Scholarly article on topic 'A framework for managing customer knowledge in retail industry'

A framework for managing customer knowledge in retail industry Academic research paper on "History and archaeology"

CC BY-NC-ND
0
0
Share paper
Academic journal
IIMB Management Review
OECD Field of science
Keywords
{"Knowledge management" / "Customer knowledge" / "Retail business"}

Abstract of research paper on History and archaeology, author of scientific article — Sourav Mukherji

Abstract Customer knowledge can be a critical source of competitive advantage in retail business. In this theoretical paper, three sources of customer knowledge are identified in the retail environment, namely customer transactions, customer interactions and customer communities of practice. Lessons for managing these three types of knowledge are derived from knowledge management practices of knowledge intensive service industries such as management consulting and software development. Finally, a decision model premised on consumer behaviour and purchase characteristics is proposed. The model would enable retailers to focus their knowledge management efforts to leverage the potential of customer knowledge for both productivity benefits and product innovation.

Academic research paper on topic "A framework for managing customer knowledge in retail industry"

IIMB Management Review (2012) 24, 95-103

available at www.sciencedirect.com

SciVerse ScienceDirect

journal homepage: www.elsevier.com/locate/iimb

IIMBMteSment

ai — - - . ...., 1 &

A framework for managing customer knowledge in retail industry

Sourav Mukherji*

Indian Institute of Management Bangalore, Bangalore, India Available online 16 March 2012

KEYWORDS

Knowledge management; Customer knowledge; Retail business

Abstract Customer knowledge can be a critical source of competitive advantage in retail business. In this theoretical paper, three sources of customer knowledge are identified in the retail environment, namely customer transactions, customer interactions and customer communities of practice. Lessons for managing these three types of knowledge are derived from knowledge management practices of knowledge intensive service industries such as management consulting and software development. Finally, a decision model premised on consumer behaviour and purchase characteristics is proposed. The model would enable retailers to focus their knowledge management efforts to leverage the potential of customer knowledge for both productivity benefits and product innovation.

© 2012 Indian Institute of Management Bangalore. Production and hosting by Elsevier Ltd. All rights reserved.

Customer knowledge as driver of success in retail business

The retail industry in India is poised for explosive growth. Traditionally the 'kirana' or neighbourhood mom and pop stores have dominated the Indian retail landscape. Of late large Indian conglomerates such as Reliance Industries and ITC have started to make significant investment in the retail industry while professionally managed retailers like Pantaloon or Shoppers Stop have started to expand rapidly.

* Tel.: +91 80 26993145; fax: +91 80 26584050. E-mail address: souravm@iimb.ernet.in

Peer-review under responsibility of Indian Institute of Management Bangalore.

Indian regulations do not allow foreign direct investment in retail, preventing participation of multinational enterprises. But experts believe that it is only a matter of time before such regulatory restrictions are removed. Therefore, Indian players who want to have a substantial stake in the retail business feel the need to create a stronghold in the market before it is opened to foreign competition. Moreover, some international players such as Metro AG have already started to operate from India by working around the regulatory framework that does not prevent foreign participation in 'wholesale operations', while Wal-Mart has announced an alliance with Bhatri Enterprises.1 Overall, it implies that in the immediate future the Indian retail industry is going to become intensely competitive with a significant change in the nature of competition. Instead of the erstwhile competition between local or regional players

1 'Wal-Mart enters India with Bharti Tie-up', The Hindu, November 28th, 2006.

0970-3896 © 2012 Indian Institute of Management Bangalore. Production and hosting by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.iimb.2012.02.003

who had marginal scale, the competition in future will be between big national and international players who will seek value by relentlessly building up scale and driving down costs.

The customer is likely to gain significantly as a consequence of this changing business scenario. Competition and lower costs would translate into lower prices for the customer. Moreover, professionally managed retail organisations would adopt superior processes and best practices that would translate into a better shopping experience for the customers in terms of choice, availability and convenience. Given such conditions, what would be the drivers of success for individual retailers? On one hand the kirana shops would continue to offer convenience, such as proximity or home delivery and customisation, by knowing in great detail the purchase habits of the neighbourhood. On the other hand national and international players would offer large variety and high quality at a low price, leveraging economies of scale and scope. In order to succeed in such an environment, retailers will need to invest significantly to build scale. At the same time they would need to attract and retain customers in large numbers such that there is adequate return on investment. Given the wide choices available, customers would prefer retailers whom they perceive to be most suitable in terms of meeting their purchase expectations. Retailers therefore would need to have a superior and fine-grained understanding of the customer and 'customer knowledge' would emerge as a key driver for commercial success in the fiercely competitive environment that retail business in India is shaping up to be.

