Ranking on booking sites: what moves a hotel up the list, and what the hotel controls

Booking sites rank properties per search, not once for all. Booking.com's published description says results are first sorted by its top picks, driven by click-through rate, gross bookings and net bookings, which in turn depend on reviews, availability, pricing, policies and content. Commission, payment speed and program membership can also influence ranking; no weights are published.

A general manager searches the hotel's own name for next weekend, sees it on the third page of results, and an hour later receives an email promising top placement for a monthly fee. Both the panic and the promise rest on a misunderstanding of what a ranking actually is. This article explains how to improve booking.com ranking by separating what the platform publishes, what the hotel controls, and what only disciplined testing can settle. You will leave with a five-lever checklist and a monthly test routine.

Why does your hotel rank differently for every guest and every search?#

Because the list is built for that one search. The dates, the room type, the length of stay and the guest's own history change which properties are eligible and in what order they appear.

Eligibility comes before ranking. A hotel with no room for the requested dates, a closed room type or a minimum-stay rule that does not fit is not ranked at all. Availability is the entry ticket, not a tiebreaker. A minimum-stay rule of three nights removes the hotel from every two-night search, no matter how good its reviews are.

Booking.com's own description of its default order (updated 31 May 2025) says results are first sorted by "Our top picks", called "Popularity" in the app, and that the customer's own interactions with the platform are a factor unless they opted out of personalised recommendations. The same page says the importance of each factor changes with the property's features and with how people use the platform. There is no single order to be first place in.

Paid placement sits outside the organic order. A property carrying an "Ad" badge has paid to appear there, so an ad at the top says nothing about ranking.

The practical consequence: when you check your position, fix the search. Same dates, same party size, same room type, same device, and note whether you are signed in. Two colleagues searching the same dates can see different orders on the same afternoon. A screenshot is one guest's view at one moment, not a measurement.

What do booking sites publish about ranking?#

More than most hoteliers assume. In the EU, platforms must describe their main ranking parameters in their terms, and Booking.com publishes a plain-language account of the factors behind its default order.

Regulation (EU) 2019/1150 has applied since 12 July 2020 (Article 19(2)), according to the EUR-Lex consolidated record of the Regulation (opens in a new tab). Article 5(1) requires online intermediation services to set out in their terms the main parameters determining ranking, and the reasons for their relative importance. Article 5(3) adds that where ranking can be influenced in return for direct or indirect payment, the platform must describe those possibilities and their effect on ranking. Article 5(5) requires enough detail for a business user to understand whether, and how, ranking reflects the relevance of what it offers. Article 5(6) exempts the algorithm itself.

Scope matters for a Turkish reader. Article 1(2) covers services offered to business users established or resident in the EU that reach consumers in the EU, wherever the platform sits. It is not a duty on your hotel, and it is not a lever you can pull. It explains why the disclosures exist and why they are worth reading. The rule is also under review: on 19 November 2025 the European Commission proposed repealing it in its Digital Omnibus package, and EUR-Lex still listed it as in force on 4 November 2026. Confirm your own position with a legal adviser.

Booking.com's "How we work" page (updated 31 May 2025) (opens in a new tab) states that to appear high up, a property needs to do well on three measures: click-through rate, gross bookings and net bookings, meaning bookings minus cancellations. It says those numbers depend on many factors, including review scores, availability, policies, pricing and the quality of content such as photos. Other things can also influence ranking: how much commission the property pays, how quickly it usually pays it, and whether it is in the Genius or Preferred Partner (+) program. Only for US properties booked by US customers, whether Booking.com organises payments also counts.

What the page does not give is a formula. No weights, no thresholds, no points. That absence is the most useful thing in it: anyone quoting a precise recipe is quoting something the platform never published.

Which levers does the hotel actually control?#

Five: what is open for sale, the price the guest compares, the content on the page, the policies attached to the rate, and the review record. Everything else on the published list is either the guest's behaviour or the platform's design.

