Volume 6, No. 1, Art. 13 – January 2005
Observations on Observation in India's Dynamic Urban Markets
Nikhilesh Dholakia & Piyush Kumar Sinha
Abstract: Urban India is witnessing a rapid revolution in its commercial retail spaces. These transforming commercial landscapes constitute a rich arena for qualitative research using, among others, observational methods. In this paper, we present observations about the changing urban retail scene of India. We provide emergent themes that we have already found in observations so far, and also reflect on the challenges of carrying out such observations in the Indian context.
Keywords: observation, retailing, India, shopping, consumer behavior
Table of Contents
1. A Unique Setting
2. India's Retail Scene
3. Challenges of Observing Retailing Qualitatively
4. Observations from Retail Observation in India
4.1 Suddenly there's choice
4.2 The invisible glass pane
4.3 Interacting shoppers, not lone stalkers
4.4 Towards linguistic cosmopolitanism
4.5 Immigrants break the stranglehold of tradition
4.6 Cart pushers
5. Lessons from the Field
5.1 Conscious of eavesdropping strangers
5.2 Navigating the linguistic melange
5.3 Traditional stores: finding a perch, taking in the gestalt view
5.4 Easy observations but elusive conversations
6. Concluding Notes
Fuelled by a fast-growing Information Technology sector and other economic forces, the urban market spaces in India are transforming rapidly. Traditional tiny retail stores clustered in busy bazaars or sprinkled in the suburbia are being joined by large, modern Western style retail malls and superstores. 
The changing urban markets of India provide a fascinating context for studying people in shopping situations. Because the change is fast and dramatic compared to the decades-long retail evolution in the Western world, the social and cultural transformations associated with retailing in India provide insights about human dimensions of rapid economic change. Not only business managers but also social scientists and humanities researchers can benefit from such insights. 
Observational methods are a great way to derive such insights (UNDERHILL 1999). Although sophisticated photo/video recording and radio-frequency (RFID) tagging methods are being deployed to "observe" shoppers in the structured stores of the West (SORENSEN 2003), in the rapidly transforming retail context of India, there is no substitute for careful human observation and reflection. Through unobtrusive means, such methods can provide windows into how traditional and emerging shopping cultures of India are interacting. 
With over 13 million retail stores, India has more retailers than any other country in the world. Yet, these retailers are tiny. The typical traditional Indian urban retail store has only 30 to 50 square meters, literally a hole in the wall. Stores in semi-urban and rural areas are even tinier than this. To maximize the use of space, merchandise is stacked along all available walls and floor space. The retailer leaves barely enough space for a stool or cushion on which s/he can sit. Like the kiosks found in European cities, a counter is put up as a barrier between the shoppers and the merchandise. 
In the urban centers, these traditional retail stores are being supplemented by large malls. Some of the malls are being created in traditional shopping locations, by aggressive land use strategies of private developers, while others are appearing in outlying areas near intercity highways. 
Like a forest fire starting at the edges, the retail change in India is hot and fast-spreading at the visible urban edges. From "serviced retailing," where the customer approaches the retailer and asks for specific items which the retailer fetches from behind the counter, urban Indian consumers are suddenly encountering "self service" models of retailing. In the major cities, every few months major shopping centers, branches of large-format retailers or new types of mega-store formats hitherto not found in India are opening up. Again, using the analogy of a newly triggered forest fire, it is too early to tell whether large swaths of the forest would be up in flames or just a few edges would be consumed by the new-retail inferno. It remains to be seen whether the "malling of India" would remain an urban phenomenon or spread widely throughout the country. 
Typical methods used to study people in retail and shopping situations rely on observations—by simple human tracking, using recording devices, or a combination of these—and sometimes supplemented by surveys. The goals of such research are often commercial, or at least in producing knowledge with potential commercial value (SORENSEN 2003). 
Because of commercials goals, even if done unobtrusively, observation in retail contexts is often highly structured. It entails pre-mapping and dividing the retail store into zones, precise recording of items visually scanned and/or touched by shoppers, accurate measurement of time spent "shopping" (examining, comparing, etc.), and so on (SORENSEN 2003, UNDERHILL 1999). 
By contrast, qualitative observation is free-form and naturalistic. As ADLER and ADLER (1998) note:
"Qualitative observation is fundamentally naturalistic in essence ... and follows the natural stream of everyday life. As such, it enjoys the advantage of drawing the observer into the phenomenological complexity of the world, where connections, correlations, and causes can be witnesses as and how they unfold" (p.81). 
