A Process Study of the Development of Capabilities for Replication in an Evolving Franchise Chain

The purpose of this paper is to extend insight about processes and routines needed for franchise replication and makes an important contribution to understanding ways through which dynamic capabilities are created. Based on a grounds-up study, this paper utilizes a storyline approach (Miles & Huberman, 1994) to present interview data obtained from interviews with elite informants (IEIs) of the host company and its archival data. Evolution of capabilities, both substantive and dynamic, are captured in the findings section. Of the seventeen interviews, seven subjects were location heads and ten corporate executives. Interviews with corporate executives included questions pertaining to their functional specialty as well. Finally, implications for future research and practice are discussed. Business and Economic Research ISSN 2162-486


Introduction
The success of a retailer in building a chain, is primarily due to replication since it allows firms to grow geographically and become a chain of even tens of thousands (Bradach, 1998), such as McDonald"s, Starbucks, Walmart, Kroger, CVS, etc. Replication in chains results when firms decide to exploit a retail business model, which proved successful in its initial settings, by copying it in many other locations (Szulanski & Jensen, 2008). More formally, replication in chains "entails the creation and operation of a large number of similar outlets that deliver a product or service (Winter & Szulanski, 2001, p. 730)." More than 60 industries, ranging from a cookie-cutter format frequently seen in the fast-food restaurant industry to a highly variable interior decoration service, have benefited from replication. Furthermore, interviews with elite informants (IEIs) present unique opportunities to explore the micro-foundations of firms" strategic plans, as well as how visionary elements translate into daily execution through organizational processes, policies, and actions (Aguinis & Molina-Azorin, 2015;Felin, Foss, & Ployhart, 2015;Foss & Pederson, 2016). Although profound to the success of chain franchises, the importance of replication is taken for granted in literature, barring a few exceptions. Presumably, once an entrepreneur, especially a restaurateur, finds an innovation that "clicks" very well in its initial settings, it can be replicated successfully. It is assumed that the differentiation initially created (Michael, 1999) --through either informal, visionary "gut-feeling" or very deliberately --would propel a quick spurt of growth (Michael, 2003) once the entrepreneur is able to find the human, physical and financial resources (Garg, 2013a). Attention frequently turns to the study of the adequacy of the resources, the problems in acquiring and managing the resources, and the interaction of those resources. For example, franchising literature has used multiple theories such as resource scarcity (Castrogiovanni, Combs, & Justis, 2006), resource augmentation (Garg, 2005a), agency (Garg & Rasheed, 2003;Lafontaine & Slade, 1997;Shane 1996;Michael, 1999), transaction cost economics (Mathewson & Winter,1985;Rubin, 1978), intra-channel relationships (Baucus, Baucus, & Human, 1996;Grunhagen & Dorsch, 2003), etc., to study those problems (Combs, Michael, & Castrogiovanni, 2004;Combs, Ketchen, Shook, & Short, 2011).
Somewhere in that typical and very useful approach-as the expanding literature on chains and franchising proves-an important set of questions arise: Is replication really that simple? And, what complexities have to be confronted in replication? Answering those two questions is the general research agenda of this study. Winter and Szulanski (2001) drew attention to the fact that despite being a prominent mechanism for growth in retail industries, the phenomenon of replication has not been studied in great detail by scholars interested in chain organizations. While Winter and Szulanski have contributed a body of work to build theory concerning replication in chains (see references for that body of literature), other research remains scarce (Gupta, Hoopes, & Knott, 2015). The result is that important questions regarding capabilities-defined as "how resources are used in combination" (Eisenhardt & Martin, 2000, p. 1107)-in building chains through replication have not even been identified, let alone Economic Research ISSN 2162-4860 2021, Vol. 11, No. 2 answered.
The current paper addresses this important paucity in literature. It describes processes used, issues faced, and solutions implemented by an Indian firm as it began to develop a chain of high-end restaurants in India. That firm was thus the site of a quasi-experiment dealing with replication. A grounded-theory approach, appropriate when literature in a field is in infancy (Glaser & Strauss, 1967), was chosen. This approach would allow the use of an inductive method, such that in-depth information of one firm would be studied to enhance a theory of capabilities for replication. Researchers have used an inductive approach to study related topics in the context of a single firm. For example, Szulanski and Jensen (2006) studied presumptive adaptation and transfer of knowledge, and Gupta et al. (2015) studied redesigning of routines for replication. This field study also reveals many variables, shown in Table 3, which can be used in constructing a more comprehensive model. Such a model would incorporate compelling mediators for examining the effect of replication capabilities on performance, thereby addressing recently noted shortcomings in empirical research on dynamic capabilities (Schilke, Hu & Helfat, 2018).
When a firm runs a business in an "equilibrium" stage using stationary processes, it exercises zero-level capabilities without which the firm can "not collect the revenue from its customers that allows it to buy more inputs and do the whole thing over again (Winter, 2003, p. 992)." But disagreement exists among scholars regarding hierarchies of capabilities (Srekovic, 2018). A simpler term-substantive capability-is used in this paper instead of zero-level capabilities, following Zahra, Sapienza, and Davidson (2006). Accordingly, a dynamic capability is "a higher order capability that allows reconfiguration of substantive capabilities (Zahra et al., 2006: 917)." For F&B service business specifically, "the capabilities that support the creation of new outlets by McDonald"s or Starbucks are prototypical examples of dynamic capabilities, focused on the domain of scale and geographic markets rather than product attributes (Winter, 2003: 992)." Whereas growing business in the same location requires substantive capabilities such as developing promotional campaigns, offering daily menu specials, etc., dynamic capabilities needed for replication include processes and routines "used by managers to copy, transfer, and recombine resources within the firm (Eisenhardt & Martin, 2000, p. 1107." But where does the insight about processes and routines needed for replication come from, and how? Understanding those mechanisms is the general purpose of this field study. The specific research questions that this paper attends are: How do capabilities needed in replication evolve? Do executives have those capabilities before they begin replicating a business model? Do they even recognize the importance of capabilities they don't have? By exploring these questions, this study makes an important contribution to understanding the processes through which dynamic capabilities are created. Understanding how dynamic capabilities are "built" has been central to research on firm evolution, growth, and survival (Ambrosini, Bowman, & Collier, 2009;Makadok, 2001).
Domains of product development and manufacturing have dominated this field (see Eisenhardt andMartin, 2000, &Zahra et al., 2006, for reviews of many such studies). This Economic Research ISSN 2162-4860 2021, Vol. 11, No. 2 study adds an important piece to a growing body of literature on dynamic capability (e.g., Drnevich & Kriauciunas, 2011) by exploring underlying processes in two important but scarcely studied contexts: chains, and emerging economies. The appropriateness of the restaurant industry in the context of chains has been noted severally (e.g., Bradach, 1998;Dant & Nasr, 1998;Garg, 2005b;2018;Kalnins & Mayer, 2004;Michael, 1999;Sorenson & Sorensen, 2001) and that context, thus, still remains modal. Emerging economies are at the forefront of growth of chains (Welsh, Alon, & Falbe, 2006), and in India, chains are gaining popularity (Franchising Focus, 2008). India offers the large size advantage that has been shown to attract great interest in chains in Brazil (Gomes, Carneiro & Dib, 2018).
Based on a grounds-up study, this article is organized somewhat differently from a typical roadmap. This introduction is followed by methods. Next, a storyline approach (Miles & Huberman, 1994) is used to present data obtained from interviews of executives of the host company and its archival data. Evolution of capabilities, both substantive and dynamic, is then captured in the findings section, which is divided into two phases. A discussion section follows next, where first findings are interpreted in the light of a multifaceted perspective of dynamic capabilities for replication. Finally, implications for research and practice are discussed.