The importance of customer knowledge is well understood and emphasised in the world of business. It has been realised that customers can be sources for innovation (Thomke & von Hippel, 2002) and customers can provide perspectives and suggestions that might have been overlooked or not seen by the organisation. The challenge in the retail industry is to develop a superior understanding of the customer along with the creation of large scale of operations. It is easy to understand customer preferences and accordingly customise products or services when the scale of operations is small. The kirana or mom-and-pop stores epitomise such customisation on a small-scale. However, the retail industry needs customisation on a large scale in order to attract and retain its customers. This would only be possible by adopting a systematic and process oriented approach towards acquisition, storage, analysis and application of customer knowledge — an organisation practice that can be broadly described as 'customer knowledge management'. This paper talks about the critical issues pertaining to the customer knowledge management system that is relevant to the retail industry.

In the following sections, we discuss how retailers worldwide are leveraging data captured from customer transactions. However, there is a growing realisation that such transactional data needs to be complemented with data captured from customer interactions in order to develop insights that can translate into product and service innovations. Since acquisition and management of interactive data is challenging, we look at the knowledge management practices of two knowledge intensive industries, management consulting and software development,

to derive lessons that would enable retailers to deal with such challenges. We classify three sources of knowledge in the retail environment, namely customer transactions, customer interactions and customer communities that retailers can leverage to build a robust knowledge management system. Thereafter, we propose a typology to determine the focus of retailers' knowledge management initiatives depending upon the behaviour and purchase pattern of the consumers. The contribution of this paper is discussed in the concluding section.

Data processing in retail environment

One of the biggest advantages of the retail industry in terms of developing customer knowledge management systems is the availability of data about customer purchase behaviour. However, with the advancement of information technology, organisations today have enormous capacity to store and process data and generate information. Specifically in the retail environment, because of extensive automation such as deployment of point of sales (POS) terminals, or radio frequency identification (RFID) transponders, today it is possible to capture data about consumer behaviour at multiple points. This has led to large retailers worldwide running sophisticated applications for processing the data that is captured, such as data warehouse applications for decision making or data mining applications for obtaining hidden relationships among apparently unrelated variables.

As illustrated in Fig. 1, a data warehouse is a repository of data collected from multiple transaction processing systems. Such data might originate within the organisation (e.g., from the POS terminal within a retail shop) or outside (e.g., data on consumer purchase and credit history

Figure 1 Decision support systems in transaction intensive environment.

obtained from a credit card agency). Data warehouses are intentionally kept separate from transaction processing systems because they are designed specifically for query processing. For example, while transaction intensive systems would typically avoid any redundancy of data, a data warehouse intentionally builds in certain data redundancy to ensure faster response to queries. Some organisations also run sophisticated query and analysis tools on the data stored in the data warehouse. Such tools often deploy algorithms based on artificial intelligence or neural network principles to find out hidden relations among variables (such as the fabled correlation between sales of beer bottles and diapers) and are known as data mining tools. Collectively, data warehouses and data mining tools form part of the repository that organisations, especially retailers, are now deploying extensively for creating sophisticated decision support systems.

While such decision support systems are extremely useful in the retail environment, they suffer from two limitations. First, data warehouses have minimum efficient scale, i.e., they become useful only when an organisation is able to digitally capture very large amounts of data, often running into terabytes. While several organisations have experimented with smaller volumes (often naming them as data marts instead of data warehouses), generating statistically significant relationships between variables have remained quite challenging. This leads to the second limitation of data warehouses, i.e., while information technology driven decision support systems are efficient in capturing and processing transactional data, they do not necessarily generate rich insights that can be used by organisations for decision making (Davenport, Harris, & Kohli, 2001). Transaction processing systems can efficiently capture data about customer behaviour, but they are not adequate to capture the knowledge that the customer possesses (Gibbert, Leibold, & Probst, 2002). Thus, a robust information technology infrastructure is a necessary but not sufficient condition for aiding decision making, a limitation that can be generalised to most enterprise wide knowledge management initiatives (McDermott, 1999).

