Marketing research has identified several benefits of dynamic pricing models. For example, dynamic pricing in terms of inventory considerations and time horizons, bundling or personalized offerings has been found to increase sales volume, customer satisfaction and to skim reservation prices. However, today's retailers lack the capability to apply dynamic pricing models because of missing services that realize them and technologies such as smart product infrastructures that deliver the resulting prices to customers. Moreover, dynamic pricing models rely on various price parameters provided by several stakeholders such as retailers (e.g., inventory data), suppliers (e.g., recommended sales price), customers (e.g., buying history or products in the shopping basket), or the government (e.g., taxes). In this sense, interoperability between the price parameters of these stakeholders is required and can be addressed with the help of semantic technologies. Because unprecedented, our objectives are therefore to model, implement and evaluate a framework for pricing services that rely on explicit semantic descriptions and rules. We call them knowledgebased pricing services (KPS). In contrast to dynamic pricing models that are solely based on historical data about prices and customers, the knowledge-based approach uses logical statements to individualize a price. In the current work, we propose a conceptual model for KPS and exemplify its use for a personalized pricing scenario within an in-store shopping situation. Furthermore, we draw implications for business models in the retail industry to motivate the adoption of KPS. And finally, existing tools (e.g., ODRL-Services, SPDO or the Tip 'n Tell smart product infrastructure), which may play a major role for the implementation of KPS, are discussed in order to guide future work. This paper is therefore a first step towards the application of dynamic pricing strategies in retail stores that are based on explicit semantics and which have the potential to improve operational agility in the retail industry through an improved availability and quality of price information. Thus, KPS may foster the evolution of a new business ecosystem around pricing services.
Today’s retailers lack the capability of applying sales strategies that rely on an instant pricing of
products. For example, dynamic pricing in terms of inventory considerations and time horizons
To address this challenge, a price delivery infrastructure and pricing services must be available. On
the one hand, smart products could support the presentation of dynamic prices as they incorporate
information technology for business purposes (Konana and Ray 2007, Maass and Varshney 2008).
In contrast to Electronic Shelf Labelling Systems (Southwell 2002), the concept of smart products is
more flexible because products can directly be identified through the attached barcode or RFID tag
in order to request price information. Accordingly, smart product infrastructures
However, on the other hand, the application of dynamic pricing models does not only require a delivery infrastructure but also a service platform for the management and configuration of pricing models. In particular, dynamic pricing models rely on various price parameters provided by different stakeholders such as retailers (e.g., inventory data), suppliers (e.g., recommended sales price), customers (e.g., buying history or products in the shopping basket), or the government (e.g., taxes). Thus, the evolution of a new retail ecosystem around pricing services is most likely if early adopters apply them. And similar to IT-based ecosystems like the App Store with Apple being the keystone (Iansiti and Levien 2004), also developers and providers of pricing services would be part of this new retail ecosystem (Tian et al. 2008). Figure 1 exemplifies a pricing service ecosystem in the retail industry.
Having a delivery infrastructure and a pricing service platform, retailers and their suppliers would be able to promote dynamic price information to their customers with all the benefits identified by marketing research (see above). This also implies an enhancement of operational agility in terms of improved availability and quality of price information in the retail industry. But bearing in mind many different stakeholders as shown in Figure 1, it is obvious that interoperability is a key challenge for a successful application of pricing services as it has been in IT for over two decades (Legner and Lebreton 2007). To address this challenge, the use of semantic technologies might be a solution as they enable an automated exchange and integration of semantically rich information (Maass and Lampe 2007). Correspondingly, there are three primary advantages for using semanticsbased services: “they promote reuse and interoperability among independently created and managed services; ontology-supported representations based on formal and explicit representation lead to more automation; and explicit modeling of the entities and their relationships between them allows performing deep and insightful analysis” (Sheth et al. 2006, p. 56). 1 This does not apply to high-cost or high-complex products such as individual software or cars. In this case, sales representatives begin negotiations with customers leading to personalized prices.