Methods
Methods followed suggestions by Yin (1989) and were used by Szulanski and Jensen (2006), involving Mail Boxes Etc. This study began with perusal of many key documents that were made available by the host company (THC henceforth, to protect the anonymity of the firm). These included: a background note on THC and a written corporate profile of THC; brand history of the umbrella brand (will be identified as UB, henceforth, to protect the anonymity of the umbrella brand); corporate brochures and listings of outlets; profiles of the management team and a schematic of the organization structure; menus from several outlets of the business format/brand of focus. pictures of the layout of the restaurant and display of food; and brochures, "concept note,"and press releases of key special events to promote the focal brand.
To maintain anonymity, the fictitious name of Mé lange will be used to refer to the company and its brand. In order to gauge the preparedness of the company to franchise Mé lange, the following artifacts were examined: introductory material for prospective franchisees, a sample of a site approval proposal, a manual for franchisees, a new store opening manual, an employee"s handbook, and a brochure of the only franchisee of the focal brand. After becoming familiar with the archival material listed above, researchers interviewed the management team in the corporate office using a semi-structured approach. Sixteen respondents were targeted (excluding the managing director), based on the profiles of the management team and the organization structure. These were mostly one of two types: (a) head of a department, so that a variety of functional views could be captured, or (b) manager of an outlet, so that experience at the outlet level could be netted. Notes of the interviews were taken, but transcripts could not be made since many respondents had reservations. Locations, menus, layouts, atmospherics, kitchen operations, etc., were observed at the locations and notes were taken. The interviews were open ended to capture the ground-up spirit of this research, but the following questions formed the anchors of the interviews:

Q1. Compared to the Connaught Place unit (the referent unit) of Mé lange, how much different is the franchised unit Mé lange (the focal unit)?
Respondents were asked to rate the overall difference on a scale of 1 (least) to 7 (most).

Q2.
Give salient examples to explain the level of these differences.
Respondents were probed to think of many items, such as menus, clientele, size and layouts, atmospherics, location, etc., to answer this question in some detail. They were then asked to rate the focal unit, in terms of differences relative to the referent unit.
Of the seventeen interviews, seven subjects were location heads and ten corporate executives. Interviews with corporate executives included questions pertaining to their functional specialty as well. The few executives who were with THC almost since its inception were particularly a good source of nostalgic, qualitative details that otherwise couldn"t have been captured, since there was very little documented organizational memory of that nature.
All of the information gathered from the above sources was used to construct a storyline of the replication experience at Mé lange. The storyline approach permits readers access to the qualitative data itself and to evaluate conclusions of the authors (Miles & Huberman, 1994). It is crucial to an inductive composition style of presenting a qualitative process study which must provide a close-knit link between the empirical data and the model/framework that emerges from that data (Berends & Deken, 2019). This approach has been used in other studies of replication (e.g., Gupta et al., 2015;Szulanski & Jensen, 2006).