From data processing to knowledge management

It is necessary for retailers to understand the important levers beyond information technology that would enable them to derive maximum benefit from customer data. They need to understand the best possible ways to collect, store, analyse and deploy knowledge from and knowledge of customers so that such knowledge can provide them with sustainable competitive advantage. One way in which this can be achieved in a relatively short period of time is through the knowledge management practices of other industries — especially those that have been experimenting with and evolving their knowledge management systems and processes. In this paper, we analyse the knowledge management practices of software service and strategy consulting firms in order to derive lessons for customer knowledge management initiatives in the retailing industry. For both the consulting and the software industries,

knowledge is the most critical resource for their business value proposition and successful knowledge management initiatives form the basis of their competitive strategies. Both these industries, possibly more than any other industry, have been at the forefront of knowledge management.

Among the practitioners, strategy-consulting firms like McKinsey & Company were probably the first to realise that information technology focussed knowledge management systems have limited ability to capture certain kinds of knowledge (Bartlett, 1996). More often than not, knowledge that provides competitive edge to individuals or organisations is complex and embedded in a specific context. Such knowledge is difficult to articulate and therefore difficult to be captured in documents or databases. While scholars like Polyani (1966) had explained the difference between two kinds of knowledge — tacit and explicit, consulting organisations have put into practice two kinds of knowledge management systems to leverage the tacit and explicit dimensions of knowledge. Explicit knowledge, which by definition can be easily articulated and captured in documents, can be managed using information technology, e.g., computers, relational databases and communication networks. However, tacit knowledge cannot be articulated or documented. Organisations can only create facilitative conditions such that tacit knowledge can be shared through personal connections, by means of direct communication between experts who posses such knowledge. In their celebrated paper, Hansen, Nohria and Tierney (1999) distinguished between the value proposition of two kinds of consulting organisations — those like Anderson Consulting that were driven by 're-use economics' and those like McKinsey & Company that were driven by 'experts economics'.

Consulting organisations whose value proposition is based on 're-use economics' deliver standardised solutions to customers. They benefit significantly from the 'people to document' approach of knowledge management, which involves articulation and codification of explicit knowledge and disseminating the same across the organisation as 'best practice'. Such organisations discourage their employees from 'reinventing the wheel' because the strength of their practice lies in identifying the best way of performing activities and replicating the same everywhere. They build large data and document repositories and develop sophisticated classification and search algorithms to ensure ease of use, data currency and relevance. However, strategy-consulting firms such as McKinsey or Bain do not base their practices on delivering standardised solutions. Their value proposition, instead, lies in providing unique solutions to problems faced by their clients. Therefore, the purpose of their knowledge management system is not replication or dissemination, but synthesis of knowledge from experts and in the process, development of new knowledge (Ofek & Sarvary, 2001). Their knowledge management systems facilitate people to people connections and subsequent collaborations and as a consequence, these organisations do not focus significantly on building large document repositories. Instead, their knowledge management systems create directories of expertise and project 'snapshots' where employees, individually or collectively, declare their areas of expertise or write briefly about the

problem that their team had solved for the client. The organisation, by means of incentives and other cultural interventions, ensures that consultants in need of knowledge or confronting a problem on behalf on their clients can identify the expert or project team members who can provide them with insights or help them solve their problems in a collaborative manner.

Hansen et al. (1999) advised organisations to choose one of the two knowledge management strategies — codification or personalisation, over the other because they felt that the two approaches do not mix well. However, this is not what has been observed in the software development industry where organisations create knowledge management systems that can accommodate multiple modes of knowledge sharing and generation activities. For example, two of the largest Indian software services organisations, Infosys and Wipro have built up knowledge management systems that not only have a large data repository but also support applications equivalent to an experts' directory intended to bring about collaboration among experts. Such organisations do not intend to choose between one kind of knowledge over another. They believe that for competitive advantage, both explicit and tacit knowledge need to be managed simultaneously, with equal focus. Given the knowledge intensity of software development activities and the rapid change of technology, the shelf life of standardised knowledge and best practices is limited. Therefore, part of the knowledge management initiative focuses on synthesis and generation of new knowledge and insights while the rest focuses on building codified knowledge and libraries of reusable components that can increase the efficiency of software development and project delivery (Mukherji, 2005).

