WO2010068551A1

WO2010068551A1 - Market reference price determination system and method

Info

Publication number: WO2010068551A1
Application number: PCT/US2009/066576
Authority: WO
Inventors: Ronald P. Menich; Jim Rozell
Original assignee: Jda Software Group, Inc.
Priority date: 2008-12-09
Filing date: 2009-12-03
Publication date: 2010-06-17

Abstract

One exemplary aspect of this invention generally relates to the field of resource and revenue management and pricing, and more particularly to price-sensitive forecasting and optimization for the purposes of dynamic pricing at least in the fields of hospitality, car rental, passenger transport (including passenger air, passenger rail), air cargo and in general any perishable asset industry.

Description

MARKET REFERENCE PRICE DETERMINATION SYSTEM AND METHOD

RELATED APPLICATION DATA

[0001] This application claims the benefit of and priority under 35 U. S. C. § 119(e) to U.S.

Patent Application No. 61/121,009, filed December 9, 2009, entitled "Market Reference Price Determination System and Method," which is incorporated herein by reference in its entirety.

FIELD OF THE INVENTION

[0002] One exemplary aspect of this invention generally relates to the field of resource and revenue management and pricing, and more particularly to price-sensitive forecasting and optimization for the purposes of dynamic pricing at least in the fields of hospitality, car rental, passenger transport, air cargo and other related perishable asset industries.

BACKGROUND

[0003] Reference Price is a marketing term defined to be (c.f., http://www.marketingpower.com/mg-dictionary-view3861.php),

[0004] "The price that buyers use to compare the offered price of a product or service. The reference price may be a price in a buyer's memory, or it may be the price of an alternative product. "

SUMMARY

[0005] The reference price might be in the buyer's memory if the buyer has a history of repeat purchases of the same item. However, in one exemplary embodiment of this invention, another case is considered, namely, the one for which the buyer has little or no previous purchase history for the hotel room, car rental or other item the buyer wishes to rent. For example, the customer might wish to book a hotel room in a city he or she has never visited. Another possibility is that the customer has visited that city in the past and rented a room there, but that that happened so long ago that he or she has forgotten the price paid and thus that price is now irrelevant for purposes of forming a price reference for the new rental. In a passenger rail context, a salesperson from the United States planning to visit some new accounts in France might wish to book train travel from Paris to Marseilles, that train trip representing something he or she has not done before. He or she might shop some internet sites to see the ranges of prices offered by various train companies and low cost air carriers servicing that origin and destination pair.

[0006] In such cases, the Internet provides transparency to the customer and allows him or her to form a price reference for a hotel room or rental car based on the ensemble of rates displayed by various providers on websites such as expedia.com^® and hotels.com^®. When the customer forms a price reference in this manner the result is referred to as market reference price, and an exemplary embodiment of this invention describes a technique for determining this quantity for purposes of usage in Revenue Management Systems (RMSs).

[0007] Revenue management is the practice of controlling the prices and availabilities of products so as to maximize revenues or profits for the firm. An RMS typically used data representative of one or more metrics, market and/or pricing information to forecast customer demand and then recommends prices or availability controls as a result of optimizing the pricing and acceptance of that demand into capacity. Older RMSs often did not model the relationship between demand and price in great detail, and existing RMSs do not take a market reference price as described here into account when constructing demand forecasts and control recommendations. But consumer demand can reasonably be expected to respond to the difference between the price offered by a firm and the market reference price a customer constructs by viewing the ensemble of available rates. This lack of market reference price understanding constitutes an area of opportunity for improvement of forecast accuracy and recommendation relevance within revenue management. [0008] Various third-party providers of competitive rate shopping data have emerged in recent years, including QL2, Rubicon (MarketVision), AnyRate, RateGain (RateView), Ubiquick, Infare, Travelclick (Hotelligence) and others. This competitive rate data forms one exemplary input to the market reference price system described herein.

[0009] The concept of a theoretical market price for purposes of price-sensitive demand forecasting is known in the economic literature (c.f., Balvers, Ronald J. and Thomas F. Cosimano, "Actively learning about demand and the dynamics of price adjustment", The Economic Journal, Vol. 100, No. 402, (Sep. 1990), pp. 882-898, specifically the mentioned on page 884 in equation (I).) [0010] JDA, the assignee of the subject application, has an existing Reference Price process within its Demand Decomposition suite of modules. That Reference Price process constructs a weighted average of own past prices to use as a reference. This type of reference price is used in situations for which competitive rate data is not available from which to construct the market reference price described herein. Also, the usage of a weighted average of past own prices is useful in situations for which it is thought the customer has anchored his or her price expectations towards past own prices rather than current market prices; this may be the case in situations of repeat buying by a customer from a single provider over time.

[0011] An exemplary aspect of one aspect of the present invention is better-suited to situations in which customers exhibit little repeat buying behavior or do so with long gaps of time between purchases. [0012] Revenue management practitioners often use this competitive rate data for decision support, comparing the effective prices charged by a firm versus its competitors. Some car rental firms are using this competitive rate data to automatically respond to price changes by competitors. [0013] Passenger airlines have long used competitive fare information provided by ATPCO and SITA. There are automated fare response tools on the marketplace such as Sabre's AirPrice system and JDA's Fare Product Manager (FPM).

[0014] One exemplary aspect of this invention generally relates to compressing competitive rate shopping data into a single market reference price by means of a configurable summarization function such as weighted average, median, kth-percentile and so forth. One exemplary embodiment of the invention could be implemented in software in conjunction with a special purpose or general purpose computer.

[0015] In accordance with this exemplary embodiment, the software implementing this invention executes prior to the execution of and provides input to a price-sensitive forecasting and optimization software. The price-sensitive forecasting software, for example, utilizes the market reference price within a market response model which predicts demand as a function of the percent difference between the offered price and the market reference price.

[0016] Figure 1 is a flowchart illustrating an exemplary high-level usage of the market reference price concept within price-sensitive forecasting. In the Fig. 1, the market reference price is determined by the Market Reference Price Process 100 and fed to the price-sensitive forecasting process shown in steps S110-S190. The price— sensitive forecasting process begins in step SIlO and continues to step S120 with the receipt of bookings with prices. The market reference price is used as input to the process of price normalization S130 and also as input to the what-if forecasting process S170.

