Using the same pricing software as your rivals: what competition law says

Algorithmic pricing means software that sets or suggests room rates from data. Using the same tool as rivals is not automatically unlawful. What matters is whether the tool pools rivals' confidential data, and whether hotels agree to follow its output. Article 4 of Türkiye's Law No. 4054 bans price agreements and concerted practices.

A revenue manager sits through a demo of a rate tool that "reads the market", and the vendor mentions that most of the hotels in the region already use it. Two questions decide whether that is ordinary software or a legal problem: whose data produces the number, and what have the hotels agreed to do with it. After reading, you will know which mode of algorithmic pricing your contract buys, what the courts and the Turkish Competition Board have actually said, and which questions to send the vendor in writing before you sign.

What is algorithmic pricing, in plain words?#

Algorithmic pricing is software that sets or suggests prices from rules and data, with either a person or the software making the final call. The legal question is not the algorithm itself; it is where the data comes from and what the hotels using it have agreed to do.

Separate three modes and find out which one your contract buys. In rule-based mode you set the floor, ceiling and rules, and the software applies them. In recommendation mode the software suggests and a revenue manager approves. In automatic mode the software applies changes within limits you set. The difference matters because the more the software decides, the more you need to know about its inputs.

Keep algorithmic pricing apart from dynamic pricing in hotels and who sets the rates. Dynamic pricing is the commercial practice of varying rates with demand; algorithmic pricing describes who or what produces the number. A hotel can price dynamically on a spreadsheet, and a hotel can run an algorithm that never varies its price.

A tool is not risky because it is shared. It becomes a competition issue when it carries other hotels' non-public data, or when the contract or the habit of the team turns a suggestion into a common price. Ask what the software optimizes: your own revenue, a market-level target, or a target defined across all its clients in your area. The answer belongs in writing.

Ask whether the tool is a rule set with a scoring layer, and ask for a plain description of the calculation, in sentences, not a dashboard. Then write down who approves a change and how often the team approves without reading. Approval that is automatic in practice is automatic in effect.

What did the US court decide in Gibson v. Cendyn?#

The Ninth Circuit affirmed the dismissal of the guests' claim, holding that competing hotels independently buying licences for the same recommendation software, followed by higher prices, does not by itself state a claim under Section 1 of the Sherman Act. The case is Gibson v. Cendyn Group, LLC, No. 24-3576, an opinion by Judge Bea filed on 15 August 2025, brought by a putative class of Las Vegas Strip hotel guests (United States Court of Appeals for the Ninth Circuit, opinion, 15 August 2025 (opens in a new tab)).

The court's reason was causation. Section 1 requires a causal link between the contested agreement and a restraint of trade in the relevant market. Several competitors choosing the same provider, plus higher prices, is not enough to require rule-of-reason scrutiny.

The caveats matter more than the headline. The plaintiffs dropped their hub-and-spoke argument on appeal, so the court did not address an alleged agreement among the hotels to follow the recommendations. The plaintiffs did not allege that the software pooled, shared or used one hotel's confidential information in recommendations made to another hotel. And the court said it would "undoubtedly" violate Section 1 for competing hotels to agree among themselves to follow a third party's pricing recommendations, with non-binding recommendations offering no protection. It also expressly declined to decide whether delegating binding pricing decisions to a third party is a restraint of trade.

This is US federal law and a decision about whether a claim was properly stated. It does not interpret Law No. 4054 and it is not a safe harbor for a Turkish hotel.

What does Turkish competition law prohibit?#

Article 4 of Law No. 4054 prohibits agreements between undertakings, concerted practices, and decisions and practices of associations of undertakings that have as their object or effect the prevention, distortion or restriction of competition, including fixing prices or the elements that make up a price (Cumhurbaşkanlığı Mevzuat Bilgi Sistemi, Law No. 4054, Article 4 (opens in a new tab)).

The article lists the cases it covers: fixing the purchase or sale price of goods or services, the elements that make up the price such as cost and profit, and any purchase or sale conditions; sharing markets or market resources; and controlling supply or demand. The object-or-effect structure matters. If the object is price fixing, the authority does not have to show the harm; a conversation that never produced a higher rate can still be the infringement.

The presumption matters most in daily hotel life. Where an agreement cannot be proven, price changes or market conditions resembling those in markets where competition is restricted create a presumption of a concerted practice. The escape route is evidence, not explanation: each party can rebut the presumption by proving, on economic and rational facts, that it did not engage in a concerted practice. "We all use the same tool" invites the next question rather than closing it.

Associations are inside the article, not beside it. A regional hotel association that recommends rates, or circulates a rate survey for members to align on, is making a decision of an association of undertakings.

The Board has already examined pricing terms in this market. In decision 17-01/12-4 of 5 January 2017 it ruled that price and room-availability parity clauses and best-price-guarantee terms in a major online booking platform's contracts with accommodation facilities infringed Article 4, and it fined the platform (Rekabet Kurumu, final decision announcement, 5 January 2017 (opens in a new tab)). For what those clauses mean in practice, see rate parity clauses and what the official decisions say.