Pure observation per se, even when not guided by commercial goals, is likely to produce relatively less dense descriptions than ethnographies resulting from immersed participant observation in a social setting. Commercial goals can narrow the scope and "thin out" the nature of observation. In this sense, we can characterize the accounts resulting from commercial retail observation as "thin descriptions" (or snapshots) in contrast to "thick descriptions" that result from participant observation of communities over prolonged periods. It is possible, however, to derive strategic inputs from thin descriptions too. SINHA and UNIYAL (2003, 2004) have segmented shoppers and identified business opportunities for retailers in India. 
A challenge for qualitative researchers of the retail scene is to add some measure of "thickness" to observational accounts of retailing that are essentially thin, because of the ephemeral and non-participant nature of the observation. Layering of what is observed with observations regarding what is observed is one way of achieving a greater measure of thickness in descriptions. 
Urban Indian shoppers are witnessing a rapid change in the shopping options available to them. Even the most urbanized of the shoppers, however, have not shifted totally to the new open format stores. The emergence of new formats, in fact, has caused shoppers to divide their purchases between traditional and new formats. Typically, one of the retail formats—whether traditional or new—becomes the major source for a product category such as bread or shampoo, while other formats are visited occasionally. This is creating a retail canvas that is varied and interesting. As they visit the new format stores, Indian customers seem to behave in ways that are atypical of Western, developed contexts (SINHA & UNIYAL 2004). Some of these behaviors are a reflection of their earlier behavior patterns, established with decades of small "serviced" store format, while other behaviors are specific to the new context. 
The observations that follow are actually meta-observations: reflective observations based on multiple studies. The main set of studies, supervised by the second author of this article, were in-store observations of shoppers visiting both the traditional and new format stores. These stores dealt with products such as books, grocery, medicine, shoes, apparel, household appliances and lifestyle products such as greeting cards and gift items. Male as well as female shoppers were observed. The observation notes pertained to the shopper and not the accompanying person, except when the latter played an important role in decision-making. The trained observers, the field staff, followed the shoppers during the process and kept a safe distance so that they could also hear the conversations but not interfere with the shopping process. In the traditional stores, where the small size of the store did not allow movement, the investigators found a vantage position that gave them a good view and also did not expose them to the shoppers. Observational notes of those shoppers, who appeared to be conscious that they were being observed, were dropped. The agreement of the store managers was sought before starting the study in a specific store. The personnel at the stores were also briefed about the study so that they did not confuse the investigators with customers and disturb the observation process. 
The observations were recorded as field notes. The investigators would check for completeness of their notes of one observation before picking up another shopper to observe. Before starting the project, the second author of this article—supervisor of the project—observed the in-training investigators, or the "field staff," as they observed a few of the shoppers. The field notes of the supervising author and the field staff (the in-training investigators) were compared and the final observations started only after these notes were found to capture almost similar information. This exercise was carried in both formats. The field notes were then transcribed. A total of 282 shoppers were observed. 
In the meta-observations that follow only those aspects of shopper behavior that occurred often in multiple observations have been used to draw the inferences. Some of these meta-observations also draw upon a similar earlier study conducted to understand children's responses to point-of-purchase communications (MOHANKUMAR, SINHA & KRISHNA 2003). 
Customers tend to come to the stores with a choice set. This set varies according to the extent of planning that the customer undertakes before reaching the store. With more planning, this choice set narrows. While buying from the traditional format (serviced) stores, the customers carry lists. These lists, in many cases, mention the brand name of the product. Generally, the retailer simply fetches the sought brands and collects them on the counter, ready for tallying and payment. Only in case of non-availability of the sought brand, the serviced-store retailer may mention alternative brands. In some cases, retailers—presumably motivated by incentives offered by brand marketers—suggest newly launched brands to the customer. Such suggestions from the retailer are more likely to be proffered to customers who are "regulars" in terms of patronizing the store. 
Even in case of the new format stores, the customers carry lists. They keep looking at the list so as not to miss any item. Unlike the serviced store, however, a large number of customers in the new format stores browse and consider several brands before choosing. In some cases, they do change the brands that they had on their lists. The open display of products leads to expansion of their consideration sets. Many of them also buy more than their lists after browsing through the displayed merchandise. 
While product substitution in the traditional serviced stores depends almost entirely on the retailer's willingness to offer alternative brands, customers exercise this choice on their own in the new format stores. Interestingly, in many categories, customers end up buying their regular brands even after browsing and evaluating other options. Decades-long ingrained buying habits and brand loyalties do not disappear with sudden emergence of choice. 
In the traditional serviced store, since customers do not have access to the merchandise, they generally refrain from browsing and getting the touch and feel of the products. In most cases they window shop. They scan the store, but most of them do not ask the retailers to bring forth alternative brands to the counter for examination. Customers in traditional stores do not touch the merchandise, even though some of it is displayed on the counters. 