[0017] The Price normalization in step S130 is the process of restating historical bookings, each booked at possibly different prices, as though they had booked at the market reference price. [0018] From the price normalization step S130, the baseline demand is determined in step

S140, with a forecast generated in step S150 and the baseline demand outlook generated in step S160.

[0019] The What-if forecasting process S170 projects what demand will be a particular price, based on input price lists S175, the mathematical model which performs this projection understands demand to be a function of the percentage difference between price and market reference price. From step S170 a demand outlook by price is determined in step S180 with control continuing to step S190 where the control sequence ends. [0020] Price-sensitive forecasting is tied to optimization, and Figure 2 illustrates the usage of market reference price within price optimization. In Fig. 2, the market reference price is determined in process 100 and fed to the optimization processes 200. The optimization process includes a price menu, price bounds and parameters, current booked information, demand outlook, upgrading rules and projected capacity. These are fed into the network price optimization routine 202 from which the recommended rates 204 are determined. These recommended rates can then be uploaded to, for example, a reservation system 206.

[0021] The manner in which the market reference price is used in optimization is similar to its usage in what-if forecasting as shown in Figure 1; indeed, a price-sensitive forecasting model is represented in the constraints of a mathematical program which constitutes the optimization. That is, the optimization understands how demand changes as a function of the percentage difference between price and market reference price and the optimization seeks to set the price so as to maximize profit to the firm.

[0022] JDA's price-sensitive demand forecasting system accepts as input a market reference price. As the market reference price changes, so too do the associated demand forecasts. By contrast, forecasts in typical traditional revenue management systems do not change immediately with changes in competitive pricing; rather, typical or traditional revenue management systems respond only very sluggishly as the flow of own bookings increase or decrease over time in response to changes in market pricing. By analogy the response of a traditional RMS will be referred to as "driving by looking at the rearview mirror" whereas having a price-sensitive forecaster dependent on a market reference price is akin to "driving by looking out the front window:" the traditional RMS responds only long after competitive price changes occur, whereas the price-sensitive forecaster dependent on a market reference price responds immediately. Thus the market reference price is central to the ability of a demand forecaster to respond quickly to changes in market pricing. [0023] The incorporation of the market reference price concept within a price optimization scheme allows the optimizer to recommend prices that are within the range of believability, given the rates at which other providers of similar product are offering in the marketplace. The market reference price computation enables believable dynamic pricing recommendations for perishable inventory services industries.

[0024] It is anticipated that this exemplary type of market reference price will be most applicable to demand forecast units or firms associated with customer segments for which either a.) customer segments for which there is not much repeat buying behavior or b.) there is repeat buying behavior but for which the customer expects that high volatility of prices is natural behavior; or c.) there is repeat, but very infrequent, buying behavior such that the customer often does not clearly remember the price he or she previously paid. In such cases and potentially others not enumerated above, a customer does not naturally have a reference to which to compare a price other than by shopping current prices on the Internet, and it is in these cases for which the market reference price described in this form may be the most appropriate. This is often the case with hospitality and car rental, for example. In other situations for which there is a significant amount of repeat buying behavior and limited market price volatility, other functions may be used to generate a reference price.

[0025] The preceding is a simplified summary of the invention to provide an understanding of some aspects thereof. This summary is neither an exhaustive nor extensive overview of the invention and its various embodiments. It is intended neither to identify key or critical elements of the invention nor to delineate the scope of the invention, but to present selected concepts of the invention in a simplified form as an introduction to the more detailed description presented below. As will be appreciated, other embodiments of the invention are possible utilizing, alone or in combination, one or more of the features as set forth above or described in detail below.

BRIEF DESCRIPTION OF THE FIGURES

[0026] The exemplary embodiments of the invention will be described in detail, with reference to the following figures, wherein:

[0027] Fig. 1 illustrates an exemplary market reference price and demand outlook process according to this invention;

[0028] Fig. 2 illustrates an exemplary market reference price and reservation system according to this invention;

[0029] Fig. 3 illustrates an exemplary interface to manage network price optimization and in particular the competitor set weighting function in NPO for computation of market reference price according to this invention;

[0030] Figure 4 illustrates an exemplary interface for competitor search and selection in

NPO according to this invention;

[0031] Figure 5 illustrates an interface for season management according to this invention;

[0032] Figure 6 illustrates an exemplary interface reflecting market reference price in chart form by week and day according to this invention;

[0033] Figure 7 illustrates an exemplary price determination system according to this invention;

[0034] Figure 8 illustrates an exemplary flowchart outlining a market reference price determination according to this invention; [0035] Figure 9 illustrates an exemplary flowchart outlining a more detailed process for determining market reference price according to this invention; and

[0036] Figure 10 illustrates an exemplary flowchart outlining a more detailed process for determining market reference price and price sensitive forecasting according to this invention.

DETAILED DESCRIPTION

[0037] The exemplary embodiments of this invention will be described in relation to a market reference price system, as well as, techniques and methods to determine a market reference price from competitive rate data. However, it should be appreciated, that in general, the systems and methods of this invention will work equally well for other types of data and for other types of forecasting in a plurality of environments.

[0038] The exemplary systems and methods of this invention will also be described in relation to a market reference price determination architecture, and associated communications, hardware and software component(s). However, to avoid unnecessarily obscuring the present invention, the following description omits well-known structures and devices that may be shown in block diagram form or are otherwise summarized or known.

[0039] For purposes of explanation, numerous details are set forth in order to provide a thorough understanding of the present invention. It should be appreciated however that the present invention may be practiced in a variety of ways beyond the specific details set forth herein. [0040] Furthermore, while the exemplary embodiments illustrated herein show the various components of the system collocated, it is to be appreciated that the various components of the system can be located at distant portions of a distributed network, such as a communications network and/or the Internet, or within a dedicated secure, unsecured, and/or encrypted system. [0041] Thus, it should be appreciated that the components of the system can be combined into one or more devices, collocated on a particular node(s) of a distributed network, such as a distributed network. As will be appreciated from the following description, and for reasons of computations efficiency, the components of the systems can be arranged at any location within a distributed network without affecting the operation thereof. For example, the various components can be located with a forecast, optimization or sales suite of products, or some combination thereof. Similarly, one or more functional portions of this system could be distributed between a computing device and a server.