The Setting of the Empirical Case
To begin the field study, the managing director (MD) was interviewed. He identified Mé lange (started in 1978) as the format to explore in this study. He also mentioned other formats the company owned, which actually had four types: an ice cream parlor (started in 1978), a pastry shop (started in 1975), a quick service family style restaurant (started in 1980; henceforth QSR), and an express outlet (started in 2006) suited mostly for petrol pumps and airports. All four would be targeted for growth, depending on the opportunity.
The MD defined brand positioning of the two key formats as follows: Mé lange: He defined it as casual dining around a salad bar, international cuisine, and an alcoholic beverage bar. It aims to attract tourists and professionals in class A and A1 cities in India. The pricing would be positioned to be seen as one of higher value for less money by the target customers, consistent with their higher socio-economic status. There were seven Mé lange outlets at the time of the study.
QSR: It offers a wide choice of menu (a large variety of traditional Indian cuisine, in addition to burgers, pizza, ice cream, etc.) served in a quick service format. It aims to attract families and young adults in a large number of cities in India. Evidently, the average ticket amount would be much lower than that for the Mé lange tickets. There were 53 QSR outlets at the Economic Research ISSN 2162-4860 2021, Vol. 11, No. 2 time of the study.
The MD identified the following Key Success Factors (KSF, henceforth; defined as criteria that customers most frequently use in their purchase decisions, Garg, 2013a).
Mé lange: Location was the most important KSF. Most promising locations would be near high-end hotels (to attract foreign tourists), central business districts (to attract professionals for lunch), high-end shopping malls (to attract trendy demographics), and posh residential areas (to target higher socio-economic strata, mostly for dinner). Menu innovation was another KSF, given a very crowded industry which offered too many choices to the target customer segments. Western-style cuisine featuring a salad bar, sizzlers, and desserts were identified as the unique hallmarks of this brand.
Outstanding service and evident hygiene were the other two KSFs the MD identified.
QSR: While location was the most important KSF for this format too, it was characterized in terms of high traffic. Thus, QSRs are typically located in trade areas (where delivery options are popular), near schools and colleges (to attract the young demographics), and mid-level malls (to attract impulsive traffic). Attention would be paid to the presence of heavily trafficked residential areas around the location as well. Menu innovation was a KSF for QSR, too, with examples such as seasonally special fruit flavors for ice creams and milkshakes, bio-light (sugar free) options, Nutrybyte (vegetarian protein) burgers, combos involving tie-up with Coke, and gift hampers in Diwali season (known by foreigners as "the festival of lights" and is the most popular, elaborate festival in most parts of India, especially northern cities that constituted THC"s current geographic reach).
Promotion was a third KSF identified for this format, which focused on deals that will make the price-related value more attractive to the target customer segments.
The MD was asked to explain the role of franchising in growth of these chains. He said that to establish the desired brand positioning, it is risky to franchise in major metros such as Delhi, Bombay, Kolkata, and Bangalore. But once a company-owned unit is established in a metro, they would be looking to franchise in surrounding cities. Franchising was considered very important to the future growth of both formats. Single-unit franchising was desirable to begin with, due to high investment needed and control desired by THC. But if a potential franchisee insists on acquiring a state as a territory, such requests would be considered. Lastly, the MD described the expected profile of a franchisee for Mé lange. He indicated that many criteria will be considered. Companies that can invest as per the business plan for the state over time will be seen favorably. Past experience in the restaurant industry is desirable, but not necessary. More important is a successful track-record in developing and running a business. In fact, even more crucial was "a fit between their and our business philosophy and vision". ISSN 2162-4860 2021

The Evolution of the Mé lange Chain
The current chain has its roots in a group of businesses founded by two entrepreneurial brothers. More than 80 years ago, the two brothers started a variety of ventures in the lodging, food and beverage services industry. Of particular relevance were their path-breaking Chinese (1950( ), Hungarian (1960, and an all-silver Indian thali (plate) restaurant ventures. One of the founders had studied at Cornell University. A few current top executives had joined the original business as early as 1970s, and they traced the origins of the current marquee outlet in an earlier, close-by location which was later vacated. The founders" descendants had only recently (in 2006) sold the business to an investor group, which still had close family ties with those descendants.

The First Unit
From the beginning, the Mé lange brand was primarily targeted to foreigners, especially Europeans. The first unit was established in 1978 in a building where the founders had already set-up a modern, upscale hotel in 1958. Not surprisingly, the site was located in the national capital"s central commercial and retail area, Connaught Place-Connaught Circus. The area was built during the British rule of India, named after the Duke of Connaught, and was apparently modeled after the very famous Royal Crescent building in the city of Bath in England (Hindustan Times, 02/08/2011).
A secondary target clientele for that outlet comprised aristocratic Indians who perhaps, like the founders, had already traveled abroad and thus had significant exposure to European cuisine. Even rulers of many Indian princely states had their local homes in the nearby areas around King"s way (modern day Rajpath, which now houses foreign embassies and consulates), and came to Connaught Place for shopping, movies, and eating out at trendy places of that era (Wikipedia: 2015a). For example, Wengers, owned by a Swiss couple, lured the royalty, British officers, and Europe-traveled businessmen living in Delhi by offering pastries, tarts, mousses and homemade Swiss chocolates (Hindustan Times, 02/08/2011).
A European-style salad bar, along with soups, breads, and desserts, was evidently the central piece of the menu at the first Mé lange (henceforth, unit #1). This menu, unique among Delhi upscale restaurants of the era (even "cole slaw made in vinegar & pineapple sauce and dill pickle stood out," (personal interviews, 2008), was offered as an equally unique unlimited buffet. The collection of items such as lasagna, ravioli, half-pounder lamb burgers, and "an all-American breakfast (four pancakes, omelet, sausage, bacon, hash-browns, croissant, cinnamon rolls, and fresh fruits served in bowls made out of Pine shells), served all day," was unparalleled throughout entire India. The unique menu was matched by a premium-price and very clean ambience. Despite the passage of three decades, the demeanor of the veteran executive--as he effortlessly recalled the details of Mé lange unit #1-clearly indicated that pride was an important result of their innovation.
No wonder the restaurant was "jam packed during the lunch hours" (personal interviews, 2008), when pizzas, hash browns, French bread with soup, etc., were added to the menu. ISSN 2162-4860 2021 Indian cuisine (offered a la carte) was added later, presumably to broaden the appeal to an evolving domestic customer base that was neither necessarily up-scale nor trendy. In keeping with the extravaganza, even the Indian cuisine had a lot of variety. For example, it included a stand-out "tandoori (clay oven) sampler that actually came out with not only chicken entré e but also naan (bread), garden salad, and in addition, a small bowl of daal (lintel preparation)." In effect, the sampler was a full meal combo at very attractive prices for those looking for more than a bite during the lunch hours.

