What's happened in places like Levi.

The pattern beneath the cases

Before the individual stories, the structural shift that connects them.

At the centre of every destination case below is the same economic mechanism:

Distribution power > Product ownership.

The company that controls how visitors discover, compare, and book a destination dictates the terms under which the destination’s actual product — accommodation, experiences, food, transport — operates. This is not a tourism phenomenon. It is the same pattern that has reshaped:

• Hotels — where Booking Holdings (Booking.com, Priceline, Agoda, Kayak, OpenTable) and Expedia Group together control a major share of global online accommodation bookings, often above 60% in mature markets
• Restaurants — where Deliveroo, Uber Eats, and similar platforms now mediate a large share of urban food orders, taking 20–30% per transaction
• Retail — where Amazon’s marketplace dominates how products are discovered, with sellers paying for visibility on the same platform they depend on
• Taxis — where Uber and equivalent platforms restructured the entire driver-passenger relationship within a decade

Tourism follows the same pattern because it has the same characteristics: fragmented supply, high emotional purchase, and high need for trust. Whenever those three traits combine, platforms emerge to aggregate the trust on behalf of buyers — and capture the value of doing so.

For tourism specifically, the verifiable global picture is:

• Booking Holdings and Expedia Group control a majority share of global online accommodation bookings in many markets
• Airbnb has more than seven million active listings worldwide, increasingly operated by professional hosts rather than individuals
• Experience platforms such as GetYourGuide and Viator typically take 20–35% commission per booking

The four cases below show what happens when this structural shift meets a specific place. Each illustrates a different stage of the same trajectory.

Case 1: Iceland — the closest geographical and structural parallel

What happened

Between 2010 and 2019, Iceland’s annual visitor numbers grew roughly fourfold, from approximately 500,000 to more than 2.3 million. The growth was almost entirely international, driven by long-haul flight expansion (particularly the rise and fall of WOW Air), favourable currency conditions following the 2008 financial crisis, and an aggressive destination marketing push.

The structural conditions of the growth matter as much as the numbers. Visitors arrived overwhelmingly through:

• Online travel agencies for accommodation
• A small number of large tour operators offering packaged Golden Circle and South Coast itineraries
• International airlines whose route economics shaped which markets sent visitors

Local providers, particularly small ones, found themselves with limited direct access to demand. Visibility on the channels that actually delivered customers required platform commission and paid promotion.

What was observed

• Experience standardisation. The Golden Circle — a fixed loop of Þingvellir, Geysir, and Gullfoss — became the template product. Coach-loads of visitors moved through the same sites at the same times, produced the same photographs, and returned to Reykjavík for the evening.
• Pricing under platform competition. Initial price increases gave way to compression as platforms enabled direct comparison. Operators reported margin pressure even as gross revenue rose.
• Independent provider channel dependence. Small operators became increasingly reliant on Viator, GetYourGuide, and Google Ads to be discoverable at all.
• Documented direct-booking response. A specific Icelandic adventure company shifted its business model to direct sales, working with journalists and PR rather than travel agents, explicitly to avoid commission costs and improve economic viability. This is one of the few well-documented cases of a small operator successfully reducing platform dependence in this period.

What’s transferable to Levi

• Iceland is geographically and structurally the closest parallel. Small population, Arctic positioning, sudden international growth, OTA-heavy distribution, fragile environmental capacity. The trajectory Iceland followed between 2010 and 2019 is the trajectory Levi appears to be on now, roughly a decade later.
• The standardisation pattern — Golden Circle as a fixed template — has its Levi equivalent in mass safari products, packaged Northern Lights hunts, and the increasing professionalisation of “authentic Lappish experience” formats.
• The direct-booking case study demonstrates that disintermediation is possible at the operator level, but it requires deliberate strategic choice and is not how the system defaults to working.

Where the parallel breaks down

• Iceland’s growth was driven by a specific moment of long-haul flight economics that may not repeat for Lapland in the same form.
• Iceland’s response to overtourism was forced largely by COVID-19, which reset visitor numbers for two years and gave the destination unintended breathing room. Levi may not get a similar pause.

Key takeaway

Authenticity still exists in Iceland. But it is no longer the default experience. The default is the standardised version, sold at scale, that the platforms make most visible.

Case 2: Barcelona — what platform saturation does to a residential city

What happened

Barcelona experienced sustained tourism growth from the early 2000s, accelerated by Airbnb’s arrival in the early 2010s. The city became one of Europe’s highest-density tourism destinations. By the late 2010s, Airbnb listings had transformed entire central neighbourhoods — particularly Barri Gòtic, El Raval, and Barceloneta — from residential areas into de facto tourist accommodation zones.