In its steady state, a knowledge management system in a knowledge intensive service industry is likely to have at least three critical components. These are depicted in

Table 1 as 'document repository', 'experts' directory' and 'collaborative platforms'. The document repository and its associated management system focus on collection, storage and access of data and information. Organisations intend to gain efficiency from such management of explicit knowledge, and the purpose of such knowledge management systems is reduction of costs by locating previously generated solutions and adapting these to solve clients' problems. The second component, an experts' directory, intends to make connections between employees. This repository, instead of maintaining documents, contains contact information of experts, their profiles and brief descriptions of their expertise in specific contexts. Employees, when confronted with a problem, can post their queries that are either directed towards certain experts, or open to the entire community of experts. While the experts' directory is not as scalable as the document repository, its value lies in communication of tacit and complex knowledge that would have been very expensive or impossible to codify.

Service organisations have also started building and experimenting with a third component of knowledge management systems, namely 'collaborative platforms'. This is intended to serve the function of synthesis and generation of new knowledge, rather than dissemination of knowledge and is focused on innovation and creativity instead of efficiency through reuse. By leveraging the power of information technology and communication networks, organisations are creating virtual platforms where employee groups with specific interests discuss and collaborate on a topic of their interest while a coordinator tries to provide some direction to the discussions. Since the entire discussion is conducted over an information technology network, it is possible to track, categorise and collate such discussions that sometimes lead to generation of collective insights. Collaborative platforms intend to

Table 1 Three components of knowledge management in knowledge intensive service industries.

Document repository I Experts'directory I Collaborative platforms

Repository of all documents prepared by employees based on their experience of solving problems

Historically divided between documents related to technology and those related to sales and marketing

Collaterals, proposals, letter of references, updated information needed to make proposals and presentations

Reduce rework, increase efficiency

Domain specialists declare themselves as experts. Database maintains their profiles

Queries are posted at the system, or targeted specifically for an expert

System notes whether the expert has been able to solve the problem. Measures level of satisfaction of users

Make people to people connections

Creation of user groups having specific interest

Any employee can join the discussion, even though they are not part of the user group

System maintains log of discussions for future references

Access privileges

create technology-mediated 'communities of practice' (Lave & Wenger, 1991) that are deemed essential for innovation in knowledge intensive industries. In their research on knowledge management systems in professional services firms, Ofek and Sarvary (2001) found that in a competitive market, organisations derived greater leverage from their knowledge management systems if such systems were geared towards synthesis of knowledge and services innovation, rather than towards reduction of costs and increase in efficiency.

Managing customer knowledge

Unlike the services industries just described, where the focus of knowledge management has largely been on employees' knowledge, the retail industry needs to focus on customer knowledge for creating competitive advantage. However, like the software services and consulting industry, the retail industry needs to evolve multiple subsystems of knowledge management in order to derive maximum benefit from customer knowledge. While explicit data generated out of transactions, such as those collected from the POS terminals, can be managed through databases and applications running on top of such data repositories, organisations worldwide are realising the importance of data that cannot be collected through impersonal means. Over and above transactional data, the retail environment has a large potential for generating data through customer interactions. Data generated out of customer interactions is likely to be rich in its tacit content and as a result, might provide organisations with greater insights than those generated from analysis of transactional data (Garcis-Murillo & Annabi, 2002).

Let us first understand what kinds of information can be generated from transactional data. Such data can inform the retailer about a typical customer's purchase basket — the quantities of products purchased and the prices the consumer has paid for them. This could enable retailers to arrive at some measures of price elasticity. Analysis of transaction data can also reveal complementarities between products —the products that are purchased together, which would help the retailer in deciding location of products. If transactional data is linked to information about advertisements or trade promotions, it is possible to identify the impact of such initiatives on purchase behaviour of consumers. Time series analysis of transaction data can also indicate seasonality and cyclicality of consumer purchases and help retailers make decisions about inventory management. As was mentioned earlier, large retailers use various decision support systems for conducting these kinds of analysis. And just as service organisations have made significant investment in data codification, retailers need to make significant investments for capture and analysis of transactional data in order to improve operational efficiencies.

