Baidu Put a Flight Card in Its Search Results. What the Click Is Still Worth

🎫 A booking flow that starts in the results page

On 12 October, Baidu App and Umetrip announced a partnership. Umetrip is the travel app run by TravelSky Mobile Technology, a subsidiary of the company that operates China's civil aviation reservation systems. What the partnership puts in place is a flight card inside the Baidu App search results page.

Search the local term for "flight", and the results page presents a query-and-booking card. Choose an origin, a destination and a date, and the card shows live flights, fares and remaining seats, sourced from Umetrip's aviation data. Tapping to book hands the user to Umetrip's booking flow to complete payment. Reported plans extend the arrangement later to flight status, seat selection, baggage and airport transfers.

Most Chinese coverage of this described it as booking "without leaving the search box". The precise version is narrower and worth holding onto: the query happens inside the results page, and the payment happens at Umetrip. The user does not have to shuttle between several platforms to compare and switch, which is what the original report actually says. That is a real shortening of the path. It is not the elimination of a handoff.

🔍 What is new, and what is being exaggerated

Two things are being treated as one, and separating them matters.

The interactive results area is not new. Baidu App has been moving in this direction for most of the year: a local-life section, and on 20 August a card that made the whole results page interactive. A search entry that returns something you can act on, rather than a list you must work through, was already the direction.

What is new is the connection of live transaction data from a booking system into that area. An interactive card that shows you a result is a formatting change. A card that holds current fares and remaining seats, and can start a purchase, is a change in what the results page is for.

The trade coverage also overstates the outcome. Headlines have described a booking solved inside the search box; the same articles then explain that tapping to book opens Umetrip's booking flow. Both statements are in the same pieces. Only one of them is accurate.

🤝 The other half of the handshake: Umetrip was already building this

The partnership makes more sense read from Umetrip's side, and here the sourcing improves: a China Economic Times interview with a Umetrip vice-president sets out the strategy in the company's own words.

Umetrip launched its official airline direct-sales platform in July 2025, with Air China, China Southern and China Eastern on board from the start. By January 2026 it had all domestic carriers. Its stated terms are no markup on airline prices, no agency fee charged to airlines, and a set of commitments around transparent pricing, no bundling, airline-consistent refund rules and no price discrimination by user profile. Umetrip describes the ambition as becoming the airlines' second official homepage, and the strategic move as shifting from a single information-service product to "information service plus a one-stop transaction platform".

That last phrase is the important one. Umetrip's own strategy is to stop being an app people check and become a place people transact. Baidu's results page is a very large place where people are already expressing travel intent. The two strategies meet in exactly the arrangement announced.

So this is not Baidu adding a feature on a whim. It is two parties each moving toward the other's territory, and the space they meet in is the search results page.

🖱️ When the results area completes the task, the click loses its monopoly

To be precise about the mechanism, because the loose version is wrong.

The card is not an advertisement, and it does not occupy an ad slot. Paid search advertisements on Baidu still exist and still occupy their positions. Nothing about the ad auction changed on 12 October.

What changed is where the results area sends a user. Before, a search for a flight produced links, and the user chose one — or chose an ad. Now the results area contains a card that can hold current fares and start a purchase. The click has not been abolished. It has acquired a default destination that the advertiser does not control and did not bid for.

That is the part worth dwelling on. A paid search placement is bought on the assumption that the click is the scarce thing, and that the platform's job is to route it. When the results area can complete the task itself, part of the demand never reaches the point where routing is a contested decision. The user's intent was expressed, and satisfied, above the line where anyone was bidding.

The honest limits: this is one card, in one category, on one surface, and the payment still leaves the page. No figure has been published for how many flight searches use the card, or what happened to click volume around it. Anyone quoting a number here is inventing it.

🗂️ Which categories get a card, and which keep the click

The question an advertiser should actually ask is not what happened in travel. It is whether their own category can become a card.

Two conditions seem to govern it. First, the transaction has to be standardisable — the same query and the same decision for every user, with a small set of variables. Origin, destination, date, price. Flight booking is close to the ideal case. Second, an authoritative data source has to exist. Umetrip is not an ordinary partner; TravelSky operates the reservation systems through which Chinese airline ticketing flows, which makes it approximately the source of record.

Those conditions are demanding, and they point in a useful direction. Categories where a standardised, checkable answer exists and a single authoritative supplier can be identified are candidates: flights, hotel availability, cinema tickets, appointment slots. Categories where the decision depends on specification, configuration, negotiation or verification are poor candidates, because the card would have to reproduce an argument rather than a fact.

For an overseas brand entering China, the near-certain reading is that you will not be the card. The question becomes which side of the line you are on.

🎯 What this changes for an overseas brand's China plan

Three adjustments, none of which requires assuming the card spreads.

Find out whether your category already has a card, before you forecast clicks. Search your own category terms in Baidu App and look at what the results area does. This is a ten-minute check with a direct bearing on expected click volume, and it is better done before a media plan is built than after.

Accept that in cardable categories, the card takes the simplest query. A user who wants a price and a date will use the card. What remains for a paid placement is the part a card cannot carry: the comparison between options, the reason to choose one supplier over another, the specification, the service commitment, the case a card has no field for. That is a different creative brief from the one a category term usually gets.

In categories that keep the click, expect the click to get more contested. As the results area absorbs the simplest queries everywhere else, the remaining clickable demand concentrates. That is not automatically bad — it is the demand that needs an argument, which is where a well-built landing page earns its cost.

🏢 Where Baidu PPC Pro comes in

The China side of this has four prerequisites, and none of them changes when the results page changes.

An account without a Chinese entity. Baidu requires a Chinese business licence to open an advertising account. Most overseas brands do not hold one, and obtaining one takes months. Baidu PPC Pro opens and manages the account without it, which decides whether a brand can be in a Chinese auction at all.

A keyword map built from local terms. Chinese consumers search the local term for the category, the local term for the feature and the local phrase for the problem they want solved. A translated keyword list reaches people and then fails to match them. The comparison queries that carry the highest intent have to be built from observed search behaviour.

Creative that clears review the first time. Advertising in China is filed and reviewed per industry, with claims checked against permitted phrasing. A creative that performs in another market can be rejected on submission here, usually with no advance warning.

Landing pages that answer the search. Chinese-language pages that resolve the comparison the user just typed. Traffic to a page that reads as translated is traffic already paid for and then lost.

📅 What to watch, and notes on the source

Four things would settle the open questions.

1. Whether Baidu extends the pattern to other categories with an authoritative supplier, and which ones.
2. Whether the card ever sits in a paid position, is sponsored, or draws a fee of any kind. At present it is reported as a partnership, not an ad product.
3. Whether Baidu publishes click or query-volume effects for the travel category. Without that, the size of the displacement is unknown.
4. Whether Baidu App's broader move toward an interactive results page continues. The 20 August card and the local-life section are the pattern to read this against.

Notes on the source:

• The partnership was announced on 12 October 2026 and reported on 13 October by Chinese financial and aviation trade media. Umetrip's own statements about its direct-sales platform come from an interview with a Umetrip vice-president in China Economic Times.
• Coverage describes booking as completing "inside the search box". The same coverage states that tapping to book opens Umetrip's booking flow to complete payment. This article uses the narrower version.
• The card is a feature of the results area, not an advertisement, and it did not change Baidu's ad auction. Nothing here should be read as a claim that paid placements were replaced.
• Umetrip's terms — no markup, no agency fee — are the company's own description of its platform, not an independent finding.
• No click-volume, query-volume or revenue effect is quantified, because none has been published.