[0042] Furthermore, it should be appreciated that the various links, including the communications channels connecting the elements can be wired or wireless links or any combination thereof, or any other known or later developed element(s) capable of supplying and/or communicating data to and from the connected elements. The term module as used herein can refer to any known or later developed hardware, software, firmware, or combination thereof, that is capable of performing the functionality associated with that element. The terms determine, calculate, and compute, and variations thereof, as used herein are used interchangeably and include any type of methodology, process, technique, mathematical operation or protocol. [0043] The term "automatic" and variations thereof, as used herein, refers to any process or operation done without material human input when the process or operation is performed. However, a process or operation can be automatic even if performance of the process or operation uses human input, whether material or immaterial, received before performance of the process or operation. Human input is deemed to be material if such input influences how the process or operation will be performed. Human input that consents to the performance of the process or operation is not deemed to be "material."

[0044] The term "computer-readable medium" and computer-readable storage medium as used herein refers to any tangible storage and/or transmission medium that participates in providing instructions to a processor for execution. Such a medium may take many forms, including but not limited to, non-volatile media, volatile media and transmission media. Non-volatile media includes, for example, NVRAM, or magnetic and/or optical disks. Volatile media includes dynamic memory, such as main memory. Common forms of computer-readable media include, for example, a floppy disk, a flexible disk, hard disk, magnetic tape, or any other magnetic medium, magneto-optical medium, a CD-ROM, any other optical medium, punch cards, paper tape, any other physical medium with patterns of holes or deformation, a RAM, a PROM, and EPROM, a FLASH-EPROM, a solid state medium like a memory card, any other memory chip or cartridge, a carrier wave as described hereinafter, or any other medium from which a computer can read. A digital file attachment to e- mail or other self-contained information archive or set of archives is considered a distribution medium equivalent to a tangible storage medium. When the computer-readable media is configured as a database, it is to be understood that the database may be any type of database, such as relational, hierarchical, object-oriented, and/or the like. Accordingly, the invention is considered to include a tangible storage medium or distribution medium and prior art-recognized equivalents and successor media, in which the software implementations of the present invention are stored. [0045] The terms "determine," "calculate" and "compute," and variations thereof, as used herein, are used interchangeably and include any type of methodology, process, mathematical operation or technique.

[0046] Also, while the invention is described in terms of exemplary embodiments, it should be appreciated that individual aspects of the invention can be separately claimed. [0047] The setting in which one exemplary embodiment of the invention operates in the hospitality industry is as follows; comments will be given later regarding its application in other perishable asset services industries. Revenue management software provides rate recommendations for a set of properties S and each future stay night d within a set D. (For example, S might constitute the set of XYZ Hotels and Suites and D might constitute 0 to 120 stay nights in the future from the current calendar date.) For each property s e S , there is a set C(s) of competitive properties selling similar quality product in the same general area as s; the set C(s) is defined by the hotel operator.

[0048] Using a third-party provider of competitive rate shopping data such as QL2 or

Rubicon MarketVision, own- and competitive rates are collected for each own property s in S and each competitive property c ε C(i) and each date d e D . Let p (c,s, d^~) be the selling rate for date d at competitive property c ε C (s) returned from the competitive shopping robot. The market reference price p (c, -, d\ is constructed as follows:

[0049] Weighted Average Example

[0050] Let λ [s, c, d) > 0 be a nonnegative weight; the exemplary approach provides convenient mechanisms for the specification of these weights. A market reference price constructed using the weighted average method is:

∑ λ(s,c,d)-p(s,c,d)

_P{s,;d) = —

2_^ λ{s,c,d) ceC{ή

[0051] Optionally, no renormalization is performed, in which case the formula becomes: p(s,-,d) = ∑ λ(s,c,d)-p(s,c,d) ceC(s)

[0052] Facilities for manipulating the Weighted Average Method are can be operative in the

GUI for the Network Price Optimizer (NPO).

[0053] Figure 3 illustrates the ability of the NPO GUI 300 to allow users to enter weights

310 for the computation of the market reference price as a weighted sum of competitor rates. For hospitality, competitor weights are associated with a hotel property or location and are considered to be resource-centric. For passenger transport (e.g., passenger air, passenger rail), such competitor weights are associated with a passenger origin and destination ("O&D") and are considered to be package-centric. Thus, the concepts described in an exemplary embodiment of this invention may be applied to the differing aggregation levels (e.g., resource, package) at which the competitor weights naturally live in the particular industries to which the invention may be applied. [0054] In Figure 3, different competitor weights can be specified for different seasons 320, where a season is defined to be a date range and selection of days of week. The concept of multiple seasons for competitor weights applies whether those weights be specified at the resource, package or other aggregation level. Figure 4 shows an interface 400 that allows a user to search 410 for a competitor 420 and add that to its competitor-list set 430.

[0055] Figure 5 illustrates en exemplary season specification dialog 500 of NPO. In 510, two exemplary season definitions are present, showing the start date, end date, season name and days of the week.

[0056] Figure 6 illustrates an exemplary embodiment of the determined market reference price over the course of a week— the curve with legend "marketRef" 610 — appearing within the chart 620 in an exemplary NPO.

[0057] Median Example

[0058] Under the median method, the market reference price is constructed as follows: p(s, -,d) - median I (/7(5, c, t/) : c <≡

[0059] kth-Percentile Example

[0060] Under the kth-percentile method, the market reference price is constructed as: p[s,-,d) - percentile! {/?(s,c,t/) : c <≡ C[s)\ ,k) where the syntax of the percentile function described here is as in Microsoft^® Excel^®. Note that for k = 0.5 , the kth-percentile method reduces to the median method mentioned above.

[0061] Position in Range Example

[0062] Under the position in range method, the market reference price is constructed as: p( ^Vs, ;d) ' = -k) -

/ c meCi(ns) p( ^Vc, s, d) ' with k = 0.5 , the position in range method reduces to the midpoint of the range.

[0063] One exemplary embodiment of the invention provides a means of selecting between the Weighted Average, Median, Percentile and Position in Range and automatic methods for computing the market reference price.

[0064] Market reference prices so computed are stored to database tables within, for example, an enterprise architecture. Thenceforth these reference prices can be utilized as input to price-sensitive forecasting modules, and, for example, JDA's Network Price Optimizer dynamic pricing module.