None of this decides whether any particular tool is lawful. That depends on the facts of your contract and your conduct, and belongs with your legal adviser.

What has the Competition Board said about pricing algorithms?#

In its Amazon Türkiye decision of 18 April 2025, the Board said pricing algorithms can bring lower prices, but because they automate decisions and increase transparency they also make it easier for competitors to monitor each other, agree, and punish deviations. The decision is 25-15/348-164, file 2023-1-038, and it examined an automatic pricing mechanism offered to sellers on a multi-category e-marketplace (Rekabet Kurumu, decision 25-15/348-164, 18 April 2025 (opens in a new tab)).

The Board accepted the benefit side: pricing algorithms can produce lower prices, and the technology is not unlawful in itself. The risk side is transparency plus automation. Competitors can watch each other, reach an agreement on price and other strategic factors, observe deviations and retaliate. That can produce information exchanges, secret agreements or concerted practices.

The Board added a point aimed directly at shared vendors: competitors choosing the same third-party algorithm instead of developing their own may have the potential to give rise to collusion. On responsibility, it set a threshold: a user of a third party's anticompetitive algorithm must at least be aware of that conduct or be able to reasonably foresee it. A vendor's public claims about what the tool does are part of what you could foresee.

The Board also said algorithmic pricing may not amount to an infringement by object in every case. It unanimously found that Amazon Türkiye had not infringed Article 4 and that no administrative fine was needed, and the decision records the company's statement that the mechanism would be closed in Türkiye as of 10 February 2025. Read the reasoning as a framework for any shared pricing tool, hotels included. It is not a finding about hotels, and it does not settle your case.

Public data, your own data, pooled data: why the difference matters#

A tool fed by public data and your own numbers is a calculation. A tool fed by rivals' non-public data is a channel between competitors, and the channel can be the problem even when nobody agreed to anything.

Public data is your competitors' published rates, public event calendars and published tourism statistics. Everyone can see it, so using it is ordinary market intelligence. Your own data is your bookings, pace, cancellations, costs and guest mix, and nothing leaves the hotel in a form that identifies a rival.

Pooled or confidential data is the category that changes the legal character of the tool: other hotels' forward rates, occupancy, pick-up, discount practice or willingness to pay, or the recommendations the tool gives them. The mechanism is uncertainty reduction. Exchanging forward-looking, non-public data removes the guesswork that competition is supposed to create, and the exchange itself can be a concerted practice without a written agreement or a meeting.

Ask about aggregation thresholds: at how many contributing hotels does a pooled figure appear, and can a participant be identified from it? In a small resort comp set, "aggregated" can still be one hotel's number. Watch correlation too. The same tool, the same default settings and the same comp set produce similar prices, which is exactly the pattern that feeds the Article 4 presumption, even where nothing was agreed. For how to define that set fairly, see choosing a competitive set and comparing rates fairly.

The practical test: take one rate suggestion and write down where each input came from. If you cannot, you cannot answer a regulator's first question.

What should you ask the vendor before you sign?#

Ask for the data inputs in writing, and for the contract to state whether recommendations are advisory or binding. Then take both answers to your legal adviser before the renewal date, not after.

A clause that obliges you to follow output is the clause counsel will want to read. So is the comp set question: a vendor that serves most of a small market is a different proposition from one that serves a handful. For the day-to-day side of reading suggestions, see what a revenue manager should check in AI rate recommendations.

Three questions belong to counsel rather than the vendor: does your comp set definition itself create an exchange, does participation in a shared benchmark create exposure, and what record should the hotel keep of approvals?

What are the everyday risks, even without software?#

Most Article 4 exposure in hotels comes from conversations rather than code: price talk with neighboring hotels and rate discussions inside associations.

Association meetings are covered. Decisions and practices of associations of undertakings fall under Article 4, so a recommended minimum rate or a members' rate survey discussed in the room is the classic risk, whatever the minutes say. Informal channels count the same: general manager messaging groups, phone calls between neighbors, front-desk conversations about what the hotel across the street is charging next season.

Benchmark participation is an exchange. Sending your forward rates into a pool and receiving rivals' in return is the mechanism, and receiving is enough to be part of it. The safe subjects are your own costs, your own demand signals and public data. The unsafe subjects are future prices, discounts, capacity, and who will take which segment or group.

Give revenue staff a three-line rule: we discuss our own numbers, we do not discuss future prices with competitors, and we leave and record if the conversation turns. Record keeping is your defense. A short dated note of who was present and what you said is worth more than a recollection two years later.

What to do next#

Before the next demo or contract renewal, send the vendor the data questions above in writing and put the answers in front of counsel. The parts of this article fit together as one test. A tool is a calculation when it uses public data and your own; it is a channel when it carries rivals' non-public data; it is an agreement when hotels commit to follow it. Only the first is comfortably outside Article 4.