Service and Taxes Revenue
Government
Price Parameters Retailer
Pricing Service Marketplace
License Fee Pricing Module
Developer
Price Parameters
Solution
License Fee Customer (of Retailers)
Provider
Subscription Fee and Price Parameters
Service and
Revenue
Subscription Fee and Price Parameters
Service and
Revenue
Because unprecedented, our objectives are therefore to model, implement and evaluate a framework for pricing services that rely on semantic descriptions and rules and that we call knowledge-based pricing services (KPS). The contribution of the current work is limited to the first step only. Therefore, we propose a conceptual framework for KPS and exemplify its use in the next section. Then, implications of KPS are drawn for business models in the retail industry to motivate their adoption. Afterwards, existing tools for the implementation of KPS are discussed in order to guide future work. Finally, we conclude this paper with a short summary. 2. A Framework for Knowledge-based Pricing Services Our framework for knowledge-based pricing services (KPS) is based on the work of Spohrer et al. (2007), which discusses steps towards a theory of service systems. They claim that components of a service system are “people, technology, internal and external service systems connected by value propositions, and shared information” (ibid. p. 73). Consistently, our framework also comprises four components (cf. Figure 2). But no recursive definition of service system is used for ease of presentation and which is also consistent with the Ambient Media Framework (AMF). AMF is a recent model proposed by Maass and Varshney (2009) that can be used to design service-based IS environments.
According to Spohrer et al. and AMF, our framework consists of a Social System, a Service System, an Infosphere and a Physical Realization System. The Social System describes roles and objects in a shopping environment, e.g., people such as retailers, suppliers and customers or smart products. These roles and objects interact and communicate according to implicit or explicit pricing rules and protocols. Thus, the social system is strongly related to the concept of people as described by Spohrer et al. For their interactions and communications, role-taking actors and objects make use of pricing services provided by the Service System, which covers both internal as well as external service systems (see Spohrer et al. for further details). The Infosphere provides the semantics that are associated to entities of both the Social System and the Service System, whereas these entities are realized by appropriate Physical Realization Systems, e.g., a semantic web service for pricing bundles. Consistent with Spohrer et al., the Infosphere represents shared information of all entities within a service system whereas the Physical Realization System requires adequate implementation technologies.
Infosphere Knowledge about retailers, customers, products, pricing strategies and legal constraints
Social System Buyer, seller, smart products
Service System
Pricing services
Physical Realization System
Human individuals; non-human, physical objects; data and algorithms Because it addresses the challenge of interoperability within a heterogeneous ecosystem as depicted in Figure 1, the main advantage of our framework lies in the semantics-based component, i.e., the Infosphere. This component is used to describe KPS and their interactions with people and smart products with explicit semantic descriptions and rules. In contrast to dynamic pricing models that are solely based on historical data about prices and customers (e.g., airlines pricing passenger seats), the knowledge-based approach relies on logical statements to determine a price. A literature review on pricing models reveals five domains of knowledge, which are relevant for a Pricing Infosphere. These domains cover knowledge about retailers, customers, products, legal constraints and pricing strategies. An overview of the domains with exemplary pricing aspects and supporting literature is given in Table 1. We briefly discuss each of these domains: Knowledge about the retailer: first, knowledge about the retailer covers pricing parameters that address his or her inventory management, individual terms and conditions for volume discounts or product bundles to name a few. Thus, these pricing parameters describe general requirements and constraints that are derived from the retailer’s strategic goals. Knowledge about the (individual) customer: second, knowledge about an individual customer is predominantly required to implement personalized pricing strategies and thus relevant for KPS. For example, a retailer needs to know the buying frequency or sales volume of a customer to provide a discount if a loyalty programme belongs to the retailer’s strategic goals. In contrast to knowledge about the retailer, which may also include information about customer segments, i.e., retailers’ assumptions of customer segments, this domain focuses solely on information that is related to an individual customer. Knowledge about the product: the third knowledge domain addresses all information related to products a retailer offers, such as the product description or the recommended sales price as determined by the supplier or producer. But the information provided here can also be derived from other sources, e.g., review portals like DooYoo.co.uk, eOpinions.com, Ask.com or Amazon.com, which may influence the pricing of a product as well. Knowledge about legal constraint: this knowledge domain covers all aspects that are specific to a country and their policies related to pricing. For example, the information of value added taxes or customs duty belongs to this knowledge domain. Knowledge about pricing strategies: finally, without knowledge of pricing strategies in general, i.e., the required pricing parameters, effects on sales, fit with strategic goals of retailers, suppliers or producers, the implementation and application of pricing strategies would fail because they build a parenthesis for the other four domains. Correspondingly, the basic entity in a pricing marketplace will be services in the form of instances of these pricing strategies tailored for different retail industries.