Availability is the first lever. Open dates, open room types, minimum-stay rules and closed-to-arrival settings decide whether the hotel enters the race at all. A room type closed for a Saturday cannot rank for a Saturday stay. Check the calendar for gaps before blaming the algorithm; if you have not finished preparing the listing, read what to prepare before you go live on booking sites first.

Price is the second, understood as the guest compares it: the total for the same dates, with the same meal basis and the same cancellation terms, in the currency they see. A lower headline rate with a stricter policy is not a lower price in that comparison, and it usually converts worse.

Content is the third: photos per room type, room size, facilities and a description that matches what the guest will actually get. Content quality is named in the published description; volume of photos is not. Policies are the fourth. The cancellation window, prepayment, children and extra-bed rules change both how often the listing is opened and how many bookings survive, and cancellations are subtracted in the net bookings measure. Reviews are the fifth: score and recency. A weekly routine is enough for a small team, and it is the one lever that compounds slowly and cannot be bought; see a weekly routine for hotel online reputation.

The listing itself has to be complete before any of this is measurable. A half-finished page makes every later test unreadable.

Does paid visibility help, and how do you judge it?#

It can put the hotel in front of more guests, but it is a purchase, not a ranking repair, and it should be judged by what the hotel keeps rather than by where it appears.

Three different things get mixed together: a visibility program the hotel joins, paid placements carrying an "Ad" badge, and campaign discounts on selected rates. Each has its own terms and its own cost. Booking.com's published description names Genius and Preferred Partner (+) membership among the things that can influence ranking, alongside commission level and payment speed. It does not say by how much, and no weighting is published anywhere.

Read the platform's own terms for what a program asks of the property: a discount on some bookings, a different commission, a badge, extra visibility. A vendor's summary of those terms is not the terms. This article quotes no commission or program figures on purpose. They vary by contract, property and season, and any number printed in a blog post is either out of date or wrong for your hotel.

Judge by net: revenue after the program's cost, per booking, compared with the same room sold without it over the same dates. A higher position with a thinner margin is a worse outcome, not a better one. Watch the interaction with your own website too: a discount given to one channel can undercut the price you show direct, and the guest who checks will find it. For the comparison method, see comparing distribution channels by what the hotel keeps.

Which ranking myths should you drop?#

Any rule that arrives without the platform's own text behind it, and any claim of a fixed formula. The published descriptions give measures and factors, never weights.

The test is simple and repeatable: can you find the claim on the platform's own published description or in its terms? If not, treat it as a hypothesis to test, not a rule to obey. Myths survive because they are cheap to repeat and expensive to disprove. A hotel that closes cheap room types, cuts rates and joins a program in the same month cannot tell which of the three did anything. Some myths are harmless but wasteful; others cost money directly, because they lead to discounts given for a placement nobody measured.

Different platforms publish different descriptions. A rule learned on one platform does not transfer to another, and naming one platform's practice as an industry standard is how bad advice spreads.

How do you test one change at a time?#

Pick one lever, change it for a fixed period, and compare the same measure before and after with everything else held still. The discipline is in the word one.

Choose the measure before the change, not after. Use the figures your extranet already reports for the same defined search or date range: how often the listing was seen, how often it was opened, how many bookings came out. Write the baseline number down. Then change one thing. Photos and price and policy in the same week produce a result nobody can attribute, and the next test starts from an unknown base.

Fix the end date in advance. A week is usually too short for booking behaviour to settle; a month is a workable unit, and comparing with the same month a year earlier removes most of the seasonality. Read the same number you wrote down, not a different one that happens to look better. Record the change, the dates, the measure and the result in one place the next person can read. Twelve recorded tests in a year are an asset; twelve opinions are not.

Accept null results. A change that moved nothing is a finding, and it stops the same experiment being run again next season. The inputs come from the morning routine on the extranet, where availability gaps, closed room types and new reviews surface first; see the morning extranet routine for booking sites.

What to do next#

The ideas fit together in one order: ranking is per search, the platform publishes which measures and factors it uses, five of those factors belong to the hotel, paid visibility is a purchase to be judged on net revenue, and the rest is testing. The next step is small and can be done this week: pick one defined search, write down the baseline figure your extranet reports for it, and change exactly one lever for a month.