Expansion Phase 1: 1987-2006
As Delhi grew and spawned a very fast growth of surrounding areas where land was much cheaper, the founding entrepreneurs saw an opportunity to grow their business through both geographic expansion and addition of formats, often offering different types of food and refreshments such as coffee, ice-cream, pastries, and hot eclectic food served quickly. Presented below is a brief chronological account that exclusively focuses on the development of Mé lange, the chain of interest in this paper.
Unit #2 was opened in NOIDA, a newly developed, planned city in the national capital region in 1987. NOIDA was a very ambitious government project initiated in 1976 to become a huge metropolis comprising planned sectors and zones for modern industry, retail centers, financial services, office and residential towers, educational institutions, hospitals, transportation hubs, etc. Situated about 25 kilometers south-east of the first, unit #2 was located in the same building where THC had recently opened a hotel, one of the earliest in that area. The hotel carried the umbrella brand name of the company, UB. The whole building was built by THC on a big plot of land, at a low cost, since it was bought right when the metropolis land was first sold for development. Production facilities and a QSR unit were added incrementally in that complex. Production facilities now include a food processing unit, a cheese plant, an ice cream factory, a bakery, and an engineering workshop. Besides guest rooms, the hotel included banquet and meeting facilities. Mé lange unit#2 had a seamless entrance from the hotel reception area. The kitchen was built behind the reception area, and was in fact established to supply a QRS unit located on the opposite side of the reception area.
The lunch buffet offered at this unit did not include the salad bar but was rather a "small fair of typical Indian cuisine, including a few of both vegetarian and non-vegetarian dishes" and the clientele "lacked sophistication" (personal interviews, 2008). It consisted mainly of traders, office people, property dealers (since there was a boom of construction of houses and offices), and businessmen that were house-guests at UB hotel; there were "no kitty parties of ladies". It didn"t have a license to serve liquor, although a bar was opened within the adjoining hotel soon after.
Unit #3 was established in an eastern suburb of Delhi, in a commercial complex that was close to newly developed, middle to upper class neighborhoods, in 1996. This unit makes a triangle with the first two, and is less than 15 kilometers from either. The unit was located in a commercial suite, which THC had bought early, again at a low price, in a shopping-cum-retail center that Delhi government had constructed and sold to the public. Hundreds of such projects were coming up all over and around Delhi, as planned Economic Research ISSN 2162-4860 2021, Vol. 11, No. 2 development had started taking roots in the national capital region and elsewhere in this still young country; barely 50 years had passed since India had liberated from the British rule.
This unit shared many characteristics with unit #2: no salad buffet, unsophisticated clientele, lower prices since "the customers had little reason to pay a higher price for a menu similar to QSR in the same premises" (personal interview, 2008), etc. In fact, a lot of items came from QSR Kitchen. Since there was no adjoining hotel, there were no house guests either. Hence, revenues were lower than unit #2, especially during the evenings, once nearby offices and shops closed for the day.
Unit #4 was opened in 2002 in Panipat. That city lies about a hundred kilometers away from Connaught Place, going north. The city is both small and traditional. The unit is situated within the premises of a hotel in a busy part, through which passes a highway connecting Delhi to a few northern states. The hotel is managed by the parent company, and carries the umbrella brand name UB.
Its ambience was very much like unit #2, but it had more space. Unlike unit #2 and 3, it hosted many family events (including match-making meetings between families of the prospective bride and groom) and parties (birthdays, marriage anniversaries, etc.). Often, the general manager had to customize its menu to attract such sponsors in order to "survive" (personal interviews, 2008). The customized food items for those events were of a very traditional North-Indian nature, far removed from the Mé lange"s trendy menu used at unit #1. Traditional Indian food and deserts were usually demanded by the party hosts. Therefore, the unit even employed/contracted traditional Indian chefs called halwais. Their skills were in cooking, for example, deserts such as the time-tested and highly popular Gulab Jamun, rather than the modern pastry.

Expansion Phase 2: 2006-2008
Meanwhile, the original unit needed to be moved since its lease was expiring and the space was proving grossly inadequate. The top management of the company had been mulling expansion and refreshing the brand for a few years anyways. They used the move of the location to re-launched Mé lange, with a great amount of deliberation and planning. Treated as a new "avatar" of Mé lange, the unit #5 was opened in 2007, on the second floor of a building in Connaught Place within a kilometer of the site of the first unit. The company"s archival documents describe this "marquee" unit as follows: "A medley of some of the finest cuisines from around the world in a fresh ambience makes the new Mé lange unique… The use of wood, checkerboard floors, white brick walls, green glass, fresh bamboo, and natural light have created a look of casual elegance. The interiors are complemented by the new logo, reflecting the new spirit of the restaurant." Unit #s 6 and 7 were both opened in Gurgaon, a city about 35 kilometers west of Connaught Place, in 2008 (Wikipedia, 2015b). It has seen huge development of multistory, modern buildings and townships that have surpassed all newly developed metropolises anywhere else in India. Close to the biggest and most prestigious international airport in the country on the ISSN 2162-4860 2021 one hand and the busiest national highway on the other, Gurgaon is the first choice for multinational corporations and large Indian corporations alike. Many of them built either a new regional office or relocated headquarters there. In turn, Gurgaon has also attracted modern and upscale residential, commercial, retail, and entertainment developments. In tune with those emerging trends, the company opened these two units in two separate malls, one of which was considered the most popular, spanking mall in northern India. The two units were barely 3 kilometers away from each other. Those two units were modeled very much after the marquee unit #5, and therefore, it will be henceforth termed the referent unit for this study.