[0065] Special case processing is required when no rate is returned from the competitive rate shop, either because of error or because of lack of availability. In the case of lack of availability, the p [c,s, d) is set to a high value p_hι {s,d) ; this is done so that the downstream price-sensitive forecasting and optimization components can interpret lack of product availability in the marketplace as an opportunity for property s to raise its rates. The p_hι (s, J) can be set, for example, to the maximal observed rate for competitor c for property s over some historical timeframe.

[0066] In the special case for which the rate shop does not return a rate due to a processing error, the p (c, s, d) is estimated to be the same as the most recently observed value of that competitive rate. Competitive rate shops are typically done every day. So if the p ic, s, d\ is missing in the rate shop performed on "May 13^th", then the real value of p (c,s, d^~) observed on

"May 12^th" may be substituted.

[0067] In other perishable asset industries the concepts are similar to what is described for hospitality above. For example, in passenger rail, the set is the set of own (train number, departure date, passenger origin station, passenger destination station, cabin/class, further customer or market segment or product type descriptors) and the set is a set of comparable competitive offers for rail services, coach services and even low-cost carrier airplane flights. In car rental, the set is the set of rental locations and the set of competitive providers of car rentals at the same airport as own location.

[0068] This section describes one exemplary design of the competitive pricing module, which is intended to generate a market reference rate given competitive shopping data from a third- party vendor such QL2, Rubicon, MarketVision, etc. The module is generic, and applicable to different industries. However, there are two aspects of this process for which a custom approach is taken for handling the different input files and steps to be performed on the input data. 1.) Third party vendor for competitive shopping data: A different input file format will most likely be received from each vendor.

2.) Vertical of the client: Hospitality, rail, rental car, low cost carriers, etc. will have different columns in the input files and different shopping patterns.

[0069] There are separate libraries for each of these cases such as QL2 - Hospitality, QL2 -

Rental Car, QL2 - Low Cost Carrier, Rubicon-Hospitality, etc. For convenience, the QL2-Hospitality pairing will be used to describe the design in this document; it is understood however that the use of QL2-Hospitality as a means of easy explication does not limit the invention to that pairing. [0070] It is standard within the industry to use QL2 for shopping a property's own price as well as its competitor's prices. In the NPO database, the list of competitors for all properties (including the own properties) is stored in the N PO_COM PETITOR table. The competitive pricing module first runs a high level filter to get only the list of competitors in the NPO database from the raw QL2 shopping data. It is assumed for the rest of the document that this filter has been applied before proceeding with the following steps.

[0071] The mapping between a property and its associated competitors are specified in the

NPORSRC_COMPT_MAP table. Users assign weights for each competitor in the Ul based on calendars, which is stored in NPORSRC_COMPT_WT.

[0072] The next section describes the steps to be performed by the competitive pricing module once the QL2 shopping data arrive and the filter has been applied. There are two types of files that are received. The first type of file (e.g., "out.csv" for QL2) contains all the rates that the shop successfully returned and the second type of file (e.g., "error.csv" for QL2) contains all the error messages from the shop such as the rate being unavailable.

[0073] The competitive pricing module includes one program/binary with two processes.

The first process operates in a custom fashion pulling the associated third-party vendor and vertical library and perform input data processing until it finally populates the NPO_RSRCDT_PRICESHOP and

NPO_PKGDT_PRICESHOP tables. The second process uses the NPO_PKGDT_PRICESHOP table to compute a market reference rate and populate the DFUEFFPRICE and DFUREFPRICE tables in the

SCPO schema. A PROCESS_OPTION is defined to control whether only the first process runs, only the second one runs or both processes run as first followed by the second.

[0074] The prerequisites of this module are completeness and referential integrity in the following exemplary tables:

NPO_PKG

NPO_PKGDT

NPO_RSRC

NPO_RSRCDT

NPO_DFU

NPO_DFUDT

NPO_DFUPKGDT_MAP

NPO_RSRCPKGDT_MAP

N PO_COM PETITOR

N PO_RSRC_COM PT_M AP

N PO_RSRC_COM PT_WT

NPO_RSRC_CAL_PARAM

NPO_BUCKET

NPO_RSRC_BKT_PARAM

NPO DFUDT PRICEPT BKT DFUPRICEPARAM (SCPO)

[0075] Some of these table names are modified when the invention is operating in the mode for which competitor weights are at the package (e.g., passenger air/rail industries) rather than at the resource (e.g., hospitality) level; this same comment applies to subsequent descriptions of tables over the next several pages. Thus, the module should be scheduled to run in the batch after these tables have been successfully populated. [0076] Steps to Be Performed by the Module - Process I

[0077] This exemplary process first reads raw competitive shopping data filtered by the set of competitors that exist in the NPORSRC_COMPT_MAP table by each property (RSRCJD).

[0078] The fields that would be needed from the first type of file ("out.csv") for the

QL2/Hospitality library are SEARCHJDATE, SITE (shopping source), NAME (NPO_COMPETITOR.CompetitorName), Cl (check-in date), CO (check-out date), RATEl, RATE2, RATE3, RATE4, RATE5, TOTAL_RATE and NIGHTS (If NIGHTS does not exist in the data file, it can be constructed in memory when the dataset is read such that NIGHTS=CO-CI). Likewise, the fields that would be needed from the second type of file ("error.csv") for the QL2/Hospitality library are SITE (shopping source), NAME (competitor/property name), Cl, CO and MESSAGE (SEARCHJDATE for the error file is the same as the SEARCH JDATE in the out file).

[0079] Currently, hospitality only shops for base rates, i.e., the "Selling Rates" rate category, for each stay night in the forecasting horizon, which is 120 days. The shopping is for three different lengths of stay (LOS) patterns, namely LOS = 1, 3 and 5. However, LOS>1 is only intended to fill in the holes where a one night stay rate is not available.