The dynamic was not gradual. Once residential apartments became more profitable as short-term rentals than as long-term housing, the conversion happened quickly. Professional operators acquired multiple units. The platform’s structure rewarded scale. Individual hosts renting a spare room became a small minority of the listings, while professionalised operators dominated.

What was observed

• Housing pressure and resident displacement. Long-term rents rose sharply in central districts. Working residents were priced out of neighbourhoods their families had lived in for generations. The displacement was demographic as well as economic.
• Professionalisation of hosting. The image Airbnb sold publicly — “meet your local host” — bore decreasing resemblance to the operational reality of much of its inventory.
• Cultural dilution in central areas. Local shops and services that depended on resident trade closed. They were replaced by businesses serving tourist throughput.
• Regulatory backlash. Barcelona’s city council introduced increasingly aggressive measures: a moratorium on new tourist licences, registration requirements, fines for illegal listings, and eventually a 2024 announcement that all short-term tourist apartment licences would be eliminated by 2028.
• Anti-tourism sentiment. Public protests, slogans aimed at visitors (“tourists go home”), and a wider political backlash against the visitor economy as a whole.

What’s transferable to Levi

• Barcelona shows what happens when the platform’s logic — listings as professional inventory rather than residents-with-spare-rooms — meets a residential community. The intermediate stage is where local providers and residents notice the change without yet having levers to respond. That stage is happening in Levi now, in early form.
• The political dynamic is also transferable. When residents lose access to housing and neighbourhood character simultaneously, the political response is rarely measured. Anti-tourism sentiment, when it arrives, is hard to dial back.

Where the parallel breaks down

• Barcelona is a major city; Levi is a village of fewer than a thousand permanent residents. The mechanisms are similar but the scale and context are different.
• Barcelona had decades of accumulated tourist infrastructure before Airbnb arrived. Levi’s transformation is happening on a much smaller and more recent base, which may make either intervention easier (smaller systems can change faster) or harder (less institutional capacity to push back).

Key takeaway

Platform growth → supply professionalisation → cultural dilution. The pattern is reproducible because the platform’s economics reward it. Without intervention, the trajectory does not reverse on its own.

Case 3: Bali — when discovery becomes algorithmic

What happened

Bali’s tourism economy is highly fragmented on the supply side — thousands of small accommodation providers, experience operators, restaurants, and tour guides — and concentrated on the discovery side through a small number of platforms (Booking.com, Airbnb, GetYourGuide, Viator) and one dominant inspiration channel (Instagram).

The discovery layer is what produced the structural change. Visitors increasingly chose where to go, what to do, and what to photograph based on what surfaced on Instagram or ranked well on platform search. Operators noticed. The product changed in response.

What was observed

• Experiences redesigned for photographs. Swing-over-jungle setups, infinity pools framed against rice terraces, “Instagrammable” café aesthetics. The product optimised for the medium of its discovery rather than for substantive experience quality.
• Repeatability over distinctiveness. When one operator’s swing photograph went viral, dozens of others built the same swing. The visual template became the product. Geographic specificity dissolved — the same aesthetic could be produced anywhere.
• Price competition on platforms. Once products converged on shared visual templates, they competed on price. Operators offering the same swing at a lower commission climbed the rankings.
• The disappearing local product. The substantive Balinese cultural experience — temple ceremonies, traditional weaving villages, agricultural rhythms — became harder to find for the average visitor, not because it disappeared but because the discovery layer surfaced something else first.

What’s transferable to Levi

• Bali shows the endpoint of algorithm-driven discovery applied to a fragmented supply base. The Northern Lights photograph, the husky safari pose, the glass igloo aesthetic — these are already standardising in Lapland for the same reasons.
• The lesson is structural: when discovery is mediated by a platform optimising for engagement, the products optimised for engagement displace the products optimised for substance. The platform is not malicious. It is doing exactly what it was built to do.

Where the parallel breaks down

• Bali’s tourism economy is much larger and more diverse than Levi’s. The aggregate effect is more visible because the volumes are larger.
• Bali’s specific Instagram-driven dynamic may be replaced by other discovery patterns (TikTok, AI travel agents) before Levi reaches the same point. The lesson is the structural mechanism, not the specific platform.

Key takeaway

When discovery is algorithm-driven, uniqueness declines. The destinations that remain distinctive are the ones whose operators retain enough independence to refuse the optimisation pressure.