All in all, each of these knowledge domains must be made explicit for interoperability reasons, such that the configuration and application of a specific pricing service reduces time and effort for all stakeholders shown in Figure 1, which would in turn increase operational agility in the retail industry.
Examples of pricing aspects Supporting Literature Terms and conditions, Aviv and Pazgal 2005, Bichler and contract, stores, inventory, Kalagnanam 2006, Chinthalapti et al. configuration of pricing 2006, Choudhary et al. 2005, strategies, assumptions about Elmaghraby and Keskinocak 2003, customer segments, sales Gallego and van Ryzin 1994, Kelkar et observations al. 2002, Su 2007, Zhiqiang and Xiong
2008,
Reservation price, shopping
Zhang 2006, Tellis and Gaeth 1990,
Zhiqiang and Xiong 2008 Product description and Aviv and Pazgal 2005, Bichler and specification, product costs, Kalagnanam 2006, Chinthalapti et al. recommended sales price and 2006, Choudhary et al. 2005, its validity time period Elmaghraby and Keskinocak 2003,
Gallego and van Ryzin 1994, Kelkar et al. 2002, Zhiqiang and Xiong 2008
price type, bundling, personalized pricing, dynamic pricing, inventory, price aversion
2007,
b) Value added tax rule:
In summary, the framework can be used to model particular purchase situations that are relevant for retailers, suppliers or producers of products in finding new opportunities to apply dynamic pricing models. Correspondingly, this framework helps to identify the required knowledge-domains, services and technologies, i.e., components that are based upon explicit semantics.
Infosphere Ontologies
Instances Knowledge about
Customers
Knowledge about
Retailers
Knowledge about
Products
Knowledge about Pricing Strategies
Knowledge about Legal Constraints
Mayer's digital image
Smith's digital image
Digital camera's digital image
Personalized Pricing Strategy
Value Added Tax for Greece played by played by played by participates in participates in Social System
Retailer Purchase Situation
communicates Customer realized by
realized by Mayer
Smith Human Individuals
Smart
Product realized by Digital Camera
Non-Human, Physical Objects Physical Realization System e r u g fi n o c y b d e s u
Service System
Service Composition Personalized Pricing Service
Value Added Tax
Service realized by
realized by
Personlized Pricing Service
Value Added Tax
Service Semantic Web Services
In this section, we draw implications of the use of knowledge-based pricing services (KPS) for
corresponding business models in the retail industry to motivate their adoption. One should note
that the following discussion assumes that a smart product infrastructure and mobile devices are
available through which prices can be presented on demand to customers, who stand in front of
products. The following implications cover dynamic pricing models in terms of personalized
offerings, bundling, inventory considerations and time horizons, because these aspects have been
shown to be relevant for the retail industry
Personalized pricing: personalized pricing services allow the application of first degree price
discrimination strategies according to Pigou (1920). Their main objective is to skim consumer
surplus and therefore to increase the retailer’s sales volume. With the exception of individual
negotiations for high-cost or complex products such as individual software products or cars, the
application of personalized pricing strategies in in-store shopping situations is currently
restricted to loyalty cards, that promise a discount after the consumer has made enough
purchases. Smart product shopping scenarios would change these kinds of business models. For
example, if the digital instance of a consumer provides information that is relevant to the
retailer (e.g., whether he is a student or an employee), the latter may provide an individual price
and present it on demand through the channel of smart products. Furthermore, retailers have the
potential to deeply listen into their customers’ interests and pricing needs if they capture the
shopping interactions and buying behaviour. This kind of ‘listening in’ might also lead to new
personalized pricing models (Urban and Hauser 2004). In addition, also suppliers or producers
might price their products according to the input of individual consumers. For example, to
increase customer retention a producer may give a discount on his product because the
customer already owns another product of that producer. In this sense, business models related
to personalized pricing might also foster the cooperation between retailers, suppliers, producers
and the end-consumer.