[0080] Note that for car rental, there could be two exemplary shopping patterns, 1 day length of rent (LOR) and 7 day LOR. For an extended stay property, a tiered rate structure could be used and the different shopping patterns would be defined accordingly. However, it is safe to assume that the shopping will be at the NPO Package level, which is defined by an arrival date and length of stay for hospitality and (train number, departure date/time, passenger origin station, passenger destination station, class/cabin, other customer or market segment or product type descriptors) in passenger air/rail environments. Room types and type of car could also be included in the shopping patterns and the keys in the NPO Package would be defined accordingly. [0081] The exemplary steps to be performed are as follows:

[0082] The rows from the first type of input file ("out.csv") should be pulled into a

NPO_PKGDT_PRICESHOP memory object with the correct mapping from the fields in the file to the columns in the NPO_PKGDT_PRICESHOP table. The exemplary embodiment only pulls PKGJD, STARTDATE that exist in the prerequisite table NPO_PKGDT. [0083] QL2-Hospitality Example:

[0084] The Status code for the rows that come from "out.csv" should be O, while the rows that come from "error.csv" should be a particular error code such as the rate being unavailable. From the "error.csv" file, only the rows for which the SITE, NAME and Cl do not exist in "out.csv" should be included. For the rows from "error.csv", CUR_PRICE is initially NULL. Finally, if COMPTJD : RSRCJD, then IS_0WN=Υ, otherwise, IS_0WN='N' in the NPO_PKGDT_PRICESHOP object.

2. Construct the dataset for NPO RSRCDT PRICESHOP:

[0085] Tables needed for the mapping between resources and packages in this step are as follows:

NPO RSRCDT

NPO PKGDT

RSRCJD

PKGJD STARTDATS STARTDATE

a. Using LOS=I for creating resource prices i. Transfer all rows for PKGJD, STARTDATE, COMPTJD with NIGHTS=I to associated

RSRCJD, STARTDATE, COMPTJD (Use RSRC_STARTDATE=PKG_STARTDATE=STARTDATE). ii. If there are multiple sources (SOURCE) for the same PKGJD, STARTDATE and

COMPTJD, use the row with the minimum CURJ¹RICE (This is configurable based on a field in NPOJ³KG table allowing selection of min, max, average, etc.). iii. Set AVAI LJ N D ICATO R=' N' if CUR_PRICE=NULL, and AVAI LJ N DICAT0R=Υ otherwise. iv. If CURJ³RICE is NULL for any of the rows, proceed with Step b. v. Otherwise, proceed with Step e. b. Using L0S=3 for creating resource prices (the table for the hierarchical procession in these steps is yet to be designed) i. For any RSRCJD, STARTDATE, COMPTJD with CUR_PRICE=NULL, update the

CURJ³RICE with RATEl of PKGJD, STARTDATE, COMPTJD with NIGHTS=3 and STATUS=O (Use RSRC_STARTDATE= PKG_STARTDATE=STARTDATE) (e.g., Use RATE2, RATE3, etc. for different passes.). ii. If there are multiple sources (SOURCE) for the same PKGJD, STARTDATE and

COMPTJD, use the row with the minimum RATEl (See the footnote for a. ii.) iii. Set AVAI LJ N D ICATO R=Υ for the rows where update is successful, iv. If CUR_PRICE is NULL for any of the rows, proceed with Step c. v. Otherwise, proceed with Step e. c. Using the third level of hierarchy for creating resource prices i. For any RSRCJD, STARTDATE, COMPTJD still with CUR_PRICE=NULL, update the

CURJ³RICE with RATEl of PKGJD, STARTDATE, COMPTJD with NIGHTS=5 and STATUS=O (Use RSRC_STARTDATE= PKG_STARTDATE=STARTDATE). ii. If there are multiple sources (SOURCE) for the same PKGJD, STARTDATE and

COMPTJD, use the row with the minimum RATEl (See the footnote for a. ii.). iii. Set AVAI LJ N D ICATO R=Υ for the rows where update is successful. iv. If CURJ³RICE is NULL for any of the rows, proceed with Step d. v. Otherwise, proceed with Step e. d. For all the remaining rows still with CURJ³RICE=NU LL, some hole-filling logic based on the NPO_RSRCDT_PRICESHOP object (the design of which is yet to be finalized) should be applied. The high-level design of the logic is as follows (What is performed in this step changes with respect to different Status code possibilities. The activities described for a Status code denote the unavailability of the rate): i. If IS_0WN=Υ:

1. Find the max CURJ³RICE available over the horizon for this resource (Let MAXJZURJ³RICE denote this value).

2. Find the max rate boundary (RATEBOUNDARYMAX) from N PO-RSRC-CALJ³ARAM for this RSRCJD such that STARTDATE is between N PO-RSRC-CALJ³ARAM. Startdate and NPO_RSRC_CAL_PARAM.Startdate+Dur.

3. Take the minimum of MAX-CU RJ³RICE and RATEBOUNDARYMAX and update CURJ³RICE. ii. If IS_0WN='N':

1. Find the max CURJ³RICE available over the horizon and update the

CURJ³RICE with this proxy rate. e. Load the final dataset into NPO-RSRCDTJ³RICESHOP (after purging rows that exist in the table with the same keys, if any).

3. Complete the dataset to be loaded into the N POJ³KGDTJ³RICESHOP table. a. For any rows in the NPOJ³KGDTJ³RICESHOP object where CURJ³RICE is NULL, construct the price from summing up the resource prices:

CUR _ PRICE _PKG := ∑ CUR _PRICE_RSRC

RSRC≡SetofllSRCbyPKG For each PKGJD, STARTDATE (=PKG_STARTDATE), COMPTJD in the NPO_PKGDT_PRICESHOP object:

i. Find RSRCJD, RSRC_STARTDATE in NPO_RSRCPKGDT JVlAP ii. Sum up the CUR_PRICE of these RSRCJD, RSRC_STARTDATE and

COMPTJD in NPO_RSRCDT_PRICESHOP b. Identify any dated packages that do not currently exist in the NPO_PKGDT_PRICESHOP object by making a distinct between NPO_PKGDT and N POJ³KGDTJ³RICESHOP such that NPOJ^KGDT.STARTDATE >= N POJ³KGDTJ³RICESHOP. SHOPJDATE. For each PKGJD, STARTDATE (=PKG_STARTDATE) in these additional rows in the N POJ³KGDTJ³RICESHOP object: i. Find RSRCJD, RSRC_STARTDATE in NPO_RSRCPKGDT JVIAP ii. Sum up the CURJ³RICE of these RSRCJD, RSRC_STARTDATE and COMPTJD in NPO-RSRCDTJ³RICESHOP (note that COMPTJD is not defined in N POJ³KGDTJ³RICESHOP object for these rows and comes from NPO-RSRCDTJ³RICESHOP) c. Load the final dataset into N POJ³KGDTJ³RICESHOP (after purging rows that exist in the table with the same keys, if any).

