Hotel blog: how much do the world’s best hotels publish, and does it bring traffic?

Half of the world’s best hotels have no blog. On 12 August 2026 we read 669 of their websites, and 340 had no articles section of any kind. A live blog, with at least one dated article in the previous twelve months, was running at 153 hotels, fewer than one in four.

We wanted to know how much the most decorated hotels really publish, so we counted, site by site. We started from four rankings: the 343 Five-Star hotels in the Forbes Travel Guide for 2026, the 840 hotels on Travel + Leisure’s World’s Best lists for 2026, the 713 holding two or three Keys in the Michelin Guide that day (812 since the selection of 18 September) and the 100 hotels on the 2025 list of The World’s 50 Best Hotels. Once duplicates are removed, that leaves 1,420 hotels on 822 websites.

Of those, 708 sit on a chain’s website: 82 on fourseasons.com, 51 on marriott.com, 46 on ritzcarlton.com, 33 on mandarinoriental.com. Any articles section there belongs to the brand and serves all its hotels, so we left them out. We read the 712 hotels that have a website of their own. For 43 we could not tell either way, and we have not counted them among the hotels that publish nothing. That leaves 669.

A site that is hard to read can pass for a site with no articles, so we looked in three places. The first was the feed, the list of latest articles that many sites keep up to date for the programs that follow them. The second was the index pages reached from the menu or the sitemap. For the third we tried the commonest addresses directly, from /blog to /journal, in the site’s English, French, German, Italian and Spanish versions too. Of the 329 sections we found, 114 turned up in the feed alone.

We then checked 71 of the sites with no articles against Google’s index: 6 did have a section, and we had missed it. So 51% is a ceiling, and the true share is a little lower.

A hotel with a live blog publishes 9 articles a year, about one every six weeks. The 9 is a median: line up the 153 hotels by how much they publish, and it is the figure for the one in the middle. Half of them publish no more than that. For 49 the year’s total is 4 articles or fewer. Just 8 hotels of the 669 reach two articles a month.

Some of these hotels publish more than we counted. At 124 of the 153 the articles sit in a feed, and many sites keep only the latest ten there: that is the default in WordPress, under the option that sets how many posts a feed shows. For 41 hotels our count stops at exactly 10. The median stands all the same, because 86 of the 153 hotels are at 9 or fewer. On 8 October 2026 we reopened those feeds: 116 answered, and 88 listed exactly ten pieces.

Google does not ask for any particular frequency. Its guide to helpful content, updated on 5 October 2026, treats it as a warning sign when a site adds many new pages simply to look fresh, and says that doing so does nothing for rankings. The same page says Google has no preferred length for a text.

Top hotels publish least in Italy and Switzerland, among the seven countries where our sample has at least 20 hotels: 31 of 43 Italian hotels and 18 of 25 Swiss ones have no articles section. In the United States the figure is 87 of 202.

Live blogs run the other way: 52 of 202 hotels in the United States have one, 9 of 39 in the United Kingdom, 6 of 43 in Italy, 3 of 33 in France and 2 of 25 in Switzerland. We leave out countries with fewer than 20 hotels, where the numbers are too small to say anything.

These are ranked hotels, the top of the market. The table says nothing about the average hotel in each country.

Half of the top hotels publish nothing, and the explanation our data supports best is that hotels hardly ever hire anyone to write. The season, the obvious suspect, makes no difference.

We checked the season first, because a hotel closed for months would have a fair reason to go quiet for part of the year. The sample has 48 hotels in ski or seaside resorts that live by the season, from Zermatt to Positano: 11 have a live blog, 23%. Among the other 621 the figure is 142, again 23%.

Then we looked at who gets hired. On the same day, 12 August 2026, we read the open vacancies on three hospitality job boards, which cover hotels, restaurants and catering: hosco, hotelcareer.de and caterer.com, about 31,000 vacancies in all.

Hospitality employers almost never advertise for someone who writes. When one does, it is a group: the smallest we verified has four hotels. Vacancies show who is being hired today. They say nothing about who already works in a hotel, or about agencies and freelancers. On 8 October 2026 we repeated the reading, on 28,359 vacancies this time: a search for “copywriter” returned one, and it was for a marketing manager. Roles, pay and team sizes are in our article on who actually does the marketing in a hotel.

A hotel blog goes with more traffic, and our measurement cannot say that the blog brings it. Hotels with a live blog have 6.2 times the estimated visits from searches that do not contain their name: a median of 1,496 a month against 242.

On 12 August 2026 we chose 120 hotels with a live blog and 120 with no articles section, both at random. For each we asked DataForSEO, a service that estimates website traffic, how many visits a month come from Google’s results in the United States, in English. The estimate takes each keyword, a phrase people search for, and multiplies its monthly searches by the share of clicks that goes with the site’s position. We set aside searches containing the hotel’s name, because they measure fame. Data came back for 231 hotels: 117 with a blog, 114 without.