Product bundling: A KPS for product bundles would enable retailers and suppliers2 of
complementary products (e.g., digital cameras and memory cards) to negotiate a discount for a
product bundle instantly by semantic reasoning mechanisms at that point of time, when a
customer is interested in those products. This would not only make the product bundle more
attractive to the customer but would also increase sales volume and profit of the retailer and its
suppliers, if the customer buys two products instead of one. These kinds of scenarios would
foster the cooperation between retailers and suppliers of complementary products in terms of
price negotiations, whereby the tie-in product (here, the memory card) would generate less
profit through a higher discount, because it was tied-in by the digital camera
is still available might be priced higher in order to reorder them timely. In this sense,
knowledge-based pricing services might change existing business models due to stock
information that is instantly available at that time a consumer is interested in a product.
Accordingly, the question will arise how to use this stock information in order to improve both
availability and sale of products by adaptive pricing mechanisms.
Pricing that considers time horizons: by having a finite time horizon for selling products (e.g.,
for seasonal products), the objective of retailers is to maximize expected revenues until the end
of a selling season by pricing products adequately
In addition to these implications, it must be considered that each of the pricing scenarios discussed could have a negative impact on consumer behaviour and the image of a retailer, supplier or producer. Correspondingly, two of the most important topics that must be addressed by all business models are security aspects and price transparency. The first one is important according to the framework presented in the last section, because there exist various knowledge domains and explicit descriptions about customers, retailers and products for which access models must be defined to prevent fraudulent use. By contrast, the second aspect is crucial in terms of consumers’ negative attitudes towards prices, that vary from day to day or from one friend to another. Thus, it must be always made transparent how a particular price is calculated to make the price traceable for the consumer. Otherwise, consumers would rather avoid retail stores that offer dynamic prices without the rationale behind it.
In this section, existing tools for E-Commerce transactions are briefly described, which can be used to implement knowledge-based pricing services (KPS). First of all, we start with adequate delivery infrastructures into which KPS can be embedded. Because there exist many potential infrastructures and related systems such as electronic shelf labelling systems, we briefly describe three of them, which are particularly related to the concept of smart products. Then, front-ends for KPS are provided, before we finally list standards for pricing services and semantic data models.
First of all, the smart product infrastructure Tip‘n Tell (Maass and Filler 2007) represents an
adequate delivery platform for KPS as it uses a semantic framework and reasoning mechanisms to
provide information about products. Within a Tip’n Tell shopping scenario, physical products are
equipped with RFID tags to identify them and to start human product interaction through a mobile
shopping assistant. Then, if a product’s RFID tag was scanned by a consumer, the mobile shopping
assistant informs the Tip ’n Tell web service (Java & Axis2) about this product. The server
components manage the semantic data pool using the semantic framework Jena2 to allow the user
to request the price information of a product. Currently, we are developing a new version of Tip‘n
Tell, which allows to plug-in external OSGi modules such as pricing services. A second
infrastructure is Fosstrak
In addition to backend infrastructures, price information of products must be presented to customers
in in-store shopping situations. In contrast to electronic shelf labelling systems (e.g., Southwell
2002), mobile devices are more flexible with regard to the concept of smart products.