While this is understandable in the old format stores, similar behavior in the new format store is perplexing. One of the biggest differences in the two formats is the liberty provided to the customers to browse. Our observations show that most customers in the new open format stores do not touch the merchandise. They evaluate the merchandise visually and only after narrowing their choice, they touch and pick the chosen item. Since most of the displays in the old format stores have glass panes and hence beyond picking, customers tend to only look at brands and products from distance, and approve or reject items based on the visual cues they can discern in this manner. Interestingly, even when the customer is allowed to touch and feel the merchandise, there seems to an invisible glass pane that inhibits interacting with the merchandise. While one can explain this for fragile items that might be subject to breakage, this behavior is observed even in case of books, household appliances and apparels. 
In the new format stores, a large proportion of shoppers come with another person accompanying them. There is also a tendency to shop with the family. This leads to situations where a lot of discussion happens at the store. Shoppers discuss about the merchandise and their choices in a low hushed voice, conscious that others do not hear the conversation. Sometimes when they realize that the store is very quiet, they stop discussing—fearful that it would be easy for other shoppers and store personnel to eavesdrop. In stores where there is either no place to sit or the discussion relates to "sensitive" price or other personal choice factors, such as gifting for a close relative, shoppers step out of the store. They sort out the issue through a less inhibited discussion outside, and then walk back into the store to buy. In many of these animated discussions, the salesperson is also consulted, especially about possibility of returning the merchandise if not found suitable. With the widening availability of mobile phones, some shoppers make telephone calls to their family members for advice and to obtain "remote approvals." 
The accompanying person provides encouragement to try more items. They become the "approvers," especially in case of apparels and shoes. This is irrespective of the gender of the shopper or the accompanying person. In case of younger shoppers, the discussion is mainly about the store; its lighting, merchandise or other facilities. We have also observed that compared to men, women are more open (less inhibited) in discussing their purchases at the stores. 
Most observations on which these emergent themes are based were done in the Western Indian state of Gujarat, where the mother tongue and regional language is Gujarati. It is interesting to note that shoppers change their language as they move from the old format to the new format stores. In the traditional "serviced" old format stores, shoppers talk to the salesperson in Gujarati. But when they visit the new format stores, they start speaking either in Hindi, the national language; or English, if they have English-speaking competence. This behavior is noticed not just in the way the shoppers talk to the sales staff but even when they talk among themselves. Quite clearly, there is a felt need to project a strong image of cosmopolitanism in the modern, Westernized milieu of the new format stores. In many cases, even when the normal flow of conversation is in Gujarati, shoppers switch to English or Hindi when they want to be assertive or express strong dissatisfaction. 
In the old format stores, shoppers use not only the regional language Gujarati but also lace it with colloquialisms. At the new format stores, shoppers avoid heavily colloquial language and employ relatively more formal language in their conversations. In the traditional old format store, there is an assumed social hierarchy of the shopkeeper being servile to the shopper. The shoppers' language therefore has tonalities of impatience and intolerance. Some shoppers find it perfectly acceptable to be rude, especially to the shopkeeper's assistants, in the old format settings.1) Such behavior is more evident when the customers are familiar with the retailer: familiarity provides a license for rough language. 
In the new format settings, unsure of the social protocols and perhaps a bit awed by the new ambiance, shoppers are accommodative and tolerant towards the sales help. Besides the use of formal forms of Gujarati, and even switching to cosmopolitan Hindi or English, shoppers adopt demeanors vis-à-vis sales staff of "polite equals" rather than "master-servant." 
In Ahmedabad—the largest city of Gujarat—for many years, there was just one "partial self-service" grocery store. All attempts to establish additional large self-service supermarkets had not succeeded. It was traditional for Gujarati families (and still is for many) to buy cereals, grains, pulses (lentils), and cooking oil once a year and keep stocks at their homes. The dry desert-like climate facilitates long-term storage without spoilage. In the post-harvest months when the supplies are strong and prices are low, traditional grocers assist such bulk buying behavior by making home deliveries of bulk items. Such bulk purchases constituted more than 60% of the total grocery purchase of most Gujarati households. The remaining 40% purchases of "filler items" represented too small a market for large format, self-service supermarket-style grocery store to survive on. 
