User:Superb Owl

Still learning the lay of the land here on Wikivoyage (having mostly been a Wikipedian) so am trying to organize my thoughts and questions while learning about how Wikivoyage works to see if I want to get involved and if so, how.

• flag (every) 'dead link'
  • InternetArchiveBot recently updated to run on Wikivoyage templates, but still some issues regarding accuracy
    • Q: any recent quantification of false positive and false negative rates?
• Show dead link flags to every user
  • At the very least, can we show it to logged-in users regardless of whether they are in editing mode or not?

• Cybersecurity: Prevent many visitors going to risky external links to websites with technical debt or that have been usurped by a scam/spam site
• Usability: cut out frustrating dead ends for users who click on bad links
• Accuracy: Visible and accurate flags also can warn contributors that the page has not been updated in a while and to help improve it or move on to other resources

• Why not link every bolded page name (first mention in the lede of the article) to its equivalent Wikipedia article?
• Add a 'sister' template in 'Read' (Wikipedia) and 'Watch' (Commons) sections of articles? In the 'Learn' sections with other official tourism sites?

  Read about San Francisco on Wikipedia

  Media of San Francisco on Wikimedia Commons

  • • Existing Wikidata link requires knowing to click on the three vertical dots (which do not exist on Mobile) before scrolling down 24 lines of text (this change happened when Vector-2022 went live).
  • Has anyone tried creating a bot to match Wikipedia articles that exist in the Wikivoyage template field for Wikipedia?
  • Why do we not link to Wikipedia articles instead of external links in some (or all instances where a Wikipedia article exists) on Wikivoyage? I assume that the pre-fork Wikitravel community never considered linking to Wikipedia but curious if there has been any shift in community consensus on this idea over time?
    • some benefits of wikilinks over external links:
      • Less linkrot, better cybersecurity experience for user (who likely knows how to find external link from Wikipedia articles)
      • More objective than a website that tends towards promotional (discourages WP:REFSPAM)
  • Breakdown in offline usage: How many times per month is the "Printable version" button pressed? How many times per month is the "Download as a PDF" button pressed? How many people use Wikivoyage data in an offline app?
    • If offline apps that integrate both Wikivoyage and Wikipedia are growing in popularity, then that strengthens the case for wikilinks (that would work offline) over external links (that would not)
  • Cooperating with OpenStreetMap on POI

  How is Wikivoyage used?

  [edit | edit source]

  We have basically no visibility into how many people actually use Wikivoyage and why (unless there is some research I have not seen).

  User traffic

  [edit | edit source]
  Tracked in Phabricator ^(?)
  Task T440733

  This spike in 'user' traffic starting at the end of 2024 strongly correlates with spike in 'automated' traffic, which suggests significant undetected bot usage as the main/only driver of growth since then. (Wikinews 'user traffic' also spiked inexplicably at the same time). Also, how else would Seychelles, a country of 121k people, ever rank second only to the U.S. in using English Wikivoyage around August 2026? Bot traffic and use is not necessarily bad, it's just that some (but not all) users would like to have a decent sense of human vs. bot traffic as it changes where many editors would donate their time and attention. The spike in traffic is especially misleading given that it could just be from an AI scanning hundreds of pages before delivering one answer and not generating much impact at all.

  Travel information is of especially high interest to AI companies

  [edit | edit source]

  Wikivoyage could be something that generates revenue for the Foundation through its API services even as its human readership is likely to continue declining

  1. Travel websites have seen greater drops in traffic than most other websites since the rise of AI (which makes any unexplained Wikivoyage spike is users that much more suspicious) - Tripadvisor, for example, lost 1/3 of its traffic between 2024 and 2025 but as of 2026 gets 60% of its revenue from partnerships.
  2. A Wikimedia Enterprise technical partnerships lead reported in early 2026 that when promoting Wikivoyage to AI companies, that Wikivoyage was the project that they had not heard of that they were most excited about.

  Future wishlist/brainstorm

  [edit | edit source]

  Would Wikivoyage consider narrowing its scope/target audience to distinguish itself more from its many commercial competitors, some of which use wisdom of the crowd - and have much bigger crowds (TripAdvisor has 1 billion reviews) and thus more wisdom... Wikivoyage's competitive advantages could include not being part of the surveillance economy and offering a more (cyber)secure experience that also focuses more on quality than quantity of information. It could also embed preferences for nonprofit/public attractions over equivalent for-profit ones, small businesses over chains, attractions that benefit local community (contribute to the tax base instead of relying on subsidy), among other editorial decisions?

  Any exploration of training an LLM to retrieve and summarize the information for users and be able to create custom itineraries, etc. based on information from free knowledge projects. Apparently travel is an area that has really shifted toward chatbots and people might want a more private/ethical LLM experience...