Correspondingly, first applications are being developed for consumers to communicate with
physical products (Maass and Varshney 2008). Examples are Shoppers Eye
The last building block for the implementation of KPS is related to standards for price descriptions and semantic data models. The latter are useful, because they can integrate standardized and nonstandardized product information (Maass and Lampe 2007). In order to store and maintain price information of products, there exist several standards. Consistent with our knowledge-based approach and explicit product descriptions, Kelkar et al. (2002) reviewed existing standards for electronic product catalogues that are based on XML and which can be used to model and define prices. In detail, they evaluated the following standards: cXML, xCBL, BMEcat, EAN.UCC, OAGIS, RosettaNet. As a result of their theoretical and empirical analysis, Kelkar et al. proposed a new general price model, because the evaluated standards cover real world price models in a limited way.
Furthermore, contracts and prices can be modelled with the Smart Product Description Object (SPDO, Janzen and Maass 2008b). SPDO is a semantic data model for products and works hand in hand with the Tip‘n Tell infrastructure (see above). The SPDO consists of five facets of which the product description and business description are used to model the corresponding price and contract information of a product. GoodRelations is another example of a semantic data model for products (Hepp 2008), which is also relevant for future work related to KPS. In addition, semantic data models can be complemented with the use of rule languages (e.g., SWRL) and reasoning mechanisms in order to request or calculate prices in a knowledge-based fashion. And finally, because laws and contracts are required to deploy and maintain sophisticated and complex services (Spohrer et al. 2007), the licensing of pricing services could be managed with the ODRL Services (ODRL-S) profile (Gangadharan et al. 2008). This profile is based on XML and fits therefore to our framework for KPS.
In summary, all of these tools, i.e., delivery infrastructures, consumer front-ends, pricing standards and semantic data models are starting points for our future work that will predominately address the implementation and evaluation of KPS.
In the current work, we proposed a framework for knowledge-based pricing services (KPS) that rely on explicit semantic descriptions and rules. In combination with adequate delivery infrastructures such as the smart product infrastructure Tip’n Tell, they have the potential to increase operational agility in the retail industry, because they enable retailers to implement dynamic pricing models in terms of inventory considerations, time horizons, bundling and personalized offerings, i.e., all kinds of price models, which cannot be applied dynamically with static price tags as of today. KPS use logic statements to derive prices and therefore extend dynamic pricing models that are solely based on historical data. In order to motivate the adoption of KPS, we have also drawn implications for corresponding business models in the retail industry. Finally, a brief overview of existing tools was given that will be helpful for a reference implementation of KPS we aim to develop and evaluate in future work.
Gangadharan, G. R., D’Andrea, V., Iannella, R., & Weiss, M. (2008). ODRL Service Licensing Profile (ODRL-S). In R. Grimm, B. Hass & J. Nützel (Eds.), Virtual Goods: Technology, Economy, and Legal Aspects. Hauppauge NY, USA: Nova Publishers.
Hardestya, D. M., Beardenb, W. O., & Carlson, J. P. (2007). Persuasion knowledge and consumer reactions to pricing tactics. Journal of Retailing, 83(2), 199-210.
Web. Proceedings of the 16th International Conference on Knowledge Engineering and Knowledge Management (EKAW2008). Acitrezza, Italy
School Press.
Workshop AmI Blocks at the European Conference on Ambient Intelligence (AmI-08).
Nürnberg, Germany Janzen, S., & Maass, W. (2008b). Smart Product Description Object (SPDO). Poster Proceedings of the 5th International Conference on Formal Ontology in Information Systems (FOIS).
Saarbrücken, Germany Karpowicz, A., & Szajowski, K. (2007). Double optimal stopping times and dynamic pricing problem: description of the mathematical model. Mathematical Methods of Operations Research, 66(2), 235-253.
Keegan, S., O'Hare, G. M. P., & O'Grady, M. J. (2008). EasiShop: Ambient Intelligence Assists
Catalogs Based on XML. Proceedings of the 11th International World Wide Web Conference (WWW2002). Honolulu, Hawaii, USA Konana, P., & Ray, G. (2007). Physical product reengineering with embedded information technology. Communications of the ACM, 50(10), 72-78.
