The big data that Beijing consumer society released recently " kill ripe " findings shows, the net buys platform, online travel and net to make an appointment with the domain such as the car to make big data " kill ripe " heavy disaster area.
Among them, the judgement user such as habit of the OTA platform search frequency according to the user, consumption the urgent pitch to airline ticket, hotel, undertake dynamic float increases price thereby, was labelled " big data is killed ripe " one of main show.
In real life, also have consumer report really, the airline ticket value that sees obviously when the search is very cheap, the fare when happen to pays went up however, or two people are searched at the same time same the same flight number of a time, the value that see is different however.
Want to clear up these problems, want to speak of from the metabolic principle of airline ticket price actually.
The price of a piece of airline ticket wait for composition by tax of surtax of fare, fuel, airport. Among them airline ticket price basically is by airline basis the element such as seasonal, transport power, supply demand relations consolidates adjusting control.
Airline is to be below the examination of the civil aviation bureau, constitute the admission fee of total value airline ticket of every course, the shipping space that gives different discount differentiates on the foundation in total value airline ticket.
The cabin seat or berth that says here, not be the physical cabin seat or berth with the different service when first-class stateroom cabin, tourist class takes a plane so, the sale shipping space that discriminative admission fee uses when pointing to sale airline ticket however (come to those who differ differentiate with different letter commonly child cabin seat or berth) , it is the direct tool that airline has price of airline ticket of freight rate management, control.
For instance country of not long ago is announced " 5 · one " in 1 hour after small long holiday increases, where to go to the quantity of airline ticket search on the net rises sharply, airline ticket price also resembles a rocket kind rapid growth, what 1000 multivariate price difference happen in new break to search is short between zero second.
The backstage of course accrual administrator that this is airline actually revised the airline ticket price of popular destination, also be their job is daily -- in airline ticket by mad the price is given off when grabbing tall child the value is put when cabin seat or berth, nobody is bought low child cabin seat or berth, of pursuit is the balance between rate of every course seat of honour and accrual.
Again for if where,saying airline ticket price is carry Cheng, those who come go showing on the platform such as where net. Consumer arrives from the search next sheet, the process of backside actually very complex.
Differ certainly child after amount of cabin seat or berth, airline can type freight rate data GDS, include to carry Cheng, where to go to waiting for platform and airline ticket agent, can inquire cabin seat or berth and price from GDS system, show the result customer.
GDS is global cent sells a system (the abbreviation of Global Distribution System) , department of special side boat provides the management of inventory, freight rate and order, what domestic aviation uses is medium boat letter, international is Amadeus of business of operation of 3 big GDS more relatively, sabre, travelport.
On platform when consumer input date is mixed set out after reaching destination, platform takes relevant data to GDS attune with respect to need, GDS can reveal place to need all airliners below course and this airliner to the amount of surplus shipping space of every cabin seat or berth gives platform or act as agent subsequently, and platform shows to consumer, it is the different flight number that different boat department provides.
After airliner of some of consumer make choice of, platform need feedbacks to GDS again data information, GDS also needs to check with boat department the price reachs information of cabin seat or berth, boat manages the price with will newest meeting and feedback of information of cabin seat or berth give GDS, repass GDS feedback gives platform, show user mobile phone to go up then.
GDS data accuracy rate also is not 100% , each airline can choose a GDS do Host, doing this GDS of Host to be able to obtain this data of excellent airline is the most accurate, and if other GDS must pass the GDS that does Host to just can obtain data, this exists 2 times to data is transmitted and be handled, be transmitted possibly in data or calculative link occurrence accident.
With where going to the net is exemple, many 680 thousand line is covered on platform, have inquiry amount on 100 million times everyday, inquire need searchs 10 thousand right-and-left cabin seat or berth every time, to inquiry of cabin seat or berth, have inquiry amount one billion two hundred and fifteen million seven hundred and fifty-two thousand one hundred and ninety-two times everyday. These data can be in airline, GDS, agent, be on the move between GDS and GDS even, request data every time how many of metropolis basis node, speed differs somewhat, sometimes an one-way request (the flight number of complex joining together that needs different boat to manage diverse line is combined for instance) be as long as possibly one bell.
That is to say, if you are searching the popular period of time that is a popular line, if arrive from the search,next only time grow too, the price has the possibility that becomes tall really, because consider to leave this paragraph of only time in you, the airline ticket of relatively low cabin seat or berth has been reaved by others (next sheet) .
Of course, appraise at the current rate is the price becomes high not merely, also can produce the situation that the price becomes low, for instance boat manages discovery sells guest to lead under when anticipating, depreciate possibly sales promotion.
In addition, platform every time requests to inquire data to GDS, it is to should pay inquiry fee, and the price is very expensive (where to go to for instance the data fee that the net pays every year is ten million yuan level) , so a lot of platform can choose to lie between period of time to inquire, put the data of inquiry cache supply user.
Such, consumer is when airliner of some of OTA platform search, most the may be the cache in the system data that sees first. And when he will pay link, platform can pass GDS, with airline newest " inventory " undertake matching. If the relevant cabin seat or berth of this airliner already sold out, can give out " this price cabin seat or berth already sold out, search afresh please " clew, cause a search possibly also so from time to tome ticket, the problem of the ticket is not had when paying, namely the stock that technical cache causes replaces different situation.
Accordingly, to going for the platform such as where net, main energy is to pass technical measure to obtain more accurate data from GDS actually, the time that reduces to pay to finally from the search paragraph in the circumstance of appraise at the current rate. Book more popular line to a few users for instance, can dominate the time of cache according to time of far and near or amount of surplus shipping space, increase to the frequency that GDS data requests namely, also mean those who inquire cost to rise of course.
On the whole, in the times with airline ticket more and more transparent nowadays price, support is so called " big data is killed ripe " undertake valence is sold adding, do not have vivosphere more and more, manage to basically relying on discharge, boat commission and news service fee go especially now, for the platform with the higher and higher cost that win a customer.
Of course, need not big data " kill ripe " , can undertake with big data essence of life follows sale however. Already boat manages peaceful stage is studying user picture, the demand preference according to every user, buy a habit, buy a preference, behavior characteristic, undertake to specific product directional put in, attract be bought quite and go out row.