From transactions to interactions

Data captured from transactional systems would not be able to answer questions such as why customers did not purchase certain products even if they had intended to, or

why they chose one product over another. It would be important for a retailer to know what prior knowledge a customer had about a particular product when s/he stepped into the shop and how such knowledge was modified based on the shopping experience. Transactional systems would not be able to identify the compromises that customers make during their purchase or the levels of satisfaction associated with their purchase decisions. While such information about consumer behaviour is invaluable, it can only be captured through a process of interaction or socialisation with the customer. Therefore, customer knowledge management in the retail industry would need to develop systems and processes that would be tuned to facilitate generation and capture of interactive data. Interactive data adds the 'human element' (Davenport et al., 2001) to the transaction data and the knowledge thus captured can be effectively utilised for customisation or even for product innovation. While qualitative market research techniques such as in-depth interviews or focus groups were intended to capture interactive knowledge from customers, these are often sporadic events, extraneous to the regular business activities. What we are discussing here pertains to organisational routines that capture interactive data and utilise knowledge thus generated for decision making. Just as service organisations have been able to design their knowledge management processes for capturing tacit knowledge, retailers need to institute systems and processes that can capture interactive knowledge in a systematic manner.

This is however easier said than done. Organisational hierarchies have been found to be more efficient in solving agency problems than markets (Barney & Hesterley, 1996). Therefore, it is easier to capture tacit and complex knowledge within organisational boundaries by mandating or motivating employees. The challenge for retailers is to ensure the same beyond the organisational boundary, because in their case rich data needs to be captured from customers whose relationship with the organisation is not conducive to sharing or collaborating. Traditionally, customers have been perceived as a source of revenue rather than a source of knowledge. As a consequence, organisations need to devise suitable incentive mechanisms — financial, social or moral, to induce or motivate customers to share their knowledge. On the supply side, organisations need to understand why customers would spend time to provide information to the retailer and whether customers would deem solicitation of such information as invasion of privacy. On the demand side, organisations need to train their employees such that they can elicit information from customers through meaningful interaction. Socialisation — the predominant vehicle for sharing tacit and complex knowledge is premised on depth of relationships between individuals (Nonaka & Konno, 1998). The challenge in the retail environment would be to develop such relationships with the customers within a time period that is long enough to create meaningful interaction, and yet not so long as to make a customer uncomfortable. Organisations also need to act on the information collected and show visible impact of such interactions to the customers in order to motivate the customers to share information multiple times. Davenport et al. (2001) have also warned that not all information provided by the customer is valid. Therefore, organisations

need to have powerful analytical and triangulation processes to ensure the validity of the information that they collect from the customers before they can act on such information.

In effect, capturing interactive data from customers would be both difficult and expensive and organisations need to be conscious of the returns that they get from such investments. Given its potential, it might be tempting for every retailer to start making investments for capturing interactive data. However, as it is difficult to collect such data on a continuous basis, managing interactive data might not be cost effective for every kind of retail business. Just as service organisations need to choose when and where to deploy a people-to-document based knowledge management system and where to utilise a people-to-people based system, retailers need to develop some understanding regarding the utility of transactional and interactive data as contingent on specific kinds of business. This is discussed in the next section.

Contingent theory for knowledge management in retail environment

Research in consumer behaviour indicates that customers do not spend equal amounts of time or attention on every purchase decision and their involvement with purchase decisions varies across a continuum. Degree of customer involvement is a function of the product, the context and the attitudes and values of the customer (Bloch & Richins, 1983). Overall, researchers concur that customers' involvement with purchase decisions is a function of the value that they attach to the product (Zaichkowsky, 1985). Customer involvement can be a discriminatory variable for

knowledge management systems because degree of customer involvement would determine to a large extent the ease with which data or information can be collected from the customer. The greater the customer involvement with a purchase decision, the easier it would be to generate and collect interactive data from the customer. When customers are not involved significantly with purchase decisions they are unlikely to get into a meaningful engagement with the retailer to provide interactive data.