[0086] Steps to Be Performed by the Module - Process Il

[0087] In this exemplary process, the DFUs in the NPOJDFUDT table are found such that

N POJDFU DT.STARTDATE>= max(NPOJ³KGDT J³RICESHOP-SHOPJDATE). Note that if the

PROCESSJDPTION is to run both processes one after the other, N POJ³KGDTJ³RICESHOP memory object can be used in the second process. Otherwise, one should read from the

N POJ³KGDTJ³RICESHOP and construct the required dataset for this process.

[0088] In addition to N POJ³KGDTJ³RICESHOP, the tables that are needed in this step are:

[0089] For each DFUJD, STARTDATE:

1. Find PKGJD, STARTDATE that they use.

2. Find the RSRCJD, BKTJD that they use to get the associated rate menu offset types (menu_offsetjype, pMOT) and offset values (offset_value, pMO), competitive menu offset types (menu_offsetjype, pMOTC) and competitive offset values (offset_value, pMOC) in the NPO_RSRCJ3KT_PARAM.

3. Compute the current price (CurPrice) for each DFUJD, STARTDATE and insert into DFUEFFPRICE (use dmdunit, dmdgroup, loc in NPOJDFUDT) using the rate menu on package based rates (NPO_PKGDT_PRICESHOP rows with ISJDWN=T).

CurPrice_DJu :=

P^M0 RSRc, BKT * ^CUR - ^PRICE PKG , ^{if '} P^M0T RSRc ,Bκτ = ' multiplicative '

(N * pM0_{RSRr BKT} ) + CUR _ PRICE _PKG , if _PM0T_{RSRC «κτ} = ' additive '

where N is the number of resources that the PKG uses. Compute the market reference rate for DFUREFPRICE.

From the NPO_PKG table, find the MktRefCompMethod and MktRefCompVal. The current method that is used for Hospitality is a weighted average computation. For each PKGJD, STARTDATE in NPO_PKGDT_PRICESHOP, take the weighted average of all competitor prices (CURJ³RICE where IS_0WN='N').

[0090] An alternative method is to take the average of the lowest and highest competitor prices.

Let REF_PRICE denote this computed rate by PKGJD and STARTDATE.

Compute the refprice for each DFUJD, STARTDATE and insert into DFUREFPRICE (use dmdunit, dmdgroup, loc in NPOJDFUDT). RefPrice_Dfu :=

pMOC^c^ * REF _ PRICE_PKG , if p MOTC ^ _BKT = 'multiplicative'

(N * pM0C_{RSRC BKT} ) + REF _ PRICE _PKG , if pM0TC_{RSRC BKT} = ' additive '

where N is the number of resources that the PKG uses. [0091] Automated Market Price Determination [0092] Market reference prices can be determined using user-specified static weightings, as described elsewhere in this document. It is also possible to automatically estimate competitive weightings. This automated weight estimation (AWE) can become vitally important in a situation for which hundreds of hotels or trains or car rental locations must have market prices established quickly, something that can be difficult to achieve relying upon humans to accurately specify these weightings.

[0093] For the purposes of this discussion, the hospitality language will be used; however, the method is general and can be extended to passenger transport, car rental and other industries as well. Automated weight estimation is achieved using linear regression with constraints on the values of the parameters. A preprocessor computes average pairwise offset values of each competitor from the own hotel property prior to the AWE procedure; this offset estimation ensures that the own property's rates lie within the cloud of offsetted competitor rates. The own rates for a hotel are the dependent or Y-variable in the regression, and the competitive rates from the different competitors form the X-matrix or independent variables in the regression. Weightings are constrained to be non-negative. Different behaviors of the market price can be achieved by either allowing the sum of weights across all competitors to float, or to require that the sum of weights across competitors to equal unity (1.0). Each row in the regression constitutes an observation of prices on a particular date in the future; the ensemble of rows in the matrix typically contains a horizon of either 120 or 365 dates into the future.

[0094] Solution of the constrained linear regression can be achieved using a quadratic programming solver such as CPLEX. The objective is to minimize the sum of squared errors between the hotel's own rates and the prediction arising from the weighted competitor rates. The decision variables in the optimization are the estimated competitive weights. Linear constraints require the weights decision variables to be nonnegative and optionally sum to unity across the set of competitors.

[0095] The output of the automated market price determination module is the set of estimated weights for the competitors. That is, the output of the automated procedure is exactly the same set of weights that users can enter directly using the static procedure described elsewhere in this document.

[0096] Fig. 7 outlines an exemplary market reference price determination system 700 according to this invention. The market reference price determination system 700 comprises a CRD selection module 710, a weighted average MRP determination module 720, a median MRP determination module 730, a Kth percentile determination module 740, a position in range MRP determination module 750, an exception determination module 760, processor 702, memory 704, display module 706, I/O interface 708, forecasting engine module 770, preferences module 780 in addition to well known componentry. The market reference price determination system 700 can be connected to one or more displays 701 and input devices 707, and receives competitive rate data from one or more feeds via network 10 and links 5. As discussed, one or more of these competitive rate data feeds are provided to the CRD selection module 710 for selection for use in the market reference price determination system 700. Once the competitive rate data is selected by the CRD selection module 710, usually at the direction of a user, a market reference price determination method is selected. As discussed, these methods include the weighted average method, the median method, the Kth percentile method, and the position in range method.

[0097] For the weighted average method, the weighted average MRP determination module 720 receives information regarding the season and attribute. Competitor weights are then selected as well as boundary information, priority information, as well as the optional configuration of an autopilot. These various weights and information can be input, for example, by a user associated with input device 707. The weighted average MRP determination module 720 then determines the weighted average market reference price, which is output.

[0098] For the median method, the median MRP determination module 730 determines the median market reference price as discussed which is then output.

[0099] For the Kth percentile method, the Kth percentile determination module 740 determines the Kth percentile MRP as discussed above, which is then output.