Business and Economic Research
Unit #6 was very similar to #5 in terms of menu, hostess desk area, an all-glass front, bright green colors, decorations, and other elements of ambience, staff, kitchen, and above all, the centrality of the salad bar. But, one particular element was different: a food court was nearby. Another very obvious difference was that unit #6 was located in a suite which had a large side as an outside wall, and THC had capitalized on that aspect by using glass lavishly on that side. This permitted a highway-side view from tables along that wall.
Unit #7 presented no such view, however. Also, although the "cream of cream" lived within a distance of two kilometers or so, its mall was not really as chic as that of unit #6 and therefore, not as popular. Further, the location within the mall was not as prominent either. Indeed, it seemed to be tucked in a corner far from the entrance, with no crowd-pulling attraction (such as cinema halls or coveted international fashion brands that unit #6 benefitted from) in the immediate vicinity. Further, the suite presented rather awkward space, with a small front and a sectional seating arrangement. As a consequence, the salad buffet was neither visible nor readily accessible from a majority of the tables.
In 2008, unit #8 was established too. Like the other two that were opened in the same year, it reflected the trend of shopping and entertainment moving to modern malls throughout the country. It was located in Jaipur. Hardly any foreign tourist misses-out on the "pink city", which constitutes the third leg of the most popular tourist itinerary in India-the golden triangle.
Jaipur is the capital, largest city in Rajasthan state. Having a rich history of a princely state pre-independence of India, it is internationally known for the architecture of its fort, palaces, and other old buildings. Tourists love its bazaars, whose merchandise showcases Rajasthan"s resplendent culture, and whose streets still illustrate the pattern of everyday commerce and entertainment signatures of the nostalgic past (Jaipur, 2015). The Rajasthan government has been consistently focusing on tourism as a main source of revenue and thus, has invested a lot in promoting it to both foreigners and nationals. Many movies, including some under the British banners, have been shot there (Wikipedia, 2015b). Just like any other class-A city in the very populous India, it is bursting at its seams as construction projects spring up all over and around it. Residential townships, industry, higher education, retail, hospitals, etc., are all parts of the modern development here, much in the footsteps of NOIDA and Gurgaon. Also, a metro transportation project linking the far-flung, under construction, modern satellite suburbs to the core of the city would begin soon. ISSN 2162-4860 2021 With so much going for it, a franchisee decided to locate the Mé lange unit #8 in a new mall in Jaipur. Being a very successful franchisee of a few other restaurant brands, including Subway, and having developed many outlets for those other brands already, he very well understood the importance of a top-flight mall to the success of an upscale restaurant. He had bought space very early when the construction project of the mall was launched. At the time, he had planned to open a unit of another company"s brand, but later, thought Mé lange to be more suitable for that location. The changes in his plans were still timely; unit #8 opened just a few months after unit #5 and coincided with grand opening events of the mall as well.

Business and Economic Research
While it was very much designed after unit #5, a distinguishing aspect of unit #8 was that it was the first, and the only, franchised unit of Mé lange. It is about 270 kilometers west of Connaught Place, making it the farthest from the flagship unit #5 and production facilities of Mé lange. The mall enjoyed the most prestigious status in Jaipur, and it was situated in a premium, central location of the rapidly expanding city. The unit was built in a spacious suite of the mall. It also featured a large all-glass front, and was situated in a prominent part of the mall that also carried a few other trendy casual dining restaurants.

Data
In addition to the above narrative, Table 1 summarizes some important data such as the date of establishment, capacity, space available, and estimated rent of each unit in the chain. Table  2 shows a compilation of the differences, relative to the referent unit #5, as perceived by the respondents regarding each of the seven units. Differences on a wide range of attributes are captured and explained in the findings section below. Note 1: Unit #1, which was located in Connaught Place, is excluded here since it was closed and made way for a new unit nearby. The brand was relaunched with the opening of that unit, unit #5, which is the referent unit. ISSN 2162-4860 2021  Note 1: Unit #1, which was located in Connaught Place, is excluded here since it was closed and made way for a new unit nearby. The brand was relaunched with the opening of that unit, unit #5, which is the referent unit.

Business and Economic Research
Note 2: A negative sign meant the attribute was rated less desirable relative to unit #5 and a + meant it was rated more desirable.

Findings: Evolution of Capabilities for Replication
A key finding based on the storyline data presented above is that in the process of growing the business, THC developed capabilities in two phases. These phases were separated by the advent of a new top management team, which brought in a different set of perspectives than the previous team. These differences led to a reconfiguration of resources and hiring/creating new resources as well. Since THC had previously been operating the Mé lange concept in an equilibrium stage using stationary processes for many years, it was in effect, exercising substantive capabilities. When THC embarked on a replication strategy, it was exercising at least one higher order capability. Thus, in each phase, a transition from substantive capability to a dynamic capability can be discerned, as explained next.

Phase 1 Substantive Capability: The Capability to Create Differentiation
Entrepreneurs "create, define, discover, and exploit opportunities" (Zahra et al., 2006, p. 917). The founding of unit #1 was truly remarkable in that in those days, there was no other restaurant in perhaps the entire nation that could boast of a European-style salad bar, soups, breads, and dessert. The menu featured items such as lasagna, ravioli, half-pounder lamb ISSN 2162-4860 2021 burgers, and "an all-American breakfast (four pancakes, omelet, sausage, bacon, hash-browns, croissant, cinnamon rolls, and fresh fruits served in bowls made out of pine shells), served all day" (personal interviews, 2008). Not only this menu was unique, but it was also offered as an unlimited buffet, something the consumers had not been offered before. Other elements of marketing mix such as very clean ambience and premium pricing supported the differentiated product.

Business and Economic Research
The ability to differentiate was a complex capability, as it arose from a multitude of interacting resources. The foreign travel experience of the founders was an important one. They also had uncommon entrepreneurial foresight that there would be growing interest in western food and buffets, and that many patrons would be looking to pay a premium for a clean and chic ambience. In addition, the combination of the restaurant and the hotel created a synergistic interface of two complementary products that aided differentiation. There were few trendy, premium facilities anywhere in India, Thus, UB was a brand for affluent tourists and the local elite of Delhi.