Collection of transactional data is facilitated by greater frequency of purchase. The more number of times a consumer purchases, the greater is the possibility for transactional systems to capture data related to the consumer purchase process. Therefore, frequency of purchase forms the second discriminator for customer knowledge management in the retail industry. Combining these two variables, i.e., frequency of purchase and customer involvement in purchase, we propose a two-by-two matrix that can act as a decision framework for customer knowledge management initiatives. This matrix is depicted in Fig. 2. In an ideal scenario, the two dimensions of such a matrix need to be orthogonal, which is not the case here. In other words, it is possible that frequency of purchase and involvement in purchase might be correlated to one another. For example, the purchase of an item like soap, which is likely to be high on frequency, is a low involvement purchase for the consumer because it is a routine purchase. But purchasing a watch, a relatively infrequent purchase, is likely to get the consumer involved significantly. However, there are frequently purchased items such as food for the calorie conscious or even soap for the beauty conscious that can be a high involvement purchase even though these items are purchased frequently. Likewise, purchase of a digital watch can be of

Frequency of purchase

Transactional data

Consumer Community

Interactive data

High frequency of purchase will generate large quantities of transactional data

Useful for capturing data beyond purchase behaviour, especially in situations of high involvement

High involvement purchase will provide opportunities for interactions that in turn will generate rich data

Consumer involvement with purchase

Figure 2 Typology of knowledge management in retail business.

low involvement because of its commoditised nature, even though customers might not be making such purchases very frequently. Therein lies the strength of this matrix, where all four quadrants would be relevant for differentiating customer knowledge management initiatives in the retail environment.

It is proposed that for items that are high on frequency and low on involvement, customer knowledge management systems should be focused on collecting transactional data. High frequency of purchase will generate significant quantities of data, but because such purchases are of low involvement, it is unlikely that customers would be in a position to provide rich data even if it were possible to interact with them during the purchase process. This is because low involvement purchases do not involve extensive search, neither do they involve comprehensive evaluation of choice alternatives (Olshavsky & Granbois, 1979). On the other hand, for purchases that are of high involvement but of low frequency, customer knowledge management systems would be focused on collecting interactive data. High involvement purchases are likely to be the consequence of active information processing by the customer and the products thus purchased would have significant relevance for the customer (Greenwald & Leavitt, 1984). It is also conceivable that the customer would devote a lot more time to the purchase process, which would provide the retailer with enough opportunities to extract rich data from the customer by means of interpersonal interactions.

The decision quadrant for purchases that are both high on frequency and high on involvement poses a unique challenge. As such purchases involve active information processing by the customer, they have the potential to generate interactive data. However, given the high frequency of purchase, it might be difficult to extract such data from the customer because the customer might not be devoting a lot of time to the purchase process within the retail environment. For example, purchase of breakfast cereals for consumers who are conscious of their weight would be high involvement purchases, because consumers would prefer to make informed choices about the cereals in terms of calorie and nutrition content. However, given that breakfast cereals are purchased frequently, such information acquisition and integration might not be made every time the consumer comes to the retail outlet. Rather than discrete information processing during the occasion of purchase, purchases that are made repeatedly and frequently involve continuous information processing (Hogarth, 1981) very often over a series of purchases or at locations away from the retail space (Hoyer, 1984).

As a consequence, retailers must device a knowledge management system that can pervade beyond the physical retail space in order to capture determinants of consumer behaviour. Such a knowledge management system would be similar to employees' collaborative platforms, the third component of knowledge management systems in the service industry. Like many of the Internet or 'click and mortar' companies, retailers dealing with high involvement high frequency products need to create virtual meeting and discussion places for their consumers. Such customer 'communities of practice' would discuss product attributes that are present or those that are desirable because the

existing products do not fulfil their needs. They can be rich sources of new product ideas and also useful in identifying new usages of existing products. For example, restaurant owners in some of the northern states of India use their washing machines to clean vegetables such as potatoes, or to blend edible liquids such as milk or curd on a large scale.2 While this was common knowledge among the restaurant owners, the manufacturers of the washing machine came to know about this only when a large number of washing machines were returned for repair with 'strange' defects. Once the retailers realised the novel usage of their product, they communicated the information to the manufacturers, who then modified the washing machines so that the machines could act as industrial blenders or vegetable-cleaners. If the retailers of washing machines were connected to the social networks of restaurant owners, they would have identified the novel usage much earlier than it was done in this case.