[00100] For the position in range method, the position in range MRP determination module

750 determines the position and range market reference price as discussed above, which is then output.

[00101] In cooperation with one or more display module 706, processor 702, memory 704,

I/O interface 708 and input devices 707, a user can then manipulate and select various views of the

MRP. As illustrated, these MRP views can be graphically displayed in a user interface, output and reports, output data, or the like. Furthermore, and in accordance with the preferences module 780, this information can be forwarded to one or more price sensitive forecasting engines for further prediction and analysis.

[00102] Fig. 8 is a high level flowchart outlining an exemplary embodiment of market reference price determination according to this invention. In particular, price shop data, own price data, and a list of competitors are input. Next, in step S800, the automated weight estimation process is performed. Then, in step S810, the market reference price is determined, which can then optionally be output to, for example, a price optimizer. [00103] Fig. 9 outlines in greater detail the method of determining market reference prices according to an exemplary embodiment of this invention. In particular, specifications for the item(s) to be shopped, and frequencies, as well as third party competitive rate data are acquired and stored in step S900. This raw competitive rate data (S910) can optionally be cleaned and hole filled as needed in step S920. Then, in step S930, the prepared competitive rate data is ready to be forwarded to the market reference price determination step S950.

[00104] As initial input to market reference price determinations step, a user selects the automated reference price as well as or optionally the manual market reference price determination method that will be used e.g., weighted average, median, Kth percentile, etc. These are input to the control directive step S940 which is also fed into the step for determining the market reference prices S950.

[00105] In step S960 the market reference prices are saved to appropriate database tables from which they can be extracted, in step 970, for downstream consumption by other processes such as price sensitive forecasting step S980, network price optimization step S982, graphical user interface outputs, reports, and the like, in step S984, or used for input to some other process, report, or the like in step S986.

[00106] Fig. 10 outlines in greater detail a method for determining market reference price according to another exemplary embodiment of this invention. In particular, control begins in step SlOOO and continues to step SlOlO. In step SlOlO, the competitive rate data is selected. Next, in step S1020, a determination is made whether the selected data is available. If the selected data is available, control continues to step S1030. Otherwise, control jumps to step S1022. In step S1022, a determination is made whether an error indication or no rate should be returned to the user. In the event of an error, control jumps to step S1024 where, with for example the assistance of the exception determination module, the market reference price can be set as an estimate with control continuing to step S1080. Otherwise, in the event of no rate, the market reference price can be set to a high value in step S1026 with control continuing to step S1080.

[00107] In step S1030, the market reference price determination method is selected and the selected CRD forwarded to the appropriate module for the determination. More specifically, for the weighted average method, control continues to step S1040. For the median market reference price determination, control continues to step S1050. For the Kth percentile market reference price determination method, control continues to step S1060. For the position in range market reference price determination method, control continues to step S1070. [00108] For the weighted average method, in step S1040, seasons and attributes can be selected. Next, in step S1042, competitor weights can be assigned. Then, in step S1044, boundaries can also be selected. Control then continues to step S1046.

[00109] In step S1046, priority can be assigned. Then, in step S1048, autopilot can optionally be configured. Then, in step S1049, the weighted average market reference price is determined. Control then continues to step S1080.

[00110] In step S1050, the median market reference price is determined with control continuing to step S1080.

[00111] In step S1060, the Kth percentile market reference price is determined with control continuing to step S1080.

[00112] In step S1070, the position in range market reference price is determined with control continuing to step S1080.

[00113] In step S1080, the market reference price is output. Next, in step S1085, various views of the market reference price can be selected. As discussed, these views can be any type of bar chart, graph, data representation, in general any graphical or textual or data-based display or output of the determined market reference price. Then, in step S1090, this market reference price information can be stored and optionally forwarded to, for example, one or more price sensitive forecasting engines. Control then continues to step S1095 where the control sequence ends. [00114] The exemplary systems and methods of this invention have been described in relation to databases, data analysis, data processing, market reference price determinations and data structures. However, to avoid unnecessarily obscuring the present invention, the description omits a number of known structures and devices. This omission is not to be construed as a limitation of the scope of the claimed invention. Specific details are set forth to provide an understanding of the present invention. It should however be appreciated that the present invention may be practiced in a variety of ways beyond the specific details set forth herein. [00115] Furthermore, while the exemplary embodiments illustrated herein show various components of the system collocated, certain components of the system can be located remotely, at distant portions of a distributed network 10, such as a LAN, cable network, and/or the Internet, or within a dedicated system. Thus, it should be appreciated that the components of the system can be combined into one or more devices, or collocated on a particular node of a distributed network, such as an analog and/or digital communications network, a packet-switch network, a circuit- switched network or a cable network.

[00116] It will be appreciated from the preceding description, and for reasons of computational efficiency, that the components of the system can be arranged at any location within a distributed network of components without affecting the operation of the system. For example, the various components can be located in an analytical data tool and/or expert data analysis system. [00117] Furthermore, it should be appreciated that the various links (which may or may not be illustrated), connecting the elements can be wired or wireless links, or any combination thereof, or any other known or later developed element(s) that is capable of supplying and/or communicating data to and from the connected elements. These wired or wireless links can also be secure links and may be capable of communicating encrypted information. Transmission media used as links, for example, can be any suitable carrier for electrical signals, including coaxial cables, copper wire and fiber optics, and may take the form of acoustic or light waves, such as those generated during radio-wave and infra-red data communications.

[00118] Also, while the flowchart has been discussed and illustrated in relation to a particular sequence of events, it should be appreciated that changes, additions, and omissions to this sequence can occur without materially affecting the operation of the invention. [00119] In yet another embodiment, the systems and methods of this invention can be implemented in conjunction with a special purpose computer, a programmed microprocessor or microcontroller and peripheral integrated circuit element(s), an ASIC or other integrated circuit, a digital signal processor, a hard-wired electronic or logic circuit such as discrete element circuit, a programmable logic device or gate array such as PLD, PLA, FPGA, PAL, special purpose computer, any comparable means, or the like. In general, any device(s) or means capable of implementing the methodology illustrated herein can be used to implement the various aspects of this invention. Exemplary hardware that can be used for the present invention includes computers, enterprise systems, demand chain management systems, handheld devices, and other hardware known in the art. Some of these devices include processors (e.g., a single or multiple microprocessors), memory, nonvolatile storage, input devices, and output devices. Furthermore, alternative software implementations including, but not limited to, distributed processing or component/object distributed processing, parallel processing, or virtual machine processing can also be constructed to implement the methods described herein.