Case 4: Amsterdam — what happens when destinations decide to act

What happened

Amsterdam faced a version of Barcelona’s pressures from the early 2010s onward: rising visitor numbers, Airbnb saturation in central neighbourhoods, housing pressure, and increasing tension between visitor economy and resident quality of life. By the late 2010s the city had become one of the most-discussed European examples of overtourism.

Unlike many destinations that publicly worried about overtourism without intervening, Amsterdam progressively chose to act. The interventions accumulated over more than a decade.

What they did

• Short-term rental caps. Initial restrictions limited Airbnb-style rentals to 60 nights per year per property. This was later tightened to 30 nights, with mandatory municipal registration and significant fines for non-compliance. Some neighbourhoods were exempted from short-term rentals entirely.
• Tourist tax increases. Amsterdam now has one of the highest tourist taxes in Europe, calculated as a percentage of accommodation cost rather than a flat per-night fee.
• Cruise ship limits. Cruise traffic into the historic centre was restricted, with the main cruise terminal scheduled to be relocated.
• “Stay Away” marketing. A deliberate, targeted campaign aimed at deterring specific visitor segments — particularly young British men associated with stag-party tourism. Public marketing actively designed to reduce certain kinds of demand.
• Coffeeshop and red-light district restrictions. Measures to reduce the city’s appeal as a hedonistic-tourism destination.
• A cap on overall visitor numbers. In 2024 the city formally adopted an upper limit on annual visitor nights — a hard cap, not a target.

What was observed

• Some of the worst pressures eased. Housing-market data suggests short-term rental conversions slowed; resident sentiment improved in surveyed central neighbourhoods.
• Tourism revenue did not collapse. Visitor numbers remained substantial. The composition of visitors shifted somewhat toward longer-staying and higher-spending segments.
• The political consensus held. Despite predictable pushback from accommodation and hospitality lobbies, the measures survived multiple electoral cycles.
• Implementation was difficult and slow. Most measures took years to design, pass, and enforce. Several initial versions were too weak and had to be tightened.

What’s transferable to Levi

• Amsterdam shows that the trajectory other destinations follow is not inevitable. Intervention is possible. The mechanisms — caps, taxes, registration requirements, deliberately deterrent marketing — are operational, not theoretical.
• The political feasibility lesson matters most: measures that look unimaginable when first proposed become normal once implemented. Cap, tax, register, exclude, deter. The instruments exist. The question is whether the political will to use them does.
• Amsterdam also shows the cost of acting late. Many of the housing displacements were irreversible. The cultural dilution of central neighbourhoods could not be undone. Earlier action would have prevented harms that the later action could only contain.

Where the parallel breaks down

• Amsterdam is a city with strong municipal capacity and a clear political mandate to regulate visitor economies. Kittilä municipality is much smaller and faces a different political constraint set.
• The visitor mix is different. Amsterdam’s pressures came from short-stay urban tourism; Levi’s pressures come from longer-stay nature-and-experience tourism. The mechanisms transfer but the calibration differs.

Key takeaway

Destinations on this trajectory are not passengers. The path is structural but not inevitable. Amsterdam shows what serious intervention looks like — and how much harm can be contained when destinations choose to act before the worst pressures become irreversible.

Where this leaves Levi

Each of the four cases is at a different stage of the same pattern.

• Iceland shows the trajectory in full: rapid international growth, OTA dominance, experience standardisation, and the disappearance of the “default” authentic experience. Levi is in the middle of this curve.
• Barcelona shows what happens when Airbnb saturation meets a residential community: housing pressure, neighbourhood transformation, and an angry political response that arrives too late to prevent the worst harms. Levi has early signals of this but has not yet reached saturation.
• Bali shows the endpoint of algorithm-driven discovery applied to fragmented supply: products optimised for the medium of their discovery rather than for substance. Levi is starting to show this in specific product categories.
• Amsterdam shows what intervention looks like when a destination decides to act: caps, taxes, restrictions, deterrent marketing, and the political work to make all of them stick. Levi has done none of this yet.

The point of this page is not to predict that Levi will follow any one of these paths. It is to make the trajectories visible — and to show that the choice of which trajectory to follow is, at least in part, a choice destinations make.

The destinations that did nothing got Iceland and Barcelona. The destination that chose to intervene got Amsterdam. The destinations whose discovery layer was algorithmic and unmanaged got Bali.

Levi’s choice has not yet been made.

Each case study is drawn from publicly documented sources. Specific data points (visitor numbers, regulatory dates, commission percentages) will be verified to original sources before this page is presented as final. Where verification is pending, the published version will mark it. See also: The Levi Tourism Model, Frequently Asked Questions, Research Methodology.