Phase 1 Dynamic Capability: The Capability to Adapt a Retailed Service to a Location
Exploiting opportunities to purchase low-cost real estate and establishing a hotel there led to the founding entrepreneurs" building the restaurant at the site of unit #2. But from the start, they deemed the most unique feature of Mé lange, the western-style buffet comprising salad bar, soups, breads, and dessert, to be unsuitable for the expected demographics of the customers of that restaurant. The chosen sector of NOIDA was not as upscale and lacked the prospects of being a favorite of upwardly mobile professionals and other modernity seeking demographics. The location of unit #3 suffered from very similar drawbacks right from the beginning, so it didn"t have those attributes either.
Executives of THC in those days seemed to be more opportunistic in accumulating location resources than they were brand image conscious. Either they lacked insights or, they were unconcerned about the impact any elements of adaption would have on Mé lange"s brand personality. Thus, they changed the menu drastically and dropped the marquee buffet altogether, to suit the clientele of unit #2 and #3. This is a clear example of presumptive adaptation, to fit a location, so far removed from the original business model (Szulanski and Jensen, 2006) that observers would fail to see the continuity in the brand from one to another location. Data about the perceived difference in menu of unit #2 and #3 relative to the referent unit menu (Table 2), which had retained the marquee buffet from unit #1, provides evidence for this interpretation. When we consider another data that the per sq. ft. rent was one third or less for these units than the premium paid by Mé lange in Connaught Place, the difference in spaciousness would indicate a noticeable lack of upscale-ness of the facilities, but facilities are an important cue for brand positioning (Garg, 2017). The neglect of such obvious brand cues that are crucial to a restaurant business suggests that the adaptation seemed to overlook its impact on the brand image. However, the ability to adapt is a dynamic capability nevertheless, since to say that only successful capabilities are, would be tautological . ISSN 2162-4860 2021 Winter and Szulanski (2001, p. 731) have noted: "the formula or business model, far from being a quantum of information that is revealed in a flash, is typically a complex set of interdependent routines that is discovered, adjusted, and fine-tuned by "doing." Similar processes led to Mé lange"s developing unit #5 from a scratch. It portrayed the template of western style salad and dessert buffet served in a warm and bright ambience, as the most distinguishing feature.

Phase 2 Substantive Capability: The Capability to Forge a Template
Many executives of THC regularly dined at the Mé lange restaurant of unit #2, because the same building has their offices. Thus, they were automatically reminded, perhaps sadly, of how different this unit is from the marquee unit #5, to which they kept returning in order to observe implementation of their decisions and customer behavior and for their enjoyment as well. In turn, their commitment to making the subsequent units very similar to unit #5 would be expected to get reinforced very frequently. They would probably discern the fine details and routines of the characteristics of unit #5 that were driving its success. That interaction and introspection is the learning process, which would drive the opportunity to replicate the format embedded in unit #5 more rigorously in subsequent units, because gradually the tacit knowledge about that format will become more apparent (Darr, Argote, & Epple, 1995;Szulanski, 2006). Table 2 suggests that learning from the past experiences was captured by the THC executives in their recent replication projects. Taken together, the units (#6, 7, and 8) that were designed and launched recently were far more similar to the referent unit on a wide range of characteristics than were the older units. Crucial characteristics on which the difference was obvious included menu, interior design, quality of staff, and operations. Thus, in phase 2, the executives had consciously focused on forging and replicating a template. However, not all recent units were as uniformly replicated with reference to unit #5. Unit #7 was less similar to the referent unit than was unit #6, and unit #8 was the least similar to the template of the three recent units.

Phase 2 Dynamic Capability: The Capability to Sustain Brand Image
It took Mé lange almost 20 years to decide to abandon the practice of allowing major adaptations to the product to suit location-specific customers and other attributes. The firm had to wait until the arrival of a new management team, when THC was bought out in 2006, for clarity in articulating what the brand stands for, how to translate it into a business format, and what will be the resulting business model. The new top management team rejuvenated the brand by creating the theme of the restaurant as the western style salad and dessert buffet served in a warm and bright ambience. They decided to replicate essentially the entire business model in newer units, whose locations were now chosen according to suitability to the product attributes, rather than to seek low costs. The firm developed deliberate routines (e.g., using standardized forms to capture attributes of prospective locations) that went into selection of markets to enter (e.g., Gurgaon) and specific locations (e.g., malls) within a market, to ensure that there was a precise fit between the chosen location and the espoused brand image. ISSN 2162-4860 2021 Evidence for the sharp focus on brand image can be gleaned from Table 1 as well as Table 2. Unit #s 6, 7, and 8 were all in upscale locations and modern malls. The seating capacity and customer space/person were very similar and far higher than unit #2. With few exceptions, they were not far behind the referent unit in terms of menu, location attractiveness, interior design, quality of staff, operations, and marketing promotions.

Summary of Findings
Putting the two stages together, the brand image creation capability-in this case, approximate uniformity of attributes of products and locations from which to deliver those product attributes-occurred very slowly at THC and only after a change in the top management team. It seemed the firm had finally moved away from ad-hoc adaptation and toward delivering unified experience, regardless of which outlet the customer visited. Thus, consistent replication of an upscale restaurant brand integrated with suitable locations seemed well within the grasp of THC; i.e., the firm had eventually accumulated dynamic capability for replication. But the discipline of the top management in enforcing the template rigorously was still not evident, as was seen from some clear deviations that were permitted in units #7 and #8.

Discussion
Understanding how dynamic capabilities are "built" has been central to research on firm evolution, growth, and survival (Ambrosini, Bowman, and Collier, 2009;Makadok, 2001). This paper has shown that the starting point for the evolution of dynamic capabilities is executive choice. In particular, path dependency of dynamic capabilities (Dierickx and Cool, 1989) emerges as the naturally occurring by-product of founders" vision and growth strategies for their enterprise, followed by vision and strategies of executives that succeed the founders. Succession turns out to be critical, since new executives are not overly burdened by the past legacy and can more freely reconfigure resources and routines.