[00120] In yet another embodiment, the disclosed methods may be readily implemented in conjunction with software using object or object-oriented software development environments that provide portable source code that can be used on a variety of computer or workstation platforms. Alternatively, the disclosed system may be implemented partially or fully in hardware using standard logic circuits or VLSI design. Whether software or hardware is used to implement the systems in accordance with this invention is dependent on the speed and/or efficiency requirements of the system, the particular function, and the particular software or hardware systems or microprocessor or microcomputer systems being utilized.

[00121] In yet another embodiment, the disclosed methods may be partially implemented in software that can be stored on a storage medium, executed on programmed general-purpose computer with the cooperation of a controller and memory, a special purpose computer, a microprocessor, or the like. In these instances, the systems and methods of this invention can be implemented as program embedded on personal computer such as an applet, JAVA^® or CGI script, as a resource residing on a server or computer workstation, as a routine embedded in a dedicated measurement system, system component, or the like. The system can also be implemented by physically incorporating the system and/or method into a software and/or hardware system. [00122] The present invention, in various embodiments, configurations, and aspects, includes components, methods, processes, systems and/or apparatus substantially as depicted and described herein, including various embodiments, subcombinations, and subsets thereof. Those of skill in the art will understand how to make and use the present invention after understanding the present disclosure. The present invention, in various embodiments, configurations, and aspects, includes providing devices and processes in the absence of items not depicted and/or described herein or in various embodiments, configurations, or aspects hereof, including in the absence of such items as may have been used in previous devices or processes, e.g., for improving performance, achieving ease and/or reducing cost of implementation.

[00123] The foregoing discussion of the invention has been presented for purposes of illustration and description. The foregoing is not intended to limit the invention to the form or forms disclosed herein. In the foregoing Detailed Description for example, various features of the invention are grouped together in one or more embodiments, configurations, or aspects for the purpose of streamlining the disclosure. The features of the embodiments, configurations, or aspects of the invention may be combined in alternate embodiments, configurations, or aspects other than those discussed above.

[00124] This method of disclosure is not to be interpreted as reflecting an intention that the claimed invention requires more features than are expressly recited in each claim. Rather, as the following claims reflect, inventive aspects lie in less than all features of a single foregoing disclosed embodiment, configuration, or aspect. Thus, the following claims are hereby incorporated into this Detailed Description, with each claim standing on its own as a separate exemplary embodiment of the invention.

[00125] Moreover, though the description of the invention has included description of one or more embodiments, configurations, or aspects and certain variations and modifications, other variations, combinations, and modifications are within the scope of the invention, e.g., as may be within the skill and knowledge of those in the art, after understanding the present disclosure. It is intended to obtain rights which include alternative embodiments, configurations, or aspects to the extent permitted, including alternate, interchangeable and/or equivalent structures, functions, ranges or steps to those claimed, whether or not such alternate, interchangeable and/or equivalent structures, functions, ranges or steps are disclosed herein, and without intending to publicly dedicate any patentable subject matter.

Claims

Claims:

1. A method for determining a market reference price, the market reference price representing a price used to compare an offered price or service comprising: receiving competitive rate data from one or more data sources; based on one or more of an automated method, a weighted average method, a median method, a Kth percentile method and a position in range method, and with the cooperation of a processor, determining a market reference price; and one or more of saving and outputting the determined market reference price.

2. The method of claim 1, wherein the market reference price is stored and forwarded to one or more price sensitive forecasting engines.

3. The method of claim 1, wherein one or more of the competitive rate data, fare data and price data is used as input for the determining step.

4. The method of claim 1, wherein the market reference price is used for one or more of network prize optimization, input for price sensitive forecasting, displayed on an output device and output in a report.

5. The method of claim 1, wherein the weighted average method further comprises selecting season information and attributes.

6. The method of claim 1, wherein the weighted average method further comprises selecting competitor weight information.

7. The method of claim 1, wherein the weighted average method further comprises selecting one or more boundaries.

8. The method of claim 1, wherein the weighted average method further comprises selecting priority information.

9. The method of claim 1, wherein the weighted average method further comprises configuring an autopilot.

10. The method of claim 1, further comprising selecting one or more views of the market reference price information for display on a computer display.

11. The method of claim 1, further comprising setting the market reference price as an estimate.

12. The method of claim 1, further comprising receiving one or more of own price data and a list of competitors.

13. One or more means for performing the functionality of any one or more of the above claims.

14. A computer readable storage media including stored instructions that when executed cause the functionality of any one or more of the above claims to be performed.

15. A system that determines a market reference price, the market reference price representing a price used to compare an offered price or service comprising: a CRD selection module that receives competitive rate data from one or more data sources; and based on one or more of an automated method, a weighted average method, a median method, a Kth percentile method and a position in range method, and with the cooperation of a processor, a module that determines and one or more of saves and outputs the market reference price.

16. The system of claim 15, wherein the market reference price is stored and forwarded to one or more price sensitive forecasting engines.

17. The system of claim 15, wherein one or more of the competitive rate data, fare data and price data is used as input for the determining step.

18. The system of claim 15, wherein the market reference price is used for one or more of network prize optimization, input for price sensitive forecasting, displayed on an output device and output in a report.

19. The system of claim 15, wherein the weighted average method further comprises selecting season information and attributes.

20. The system of claim 15, wherein the weighted average method further comprises selecting competitor weight information.

21. The system of claim 15, wherein the weighted average method further comprises selecting one or more boundaries.

22. The system of claim 15, wherein the weighted average method further comprises selecting priority information.

23. The system of claim 15, wherein the weighted average method further comprises configuring an autopilot.

24. The system of claim 15, further comprising a display module that selects one or more views of the market reference price information for display on a computer display.

25. The system of claim 15, further comprising an interface that allows setting the market reference price as an estimate.

26. The system of claim 15, wherein the CRD selection module further receives one or more of own price data and a list of competitors.

27. Any one or more of the aspects substantially as described herein.