Retail change accelerated as the city witnessed migration from other states. Many came from tropical regions of India where spoilage is quick and long-term food storage is not an option. These immigrant customers were accustomed to buying most of their requirements monthly, or even on a weekly or daily basis. As these immigrant customers started settling in Ahmedabad, the new format supermarket style stores started flourishing. These immigrant customers came to the store with families. They bought beyond their pre-planned lists as well as on impulse. They visited the stores even for casual purchases.2) Of course, as the new retails formats became available, many Gujarati households also started emulating these new retail behaviors. We find that in parts of Ahmedabad that do not have such migrants, the retail canvas is still dotted with old format stores. 
In most cases of new format supermarket style retailing, the husbands tend to be mere cart pushers and move along like zombies. They simply follow their wives. If the husbands impulsively pick something, the wives question them; and in many cases, the husbands put the merchandise back on to the shelf. When the merchandise is part of the list that the wives are carrying, and the husband picks the item, it generates a nod of approval from the wife. This is witnessed in categories such as biscuits/cookies, toothpastes, pickles, snacks, plastic ware and other peripheral items. Of course, husbands do sometimes inquire if the family needed items that the wives may have missed while browsing. While wives fill the cart, payments in most cases are made by the husbands. The task of dealing with the cashier is seen as an activity appropriate for the male in the household. 
In the very affluent families, these patterns exhibit variations. Rather than the husband, the cart pusher may be the maid or the driver (chauffer). Of course, in such cases, almost all the picking and paying is done by "madam," the housewife-employer. While not yet much evident in Ahmedabad, in bigger metropolitan areas such as Delhi or Mumbai, there are some observed cases of maids coming to shop on their own, using lists provided by the employer. To command appropriate respect from the store help, the maids "dress up" in such situations. 
While observational methods are very useful in studying the evolving canvas of urban retailing in India, the complex Indian context requires several adjustments and modifications to these methods. 
In American or British observational studies of retailing, since the observed shopper—the informant—is most likely on a lone shopping mission, it is relatively easy to train observers to be unobtrusive. UNDERHILL (1999) offers this advice for the Western retail context:
"We find that positioning behind the shopper is a bad idea—we all know the sensation that we are being watched. But if you stand to the side of a shopper, his or her peripheral vision "reads" you as just another customer—harmless, in other words, and barely worth noticing. From that position you can get close enough to see exactly what a shopper is doing" (UNDERHILL 1999, p.14). 
Indian customers tend to be very conscious of the surroundings while shopping. While younger shoppers are somewhat less so, most customers are very sentient about other shoppers or store personnel watching or hearing their conversations with the accompanying persons. The moment they become aware of possible eavesdropping, their behaviors change. While conducting observational research, many informants need to be dropped because they change their behaviors abruptly. The issue becomes very delicate especially when male field workers are sent to observe women shoppers. Women observers seem to be more discrete, less threatening—and usually perform better as observers than men. 
While conducting observational studies in Ahmedabad, in Gujarat state of India, it is important to ensure that the observers understand at least three languages: the regional language Gujarati, the national language Hindi, and English. Customers in Ahmedabad stores could speak any of these three languages. The order, however, is not fixed and also many customers tend to mix the three languages. The order and mixing tend to change with the format of the store and the kind of products being bought. Also, due to migration from other linguistic regions and the increasingly cosmopolitan nature of the population of Ahmedabad, one may find as many as eight other languages spoken by the customers. This poses a unique problem for the field investigators. It is very difficult to find investigators that speak or understand so many languages. In some instances, a customer and her companion may be conversing in English or Hindi, and then suddenly turn to Tamil—a South Indian language—to ensure privacy in a public space. The recording of the conversation in such cases becomes difficult, and the particular informant may have to be abandoned. The multiplicity and mixing of languages complicate transcription and translation. Added to this is the fact that the conversations many a times include very colloquial terms that are specific to a particular context. A good translation and understanding would require contextual familiarity. Researchers need to spend quite some time with the observers and the translators. 
Just as in the West, it is comparatively easy to carry out observational studies in a large format, self-service stores of India. Customers tend to be engrossed in shopping. A "mystery shopper" approach works very well in such stores, often by sending two observers who can pretend to be companion shoppers. 
In India, however, the vast majority of stores are small and serviced. In such stores, finding an observational vantage point is not easy. Even when the stores are the new format (self-service), since they are much smaller than American stores—most have 100 to 300 square meters—even in such self-service settings, it becomes difficult to maintain a safe observational distance. The observer carrying a notepad becomes too conspicuous. In many cases the observers had to start arranging the shelf or taking stock of the merchandise to remain ambiguous. 