Three things explain at least part of the gap, and the blog is none of them. The first is fame. Counting searches for the name as well, hotels that publish have 5.7 times the visits of those that do not: they were more searched for to begin with, and a hotel that is already visible can better afford to write. The second is language. The estimate looks at Google in the United States, and hotels in the United States, the United Kingdom, Ireland, Canada, Australia, New Zealand and South Africa make up a larger share of the publishers. Within that group the gap narrows to 2.4 times. The third is that the two groups overlap: 36 of the 114 hotels with no articles sit above the publishers’ median, and 30 of the 117 publishers sit below the median of the others.

More articles go with more visits only from 10 a year upwards. Below that, writing 1 to 4 or 5 to 9 makes no difference to the estimate: 881 visits against 711. The 52 hotels with at least 10 articles stand at 2,859.

A before and after on the same site would settle it, and every hotel holds that data for its own site. Search Console is the tool in which Google shows a site which searches bring it visits. Since November 2025 it has separated searches containing the brand name from the rest on sites that receive enough searches; on the others the query filter does the same job once you exclude the name. Compare the clicks from searches without the name before you publish and in the months after: nobody outside the hotel can make that measurement.

Hotels that blog write four kinds of piece: guides for travellers, the hotel’s own stories, awards and press cuttings, offers and event dates. On 8 October 2026 we reopened the 124 feeds of the hotels with a live blog: 116 answered, and 113 held at least one piece from the past year. For each we read the six latest titles, 584 in all, and filed the hotel under the kind that came up most.

Guides answer a question somebody types into Google. On 8 October 2026 the public feed of The Westin Riverfront, a mountain resort in Colorado, held ten pieces published between September 2025 and October 2026, and every title was a first-time skier’s question: what après-ski means, how many days a beginner needs, whether the town is worth a visit. On the same day the feeds of Gangtey Lodge, in Bhutan, and Kruger Shalati, in South Africa, each carried a month-by-month guide to the weather.

The other three kinds talk about the hotel to people who already know it. In 43 of the 113 feeds the typical piece is an award, a press cutting, a festive menu or a room page. So our “live blog” is a generous measure: it counts any site that publishes dated pieces, whatever they are.

Travellers search for a destination and a need together, such as a spa in the Dolomites or parking in Rome, and AI assistants check the practical facts of a stay, such as the shuttle and the opening dates. Both are answered by facts a hotel holds: minutes, hours, prices, a minimum age. On 19 September 2026 we measured searches of this kind with Google Ads estimates for Italy: “hotel spa Dolomiti” gets 5,400 searches a month, “hotel vicino stazione Firenze” (hotel near Florence station) 1,600 and “hotel con parcheggio Roma” (hotel with parking in Rome) 1,300. The words people really use are in our guide to what luxury travellers type into Google.

Assistants look up more than they are asked. On 27 September 2026 we read the searches that Gemini and Perplexity run when a question names a hotel: each hotel drew between 3 and 17, and 475 of the 892 were about something the traveller had not asked, from reviews to the shuttle to opening dates. In our study of answers from ChatGPT, Claude and Gemini, when an answer named one of the world’s 50 best hotels, the hotel’s own site was among the links shown in 24% of cases. The guide on how to get your hotel on ChatGPT and Gemini explains the way assistants read.

A small hotel team can publish nine pages a year, one every six weeks: the pace of the typical hotel with a blog among the world’s best. Choose them from the questions guests already ask. None of these pages guarantees visits: our measurement does not show that they bring any, and Google makes no such promise.

Dates do go missing: in our sample 52 of the 669 hotels publish articles with no date that can be read, either on the page or in the code. Google’s page on publication dates asks for two things: a visible date, and the same date in the structured data.

We build the pages of our hotel guide this way: for each hotel, the facts a traveller needs in order to choose, each with its source and the day we read it.

In August 2026 we read the websites of the hotels on four rankings with a program we wrote, and counted what they publish and how often.

• The sample: 1,420 hotels from four rankings, collected on 12 August 2026. We read the 712 with their own website; 669 gave a result.
• Reading the sites: feed, index pages and the commonest addresses. A live blog needs at least three dated articles, one of them in the 365 days before the reading.
• Known errors: 6 of 71 sites recorded as empty did have a section; two of 15 positive results opened by hand were wrong, and the rule that corrects them is already in the figures on this page.
• Traffic: DataForSEO estimates for Google in the United States, in English, on 240 hotels drawn at random. Searches without the name are counted on each site’s top 200 keywords. They are estimates computed from positions and search volumes, and they see little of a site that lives on searches in other languages.
• Titles: 113 feeds read by hand on 8 October 2026 by one person. Another reader would move a few hotels from one kind to another.
• Limits: the reading is of a single day and of the top of the market, and the association between blog and visits proves no cause.