Retailers can use customer communities of practice to identify emergent consumer profiles such as 'lead users' (Von Hippel, 1986), 'opinion leaders' or 'market mavens' (Feick & Price, 1987) and can target certain marketing efforts towards them. Retailers can also use virtual platforms for trials of prototypes or experimental products and tap into the customer knowledge and experience to develop their product or estimate the market potential. Overall, such a collaborative platform provides the retailer with a useful tool for generating data on consumer behaviour, attitudes and desires beyond the traditional boundary of the retail organisation. However, such a virtual space needs to be carefully managed even though participation by the customer may be voluntary (Gibbert et al., 2002). First of all, customers need to be motivated enough to participate and contribute to the discussions without feeling constrained because their discussions are being observed or monitored by the retailer. Second, retailers need to selectively mediate the discussions in order to make them valuable for the organisation. In this regard, some of the best practices from Internet based organisations or those from the service industry who have instituted collaborative platforms for employees need to be considered and modified according to the needs of the retail industry.

Fig. 2 depicts the matrix classifying consumer purchases in the retail environment along with the suitable customer knowledge management system. Table 2 summarises the functions of the three components on similar lines as observed in the knowledge intensive service industries. While organisations worldwide seem to understand the need for collecting, analysing and disseminating knowledge for creating competitive advantage, several organisations fail to get significant mileage out of such initiatives. One of the reasons for such unmet expectation is the inability on the part of the organisations to link knowledge management initiatives to specific business objectives. The classification presented in this paper is intended to address this issue in the retail environment. Segmenting different kinds of customer purchase behaviour and identifying definite

2 'Innosight in India: Five Lessons from Five Years', H Nair, V Raju & A Mehra, Strategy & Innovation (newsletter), Feb 2012, Vol 10(1).

Table 2 Evolving components of knowledge management in retail business.

features of knowledge management systems suitable for each kind of purchase is a necessary first step towards linking customer knowledge management with commercial objectives of any retail business. Such segmentation would also enable retailers to have a better idea about the returns that they get from investments towards management of customer knowledge. For example, retailers might decide to exclude low-involvement-low-frequency purchases, the fourth quadrant in the matrix, from their knowledge management efforts because given the commoditised nature of such products, the returns from such investments might not be adequate. As the retail environment has large potential to generate data, knowledge management efforts, unless properly focused, might get lost in information overload. The proposed typology would help retail organisations to understand what data to collect, how to prioritise and how to measure the effectiveness of various knowledge management initiatives.

Conclusion

This paper was motivated by the changing competitive scenario in the Indian retail industry. The entry of established players and the subsequent increase in competitive intensity will compel retailers to develop deep competencies that would enable them to survive and win in the marketplace. If customer knowledge, as is widely believed, is going to be an important determinant of success, retailers need to develop competencies and organisational processes to manage customer knowledge such that customer knowledge can translate into business insights that would enable retailers to attract, retain and capture maximum value from their customers. Retailers therefore need to develop an appreciation of both the challenges and the advantages of customer knowledge management.

Most organisations in knowledge intensive industries such as management consulting or software development have deployed knowledge management systems to leverage the collective knowledge of their employees. In this paper, we analysed their knowledge management practices to understand how they overcome various tradeoffs such as management of tacit and explicit knowledge or focussing on knowledge exploration versus knowledge exploitation (March, 1991). We classify three sources of customer knowledge in the retail industry — customer transactions, customer interactions and customer communities. These three sources are likely to pose different tradeoffs before the retailers because the nature of knowledge generated from these sources or their possible usage are going to be different from one another. It is here that lessons learnt from the management consulting or software service industry would enable retailers to institute different knowledge management practices that are best suited to the varied data sources.

