131.4. These are nights of accommodation use, not 1,972 different people arriving each day.

Pannonian Croatia showed an even more pronounced change in audience share. Domestic visitors contributed 54,432 of its 75,026 November 2025 nights. Their share was 72.6%, compared with 40.7% across the year. Total nights per day were 33.1% below their annual norm, while domestic nights per day were 19.2% above theirs.

Both regions met the same daily-index rule in November of all three years. But the content implications need not be identical. Domestic visitors remained a minority of nights in Zagreb; in Pannonian Croatia, they were the majority that November.

A business could use this contrast to review which questions its November content answers. Examples might include a clear itinerary, how an experience fits into a short stay, or practical information for returning visitors. These are content ideas to test, not explanations established by the dataset.

The data does not identify business trips, family visits, events or transport choices. Assigning one of those stories to the pattern would go beyond the evidence. A useful brief can name the audience and the questions to answer while leaving the reason for travel open.

Start with a specific combination: destination, month and audience. Then write the information that audience needs from your business. “November content for visitors living in Croatia” is a clearer starting point than “off-season awareness,” but it still needs a real customer question to become useful.

For a creator partnership, that could mean agreeing on an itinerary, an arrival explanation and a factual guide to what is open. For a hotel’s own channels, it could mean checking that a seasonal post leads to a page with matching dates and details. These suggestions concern the usefulness of the content; this study has not measured their effect on bookings.

Keep the next step small enough to evaluate. Choose a content hypothesis, record what you changed, and compare relevant results using your existing measurement setup. Do not replace the whole calendar or increase spending simply because a regional pattern exists. One property, creator or attraction may serve a different audience from the regional accommodation total.

Also separate the stay month from the publishing date. This analysis can suggest who deserves attention in a seasonal brief. It cannot supply a booking lead time. Historical inquiries, reservations or other appropriate first-party evidence would be needed for that decision.

Our screen found ten recurring combinations across nine of the 230 eligible regions. One region, Puglia, had two qualifying months. The other 221 regions did not meet this particular rule; that is not evidence that their audiences have identical seasonal patterns.

The three examples discussed above survived a targeted review of contemporary Spanish and Croatian methods. The complete table also retains the Italian and Greek numerical cases, rather than making the result look cleaner by removing them after calculation.

Three Italian cases need an unresolved comparability warning: Puglia in May and October, and Basilicata in October. ISTAT expanded published tourism coverage in 2025. We could not establish whether this Eurostat series retained the same accommodation scope across our three years, so these are not used as planning examples.^[4]

Thessaly in December and the Ionian Islands in April also remain numerical results only: we could not complete the contemporary Greek methods check. Malta’s February result covers collective accommodation, while the Polish monthly series covers establishments with at least ten bed places. These scopes matter when applying the findings to a particular business.^[5][6]

The table below gives every qualifying combination in every year, including the underlying nights. Read the indices alongside those counts. A group can be unusually active relative to its own annual pattern and still represent a small share of a destination’s activity.

A short month has fewer opportunities to accumulate nights. In Malta’s collective accommodation data, domestic nights per day in February were 19.4%, 10.8% and 41.6% above their respective annual averages in 2023, 2024 and 2025. Total nights per day were below the annual norm in each year, meeting our quiet-month threshold.^[1][5]

If we instead compare raw monthly counts with one-twelfth of the annual total, Malta’s domestic February result no longer reaches the 10% threshold in 2023 or 2024. February had 28 days in 2023 and 2025, and 29 in 2024. That difference is why the main analysis uses nights per day.

Nine of the ten main cases also passed the raw-month comparison. All ten remained when we required unflagged values only, and all ten remained when the component counts had to add up exactly. These checks strengthen the arithmetic; they cannot resolve an undocumented change in what establishments a country covers.

We used Eurostat’s monthly regional accommodation-nights dataset, TOUR_OCC_NIN2M, for 2023–2025. It combines hotels and similar accommodation, holiday and other short-stay accommodation, and camping grounds. The source was updated on October 12, 2026 and retrieved on October 18. National reporting thresholds, collection methods and coverage still differ, so we compare each group with itself within a region rather than ranking countries.^[1][2] The Asturias check also covered Spain’s current collection methods and historical series-linking information.^[8]

Before reading the observations, we fixed a roster of 244 territorial regions in the EU’s NUTS 2024 classification. We excluded nine whose boundaries changed relative to the preceding classification, leaving 235 candidates. Four had missing or confidential required values, and one had low-reliability values. The remaining 230 had complete usable panels. Missing and confidential observations were never treated as zero.

For each region, residence group and year, we divided the month’s nights by its actual number of days. We divided the sum of all twelve months by 365 days, or 366 in 2024, to obtain the annual daily average. The index is 100 multiplied by the monthly daily average divided by that annual daily average. Monthly and annual audience shares use the corresponding total nights as their denominators. A usable comparison required a positive annual baseline for every residence group.

We checked that domestic plus foreign nights reconciled with total nights, and that monthly sums reconciled with the separately published annual controls. The planned tolerance was the larger of one night or 0.1% of the larger count; every eligible panel reconciled exactly. We excluded regional panels containing incompatible reliability, definition, break or confidentiality flags. All values in the eligible panels were unflagged, which is not proof of perfect coverage.

The results describe this fixed source release. They are not a representative estimate of all European tourism businesses. Accommodation nights are not unique visitors or occupancy rates, and the file does not measure spending, advertising returns or booking dates. National method reviews support the named examples but do not make every country’s accommodation coverage identical.

The downloads provide the full results, exclusions, sensitivity checks and calculation method. They also retain the distinction between a numerical match and a case suitable for the article’s practical examples.^[7]