How we rank tours and tickets

Every ranking on this site comes from one score, computed at build time from the Viator Partner API and applied identically to every city we cover. No operator pays for placement, no position is editorially promoted, and the same code produces the city rankings, the category lists and the activity finder. This page is the whole method.

What decides the order of a ranking

A single number orders every list: the average traveller rating, weighted by how many people left one. A product with a 4.9 backed by eight thousand reviews ranks above a flawless 5.0 left by twelve, because the first is evidence and the second is a coincidence.

Concretely, each rating is pulled towards 4.35 until its review volume outweighs 150 reviews. A product with far more reviews than that keeps essentially its own rating; one with a handful is treated as close to average, because that is all its reviews can prove. The figures shown on a card — rating, review count, price — are the ones the API returned at the last build, never an estimate or a rounded-up headline.

Why some tours never appear on the site

Ranking well is the second filter. The first is whether a product belongs on a page about what to do in a city at all. Three rules run before anything is scored.

A review floor. A product needs at least 50 reviews to enter the catalogue. Below that there is nothing to compare, and padding a count with unproven listings would make the comparison look richer while making it worse.

Intent filters. 6 rules remove products that answer a different question: cruise-terminal shore excursions, airport and port transfers, chauffeur hire sold as a tour, station luggage lockers, SIM cards. They are real products, sometimes well rated, and none of them is an answer to “what should I book in Rome”. A shared transfer from the Civitavecchia pier was ranking second in Rome before these rules existed.

Price credibility per experience type. A price far outside what a theme costs means the product has been filed under the wrong one. A “skip-the-line ticket” below €12 is a luggage locker; one above €200 is a private VIP tour. The product loses that category rather than the site, so it can still appear under a theme where its price makes sense, and is dropped only when no theme fits.

How a top ten avoids being the same ticket three times

Pure score is the right order inside one experience type and the wrong one across several. Cooking classes rate 4.95 and above because everybody enjoys them; a Colosseum ticket rates lower because of queues, crowds and weather. Ranked flat, Rome's top ten is ten pasta classes, which is not what anyone opens that page for.

A city's headline ranking is therefore built under four rules, applied in this order.

1. Positions one to five go to what the city is known for. Landmark entry, priority access and museum tickets, and nothing else, so a page about Rome opens with the Vatican and the Colosseum rather than with a cooking class. Immersive shows and multimedia attractions are held out of these five whatever they score: they sell an admission ticket, but they are not what the city is famous for. They remain eligible from position six down.

2. At most two rows for the same attraction, across the whole ten. And the second only when it is a genuinely different purchase — a guided visit against a timed-entry ticket — because that is the choice you are actually making. Three versions of the same Vatican tour is one row of information printed three times. The rule groups day-trip destinations the same way, so three Pompeii excursions cannot take three of the ten.

3. At most two rows per experience type in the open positions. Positions six to ten are the best of everything else, and this cap is what keeps them from becoming five food tours in a city that happens to have hundreds of well-rated ones. It counts the experience types shown in the "by experience" table, so the rule cannot be dodged by a product filed under a narrower label.

4. Nothing under 250 reviews, whatever the rules above want. A product with eighty reviews can out-score one with eight thousand on a good month, and a diversity rule that reaches into thinly reviewed listings to fill a gap is filling it with noise. When an experience type has nothing above this floor it places nobody, and the row goes to the next best product instead.

There is a consequence worth stating plainly, because it is a real trade and not an accident: a lower-scoring product can appear in a top ten ahead of a higher-scoring one, when it is the best of an experience type that would otherwise be missing and the higher-scoring one is the third of a type already represented twice. The ranking is not a leaderboard of the ten highest scores in the city. It is the ten rows that answer "what do I book here", ordered by score within each of the two blocks. Every score, review count and price is shown on the row itself, so you can always see what a diversity rule cost.

The same reasoning governs the rest of a hub page: an experience type needs at least 8 activities to be shown as a comparison at all. Below that its products still count in the city total and can still reach the top ten, but there is no row and no page, because two listings side by side is not a comparison.

No. There is no sponsored placement, no paid inclusion and no advertiser relationship with any tour operator. Nothing an operator can do moves a product up a list.

The site earns a commission when a booking is completed on viator.com after a click from here, at no extra cost to you. That commission is the same percentage whichever product you book, so it gives us nothing to gain from ordering a list any particular way — and the ranking code has no access to it. Outbound booking links are the affiliate URLs the API returns, used exactly as returned, marked rel="sponsored" and opened in a new tab. Details are in our privacy policy.

How often are prices and ratings updated?

The catalogue is rebuilt weekly from the Viator API and the site is regenerated with it, which is why every ranking carries the date of the build that produced it. The current one covers 2,111 activities and 1,670,187 traveller reviews across 15 destinations, refreshed on August 21, 2026.

Prices move between rebuilds, and a “from” price is the lowest across departure dates rather than the price of any particular date. Treat the figure on a card as what the product cost at the last refresh, and the price on viator.com as the one you will pay.

Spotted something wrong in a ranking?

Mis-filed products are the failure mode of any catalogue this size, and the ones described above were all found by reading the output rather than by trusting the pipeline. If a listing looks wrong on a page, write to contact@toureditor.com with the link and we will check the rule that let it through. You can see the method applied on the Rome tours and tickets ranking.