Kourouthanassis, P., & Roussos, G. (2003). Developing Consumer-friendly Pervasive Retail
Kowatsch, T., Maass, W., Filler, A., & Janzen, S. (2008). Knowledge-based Bundling of Smart
Mobile Business (ICMB 2009). Barcelona, Spain
Information Systems (MCIS 09) Athens, Greece Legner, C., & Lebreton, B. (2007). Preface to the Focus Theme Section on Business Interoperability: Business Interoperability Research ? Present Achievements and Upcoming Challenges. Electronic Markets, 17(3), 176-186.
Liu, Y., & Zhang, Z. J. (2006). The Benefits of Personalized Pricing in a Channel. Marketing
Science, 25(1), 97-105.
Tangible Shopping. Making Semantics Work For Business, Part of the 1st European Semantic Technology Conference. Vienna, Austria
Environments. 20th Bled eConference (http://domino.fov.uni-mb.si/proceedings.nsf/2007), Bled, Slovenia.
Conference on Information Systems (ECIS). Galway, Ireland
Information Systems (MCIS 09) Athens, Greece Maass, W., & Lampe, M. (2007). Integration of Standardized and Non-Standardized Product Data.
141-146). Berlin: Springer.
Maass, W., & Varshney, U. (2008). Preface to the Focus Theme Section: 'Smart Products'.
Electronic Markets, 18(3), 211-215.
Drugs. SIG-Health Pre-AMCIS Workshop at the 15th Americas Conference on Information Systems (AMCIS 2009). San Francisco, USA Pigou, A. C. (1920). The Economics of welfare. London: Macmillan.
Resatsch, F., Sandner, U., Leimeister, J. M., & Krcmar, H. (2008). Do Point of Sale RFID-Based Information Services Make a Difference? Analyzing Consumer Perceptions for Designing Smart Product Information Services in Retail Business. Electronic Markets, 18(3), 216 231.
Sambamurthy, V., Bharadwaj, A., & Grover, V. (2003). Shaping Agility Through Digital Options: Reconceptualizing the Role of Information Technology in Contemporary Firms. MIS Quarterly, 27(2), 237-263.
Sheth, A., Verma, K., & Gomadam, K. (2006). Semantics to Energize the Full Services Spectrum.
Southwell, M. (2002). Beyond the POS. Business Trends, July 2.
Spohrer, J., Maglio, P. P., Bailey, J., & Gruhl, D. (2007). Steps Toward a Science of Service
Systems. IEEE Computer, 40(1), 71-77.
Stigler, G. J. (1963). United States v. Loew's Inc.: A Note on Block-Booking. The Supreme Court
Review, 1963, 152-157.
Stremersch, S., & Tellis, G. J. (2002). Strategic Bundling of Products and Prices: A New Synthesis for Marketing. Journal of Marketing, 66(1), 55-72.
Su, X. (2007). Intertemporal Pricing with Strategic Customer Behavior. Management Science, 53(5), 726-741.
Tellis, G. J., & Gaeth, G. J. (1990). Best Value, Price-Seeking, and Price Aversion: The impact of
Information and Learning on Consumer Choices. Journal of Marketing, 54(2), 34-45. Tian, C. H., Ray, B. K., Lee, J., Cao, R., & Ding, W. (2008). BEAM: A Framework for Business
Ecosystem Analysis and Modeling. IBM Systems Journal, 47(1), 101-114.
Urban, G., & Hauser, J. (2004). “Listening In” to Find and Explore New Combinations of Customer
von Reischach, F., Guinard, D., Michahelles, F., & Fleisch, E. (2009). A Mobile Product
International Conference on Pervasive Computing and Communications (PerCom 2009).
Galveston, Texas, USA Youll, J., Morris, J., Krikorian, R., & Maes, P. (2000). Impulse: Location-based Agent Assistance.
Fourth International Conference on Autonomous Agents. Barcelona, Spain
Pricing Model. 4th International Conference on Wireless Communications, Networking and Mobile Computing (WiCOM) Dalian, China