However, the uniqueness of the retail environment and specifically the additional challenges involved in collecting customer data limits the generalisability of knowledge management practices of other industries to the retail industry. Therefore, we develop a typology based on consumer behaviour and purchase characteristics that would enable retailers to segment and focus their knowledge management initiatives. Herein, we believe, lies the key contribution of this paper. While researchers and practitioners have realised the benefit of differentiating knowledge management activities to get maximum returns, such differentiation has been based on the logic of value creation (Hansen et al., 1999) profitability earned from customers (Davenport et al., 2001) or industry best practice (Gibbert et al., 2002). This analysis complements such approaches by identifying the important dimension of consumer behaviour as a contingency variable. In the

process, this becomes one of the early attempts to syn-thesise the field of consumer behaviour with that of knowledge management.

This paper is theoretical in nature and the proposed theoretical model needs to be tested and validated. This limitation is an opportunity for future research in the retail industry—first to develop testable propositions and then collect data to validate such propositions such as whether it is useful for collecting data from customer communities of practice in case of high-involvement-high-frequency purchases as suggested by the model, and how such data can be leveraged to create competitive advantage. For this theory to have practical significance, the proposed framework of knowledge management, when applied should translate into better performance and greater profitability for the retailer. We trust that this paper is an important first step towards providing a structure to knowledge management initiatives in retail business that is firmly grounded in the most important variable of such business, namely the consumers and their purchase behaviour.

References

Barney, J. B., & Hesterley, W. (1996). Organizational economics: understanding the relationship between organizations and economic analysis. In S. R. Clegg, C. Hardy, & W. R. Nord (Eds.), Handbook of organization studies. London: Sage. Bartlett, C. A. (1996). McKinsey & Company: Managing knowledge

and learning. Harvard Business School Case, Boston, MA. Bloch, P. H., & Richins, M. L. (1983). A theoretical model for the study of product importance perceptions. Journal of Marketing, 47(Summer), 69-81. Davenport, T. H., Harris, J. G., & Kohli, A. K. (2001). How do they know their customers so well. Sloan Management Review, Winter 2001, 63-73. Feick, L. F., & Price, L. L. (1987). The market maven: a diffuser of marketplace information. Journal of Marketing, 51(1), 83-97. Garcis-Murillo, M., & Annabi, H. (2002). Customer knowledge management. Journal of Operational Research Society, 53, 875-884.

Gibbert, M., Leibold, M., & Probst, G. (2002). Five styles of customer knowledge management and how smart companies use them to create value. European Management Journal, 20(5), 459-469.

Greenwald, A. G., & Leavitt, C. (1984). Audience involvement in advertisement: four levels. Journal of Consumer Research, ff(June), 581-592.

Hansen, M. T., Nohria, N., & Tierney, T. (1999). What's your strategy for managing knowledge? Harvard Business Review, 77(2), 106-116.

Hogarth, R. M. (1981). Beyond discrete biases: functional and dysfunctional aspects of judgmental heuristics. Psychological Bulletin, 90(2), 197-217.

Hoyer, W. D. (1984). An examination for consumer decision making for a common repeat purchase product. Journal of Consumer Research, 11(3), 822-829.

Lave, J., & Wenger, E. (1991). Peripheral legitimate learning. Cambridge: Cambridge University Press.

March, J. G. (1991). Exploration and exploitation in organizational learning. Organization Science, 2(1), 71-87.

McDermott, R. (1999). Why information technology inspired but cannot deliver knowledge management. California Management Review, 41(4), 103-117.

Mukherji, S. (2005). Knowledge management strategy in software service organizations: straddling codification and personalization. IIMB Management Review, 17(3), 33-39.

Nonaka, I., & Konno, M. (1998). The concept of 'Ba': building a foundation for knowledge creation. California Management Review, Spring, 40-54.

Ofek, R., & Sarvary, M. (2001). Leveraging the customer base: creating competitive advantage through knowledge management. Management Science, 47(11), 1441-1456.

Olshavsky, R. W., & Granbois, D. H. (1979). Consumer decision making - fact or fiction? Journal of Consumer Research, 6(September), 93-100.

Polyani, K. (1966). The tacit dimension. New York: Doubleday.

Thomke, S., &von Hippel, E. (2002). Customers as innovators: a new way to create value. Harvard Business Review, April, 2-11.

Von Hippel, E. (1986). Lead user: a source of novel product concepts. Management Science, 32(7), 791-805.

Zaichkowsky, J. L. (1985). Measuring the involvement construct. The Journal of Consumer Research, 12(3), 341 -352.