In traditional outlets, the customer does not move around—s/he is fixed in front of the counter. But there is hardly any space in the store and it is difficult to hide anything from the customers. In such cases, the investigator had to take the role of the bookkeeper sitting besides the shopkeeper. Taking up the role of a salesperson or an assistant in the store is unsafe as the investigator would not be aware of the merchandise, stocking and pricing. Also it would divert the attention of the investigators. In some cases, the observers had to become the distributor's salesperson and help out by fetching the merchandise. Also, in the traditional outlets there is very little browsing or lingering. Customers do not spend much time at the store and are very active in collecting information either from the storekeeper or watching the point of sale displays and posters. Thus, in contrast to retail observers in the West, observers in the traditional old format Indian store have to be trained to be very quick to take in a sweeping, gestalt view of the brief episodes of customer visits, make sense of and take notes about these episodes. 
While making notes about customer movements through a new format retail store is relatively easy, recording conversations is difficult. And yet, because of the "accompanied" nature of a lot of shopping in the new format stores, it is important to have some feel for the conversation occurring while shopping. The observers have to maintain a safe distance and many times stay out of the range to hear the conversation clearly. Customers also tend to modulate their voice and also many times change their language, often with the very intention of thwarting any type of eavesdropping. The muttering, swearing and something shared in a hushed voice also get lost. In case of the stores located on the high streets, the noise of the passing traffic—honking being the preferred mode of navigation—interferes with listening. The music in the store also becomes a hindrance. Similarly, conversations among other shoppers and salespersons also hinder listening. Very often, the records obtained offer at best a gist of the conversations rather than verbatim accounts. 
In the transforming urban markets of India, a fascinating new landscape for qualitative research, especially employing observational methods, is opening up. Retail stores of course provide wonderful and rich contexts for studying behaviors in the context of a rapid-flux consumer culture. Other possibilities in commercial settings are also opening up: food courts, amusement arcades, fast-food restaurants, food courts, coffee shops, and so forth. These offer totally new commercial spaces that the Indian consumer is stepping into with some trepidation, but with considerable zest, and trying to navigate in culturally unique ways. As these changes happen, the traditional commercial sector is not sitting still—it is innovating, modernizing at the edges, offering aggressive incentives or superior service. Therefore, from the perspective of studying retailing, "everything is new again" even in the established and serviced retail formats. 
1) Like the German Sie and Du, Gujarati has three levels of formalisms of "You"—the honorific Aap (used very sparingly), the polite Tamé, and the diminutive Tu—used for children, very close friends and relatives, and to address servants. In the old format stores, Tu is frequently used to address the store staff but in the unfamiliar milieu of new format stores, shoppers play it safe and employ the polite Tamé. <back>
2) Because India's states and provinces are mostly linguistically created, it is easy to spot the "immigrants" in retail settings in Ahmedabad: these people do not converse in Gujarati, the local language. <back>
Adler, Patricia & Adler, Peter (1998). Observational Techniques. In Norman K. Denzin & Yvonna S. Lincoln (Eds.), Collecting and Interpreting Qualitative Materials (pp.79-109). Thousand Oaks CA: Sage.
Mohan Kumar, Anuradha; Sinha, Piyush Kumar & Rajneesh Krishna (2003). Kids in a Kirana Store: Building a Case for Retail Communication. Journal of Brand Management, 10(3), 219-232.
Sinha, Piyush Kumar & Dwarika Prasad Uniyal (2003). Segmenting Shoppers on Their Behaviour. Paper presented at the 10th EIRASS Conference, Portland, Oregon, USA, August.
Sinha, Piyush Kumar & Dwarika Prasad Uniyal (2004). Using Observation Research for Segmenting Shoppers. Journal of Retailing and Consumer Research, (Forthcoming).
Sorensen, Herb (2003). The Science of Shopping. Marketing Research, Fall, 30-35.
Underhill, Paco (1999). Why We Buy: The Science of Shopping. New York: Simon & Schuster.
Nikhilesh DHOLAKIA is Professor in the Marketing, E-Commerce, and Management Information Systems areas in the College of Business Administration at the University of Rhode Island. His research deals with problems at the intersection of technology, globalization, and consumer culture.
Dr. Nik Dholakia, Professor
Marketing, E-Commerce & Information Systems Areas
College of Business Administration
7 Lippitt Road
University of Rhode Island
Kingston, RI 02881 (USA)
Piyush Kumar SINHA is Associate Professor in Marketing at the Indian Institute of Management, Ahmedabad, India. His research covers retailing, consumer behavior and media.
Dr. Piyush Kumar Sinha
Indian Institute of Management
Ahmedabad 380-015 (India)
Dholakia, Nikhilesh & Sinha, Piyush Kumar (2004). Observations on Observation in India's Dynamic Urban Markets [38 paragraphs].Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 6(1), Art. 13, http://nbn-resolving.de/urn:nbn:de:0114-fqs0501131.