Performance of Replication
Information gathered from the website of THC and physical verification in 2015 showed that units #7 and #8 no longer existed, and units #2, 3, and 4 continued in their original format, but were struggling. Only Unit #5 and #6 were apparently doing good business.
How do we explain this pattern of performance in relation to replication capabilities? Some executives of the original top management team realized that the initial attempts at replication had some glaring shortcomings. For example, according to a veteran senior executive, unit #2 should have been located in a different sector in NOIDA, and the location of unit #3 had a very small percentage of the target audience. But apparently, those executives did not have much voice in front of the founders, to whom they owed an entire career. Surprisingly, the top management team of 2006 also failed to shake-off the burden of legacy. That inability likely has its roots in the fact that the CEO of that team was still a family member. Was it inertia or nostalgia, that prevented exploring other strategic alternatives to maintaining the status quo? Information to explore this question is not available in this study. ISSN 2162-4860 2021 The new executives made marked improvement in crafting a template and thus exhibited a crucial substantive capability for replication, but still fell short on the dynamic capability of selecting appropriate markets and specific locations within those markets that would fit well with the espoused personality of the brand. The eventual closures of unit #7 and unit #8 were the result of inadequacy of that higher order capability. This would be labeled as a somewhat inadequate but not completely absent capability, as the sustained success of unit #5 and #6 would indicate. The premium, highly differentiated Mé lange brand was very suitably grafted in those locations, as they offered a large potential market. The market served was a very healthy combination of all the three bases of segmentation: geographic, demographic, and psychographic (Kotler, 1973): geographic, in relation to the proximity of international airport, the busiest part of a national highway, and the destination mall; demographic, in terms of the standing of new Gurgaon as the city of the affluent, educated, and white-collar professionals; and psychographic because that elite class likes to be seen modern, adventurous, and health conscious. They feel happy entertaining in an exclusive social setting and are ready to pay higher prices for that amalgam of non-monetary values overall (Garg, 2017b).

Theoretical contribution: A Multifaceted Perspective of Replication Capabilities
A cogent set of important conclusions can be drawn from this study. Successful replication in a retail setting requires both: (a) the initial, substantive capabilities in creating a differentiated brand, and (b) the subsequent dynamic capabilities in forging a standardized template. However, these two are not sufficient. More challenging are other, higher order capabilities: (c) Positioning the differentiated brand in a fitting location in a sufficiently munificent market and (d) Exercising discipline to say no to a location, no matter how economical or readily it is available, when the market offered lacks that fit or is munificence. Why is that cogent set of capabilities difficult to acquire? Winter and Szulanski (2001, p. 730) have noted: "Replicators create value by discovering and refining a business model, by choosing the necessary components to replicate that model in suitable geographical locations, by developing capabilities to routinize knowledge transfer, and by maintaining the model in operation once it has been replicated." The above passage provides a starting point for constructing, based on the storyline, findings, and discussion of results presented in this paper, a multifaceted perspective (summarized in Table 3) of replication capabilities. Executive choice is central to dynamic capabilities, defined as "the abilities to reconfigure a firm"s resources and routines in the manner envisaged by the firm's principal decision makers (Zahra et al., 2006: 917;emphasis mine)." However, even though dynamic capabilities "reside", in part, with top executives (Teece, 2014;Williamson, 2016), all executives do not view the environment of the firm identically. If executives have very different backgrounds, they are likely to perceive the environment differently too (Garg, Walters, and Priem, 2003). Their choices are a consequence of deliberate implementation of strategic intent as well as discovery-through experience with the implementation process-of what works and what does not. New executives can more easily detach from the past legacy compared to those who created it in the first place. Thus, the later executives can more dispassionately embark upon reconfiguration of resources and routines and forge a template that must be replicated without much alteration in new locations.
The experience of foreign travel and entrepreneurial enthusiasm, combined with persistent exploration and hard work, were key to the creation of a unique concept that the founders of THC christened Mé lange. The concept was highly successful, indicating that THC had achieved differentiating capabilities. Only a moderately dynamic environment may provide the false sense that early successes are due mainly to the founders" and their loyal employees" exceptional set of human capital, on which they can rely on in new markets as well. Aware that the environmental dynamism is on the rise, the founders of UB envisioned an explosion of growth for lodging and food services and were inclined to capitalize quickly on this foresight by grabbing cheap real estate and building their businesses at those locations. Executives also differ in their entrepreneurial style, which in turn has a very strong bearing on other facets. THC founders" intuitive decision making prevented adequate analysis of the suitability, and their concerns for costs fostered hastiness. They trusted their family brand name, UB, would draw crowds in the new locations. They did not mind that the unique brand name they created for unit#1, Mé lange, actually stood for a very different concept than the restaurants they ended up creating in units #2, 3, and 4. Each of those restaurants was a major deviation from the original concept, in order to suit their respective clientele, thus demonstrating adaptive capabilities. Although founding executives demonstrated two important capabilities, differentiating and adapting, they lacked a deep understanding of how these interacted.
The discovery of what routines should be emphasized when a business tries to grow geographically is a very challenging task, because of the causal ambiguity. New locations are seldom identical to the site of initial success, and to leave out some parts of the core routines in order to make presumably appropriate adaptation is tempting but can lead to failure (Szulanski and Jensen, 2006). Therefore, it is much safer to try to replicate the entire business model, since the all-important hidden routines are embedded in the model itself.
The advent of a new executive team changed most of the characteristics of Phase 1. They were professionals and unencumbered by loyalty to the founders, so they could engage in more rational decision-making. They recognized the strength of the differentiation inherent in the original format, so they studied it carefully and culled out its unique features. Learning from the outcomes of trial and error of the Phase 1 executives, executives of Phase 2 came to understand that if Mé lange had to grow by exploiting a truly differentiated overall offering (Porter, 1980;Porter, 1985), the unique western-style salad and dessert buffet embedded in its entire ambience would need to be replicated. They hired advertising and architecture experts to help them define the target brand image and embody it in an improvised format that they implemented afresh in unit #5. That unit, in which the refreshed CRDB brand was personified, was very successful. This demonstrated that THC had achieved the capability to develop a template. Further, in keeping with the macro changes in the economic, demographic, and socio-cultural sectors of the environment, they decided that all new locations would be developed in modern malls, primarily because of the pull factors (Calvo-Porral & Lé vy-Mangí n, 2018) and opportunity for sustain expansion (Watson IV, Worm, Palmatier, & Ganesan, 2015). As a result, they replicated the template in three units quickly. The implementation reflects the dynamic capability of integrating the template and location, thus enhancing the brand positioning relative to the market.

Implications for Research
The evolutionary economics framework of Nelson and Winter (1982) was originally explained in the context of large organizations. We have tried to extend it to an early stage replicator, who had only recently shown significant commitment to a core business format. Thus, the current study addresses the first two elements of the "theory of replication strategy" proposed by Winter and Szulanski (2001: 730-31): challenges in discovering and refining a business model, and difficulties in choosing the necessary components to replicate that model in suitable (emphasis mine) geographical locations. Future research can focus on the study of the remaining two elements in detail, namely, challenges in developing capabilities to routinize knowledge transfer, and in maintaining the model in operation once it has been replicated, especially in hyper-growth environment such as that faced in China and India, where GDP growth has clocked more than 7% for many years. That type of environment poses some unique challenges for replication.
First, in hyper growth environments executives will likely find abundant opportunities to change jobs. High turnover will pose difficulties in the transfer of tacit knowledge to new units, because "tacit knowledge will walk away" with the departed executives. Second, hyper-growth brings in new entrants that may either imitate or disrupt a successful business model. Competition of either type may generate pressure on the incumbent to adapt, thus accentuating the hazards of adapting presumptuously once again.
This study also suggests an important avenue for the theoretical development of general literature on dynamic capabilities. The centrality of executive choice to dynamic capabilities has been recognized (Teece, 2014;Williamson, 2016;Zahra et al., 2006), but how executive succession shapes the development of dynamic capabilities is yet to be adequately explored. Given that dynamic capabilities are path dependent (Dierickx & Cool, 1989), who succeeds as CEO will have a salutary effect on how a firm"s resources will be reconfigured. For example, a new CEO who was involved in a lot of acquisitions in previous positions might look for resource and capability infusions from outside more than another who hasn"t had much experience with acquisitions. In turn, greater issues of integration may arise. Thus, future research can explore the role of executive succession in determining relative importance accorded by a firm to different types of capabilities, such as operational, coordinative, and strategic capabilities (Srekovic, 2018). For example, attracting strategic investments is a capability that has been researched rarely in franchising literature (Hsu, Kaufmann, & Srinivasan, 2017). While keeping business within family is preferred by many owners (Patel, Kim, Devaraj & Li, 2017), according a high priority to executive capabilities might lead to going public so that ownership incentives can attract more capable CEO.
This study has the usual limitations of mostly qualitative field studies. Although useful numerical data and simple quantitative patterns have been shown, the analysis remains theoretical rather than statistical. Also, the subjectivity of the respondents as well as the researcher can be much more pronounced in a field study than in a large sample research. It is conducted in the context of a single organization and a single country, so generalizability may be limited.
Despite limitations, field studies like this provide necessary complement to large sample, empirical studies by identifying important variables like those identified above. Without field studies, empirical studies will suffer from inadequate research models in the first place. Phenomena will not be understood in sufficient details necessary to select variables on which data should be collected, nor will there be adequate grounds for postulating specific relationships among variables. Schilke et. al. (2018) recently noted that empirical research on dynamic capabilities suffers seriously from two weaknesses: underspecified models and relatedly, lack of work on identifying promising moderating variables. This field study addresses those two shortcomings by revealing many variables, shown in Table 3, which can be used in constructing a more comprehensive model incorporating compelling mediators for examining the effect of replication capabilities on performance.
If performance of a particular outlet is unsatisfactory, it need not be the fault of the template; the fault may lie in the inappropriate location for that template, or its faithful execution. Both of these shortcomings in capabilities seem to have caused the shutting-down of Units #7 and 8. Longitudinal, large sample studies of unsatisfactory performance at the outlet levels are now warranted to test the evaluation processes of replication, strategic adaptations, and financial performance following those adaptations.

Implications for Practice
An important implication of this study is to inform new practitioners of replication, such as prospective franchisors, that adaptation may involve a high degree of hazard. They should not presume that they have to adapt the template to local conditions, which they think are so different from those faced by the referent unit that the original template will not work. If they made that presumption (Aspara et al., 2010) and changed some core elements of the template, the adapted model may miss out not only on some key elements of differentiation, but their interconnections too. The deliberately left out elements and interconnections, as well as their interconnections with the remaining elements, may not have been properly understood but may be essential (Szulanski & Jensen, 2006) to providing a unified, distinct customer experience.
Replacing the marquee western-style salad buffet with traditional Indian-style buffet and entries may have been liked by many customers of a new location, but the adaptation took away the essence of a differentiated position of the Mé lange brand. Customers who liked the experience at Unit #1 and then chanced upon units #2, 3, or 4 would be disappointed. More so, if their main motivation to eat at those other units was the impression Unit #1 had left on them. That impression, when transferred on to those other units of the same brand due to psychological association, would be weakened significantly once key elements were found missing. Uniformity across outlets is a key feature that separates successful from unsuccessful chains (Bradach, 1998;Garg et al., 2005b), because its absence compromises investments in creating differentiated identity of the brand (Garg, 2013b).
Another important takeaway from this study for practitioners engaged in replication is to understand the role of organizational memory. Although Mé lange executives have routines for meeting and discussing all aspects of business, a focus on creating organizational memory for learning was not evident. Nascent chain executives, especially those who have a high degree of action orientation at the expense of analysis orientation, can easily get preoccupied with growth and operational activities. This is even more likely in the hyper environments described above. But they must still emphasize documentation of knowledge gained from experiences, so that it becomes more accessible to those who are deeply involved with establishing new units and operating them. Documenting both successful and unsuccessful adaptations are necessary, because both types of experiences accrue learning (Newbert, 2005) by doing (Eisenhardt & Martin, 2000) and can prevent repetition of mistakes in replication. Otherwise, those mistakes will obviously result in losing money and time. A nascent chain can ill-afford to lose either of these critical resources. Equally importantly, crucial errors in replication will cause difficult to repair damage to the differentiated brand that the firm is trying hard to establish, maintain, and exploit (Garg, 2013b).