Developing a branded house strategy for place brand management

Purpose – City branding has gained popularity as governance strategy. However, the academic underpinning is still poor, and city branding needs a more critical conceptualization, as well as more complex management systems. This paper challenges the use of a “one size fits all” city brand, which is still common practice in many places. The paper proposes that city branding involves much more complexity than is commonly thought and outlines a strategy that enables urban policy-makers, marketing researchers and (place) marketers alike to better deal with city branding. Design/methodology/approach – The authors integrate insights from literature on place branding, brand architecture and customer-focusedmarketing. Findings – The article argues that place brands (in general and communicated place brands in particular) are by definition very complex, due to their different target groups, diverse place offerings and various associations place customers could have. Thus, an advanced brand management including target groupspecific sub-brands is needed. Practical implications – The model will be helpful for place brand managers dealing with a diverse target audience, and is likely to improve the target group-specific communication. Originality/value – The paper provides an insight into the complexity of city brands and acknowledges that the perception of city brands can differ considerably among different target groups. Additionally, it offers a more comprehensive definition of place brands. This will be helpful for city brand managers and researchers alike in dealing with city brand complexity.


Introduction
Cities increasingly adopt business strategies in general and marketing techniques in particular (Bennet and Savani, 2003), trying to establish the city as a brand, and promote their place to its different target groups (Braun, 2012). By these means, marketing and particularly branding have become part of the range of contemporary urban policies (Ward, 2000). In fact, place branding has become one of the best-known applications of branding in the public sector (Klijn et al., 2012). Unfortunately, city marketers often believe that the city brand is controllable and fully manageable, and that they can deploy a "one size fits all" city brand for their city brand communication. Yet, a city brand as such is based on the perceptions of the different customer groups (Merrilees et al., 2012;Zenker, 2011), and these perceptions of a city can differ strongly, given the various target groups' different perspectives and interests (e.g. between residents and tourists, Zenker and Beckmann, 2013). Consequently, this creates a big challenge for city brand communication, as different associations have to be communicated to fit the customers' expectations, making city brand communication a very complex practice. Unfortunately, the academic underpinning is still poor, and city branding needs a more critical conceptualization (Lucarelli and Brorström, 2013). Therefore, this paper presents: a more comprehensive definition of a place brand; and develops a branding/communication strategythe City Branded House Strategyto deal with the complexity of city branding, to advance the current academic debate and to help city marketers and other urban policy-makers alike to manage city brands more effectively.

Place marketing and branding
History and status quo Even after more than 30 years, place marketing can be regarded as a relatively young field of academic research (Ashworth and Voogd, 1990;Kotler et al., 1993). O'Leary and Iredale (1976, p. 156) identified place marketing as a challenging field for the future, describing it as activities "designed to create favorable dispositions and behavior toward geographic locations". Place marketing could involve different types of places and different spatial scales (Ashworth and Voogd, 1990). In practice, it could be applied to neighborhoods, districts, tourist destinations, cities, rural areas, regions and countries. Metaphorically, place marketing is the family tree, with family members such as city marketing, destination marketing and location marketing acting as the branches (Braun, 2012). The first scholarly publications really dedicated to place marketing came from regional economists and geographers. Several of these focused solely on the promotional aspects of marketing, examples being Burgess (1982) and Gold and Ward (1994). Some urban policy studies, dealing with structural change in cities, focused on the use of marketing to reconstruct the city's image. Short et al. (1993) examined the efforts of older industrial cities seeking to replace their industrial image by redeveloping their places and become more attractive to investors. Likewise, Paddison (1993) looked at city marketing in the context of economic restructuring and urban regeneration. Van den Berg and Braun (1999) argued that the application of place marketing was increasing due to intensifying competition between cities caused by globalization and the information technology revolution. If, in early publications, the adoption of marketing strategies to the urban context was labeled as novel and unique, today we find it as a much more common practice in the public sector (Klijn et al., 2012). Despite increasing interest in place marketing, Ashworth and Voogd (1990) as a notable exceptionwere one of the few who proposed a planning framework for place marketing, highlighting the consumer side (next to the producer side) which should be segmented according to their needs and wants.
Questioning a "one size fits all" city brand From the very beginning, the use of marketing (and later branding) also met criticism. For example, Philo and Kearns (1993) argued that it only serves the interests of the urban elite and that it is part of an economic agenda to attract capital and stimulate urban (re-) development. Evans (2003, p. 417) is critical of the combination of culture and city branding contending that branding of cities through cultural flagships and festivals stimulates "a form of Karaoke architecture". Bennett and Savani (2003, p. 81) pointed to the legitimization problem when new brand identities were "determined in a top-down manner by local government authorities" and existing residents had little involvement. However, the use of marketing and branding by cities has been growing steadily and fits in the wider trend of introducing commercial practices and private sector management styles in urban governance (Klijn et al., 2012).
Place marketing gained additional attention from researchers and practitioners as a result of books by Kotler et al. (1993Kotler et al. ( , 1999 on Marketing Places. These three prescriptive textbooks were important for the recognition of place marketing; the academic impact, however, should not be overstated. They underplay the complexity of places and their customers, indeed, Braun (2012) argues that these approaches underestimate the urban planning and policy context. Although the academic attention is growing (Lucarelli and Berg, 2011), it is not commensurate with the growth in place marketing practice. A possible explanation for this phenomenon could be the nature of place marketing itself. After all, place marketing deals with numerous diverging target groups, complex and related products, as well as different political settings in which marketing decisions are made (Van den Berg and Braun, 1999;Ward, 2000). For example, other family members of the place marketing family, especially those with a single focus on tourism marketing, have received much more attention. Kerr (2006, p. 276) argues that this strong focus can be a problem as "destination marketing and the destination brand have the limitation of being tourism focused only and often preclude major stakeholders in a location".
In the past decade, the focus on place marketing has shifted toward place branding (Gertner, 2011;Lucarelli and Berg, 2011), with places trying to garner positive associations in the place consumer's mind. In public administration especially, the influence of stakeholder involvement (Klijn et al., 2012) and the different obstacles implementing place branding activities (Eshuis et al., 2013) were discussed. Researchers from other fields tried to measure and conceptualize place brands (Kavaratzis and Hatch, 2013;Zenker, 2011), or discuss the definition of place in place branding (Boisen et al., 2011;Kavaratzis and Kalandides, 2015;Warnaby and Medway, 2013). In addition, the first meta-analyses of place branding literature by Lucarelli and Berg (2011), as well as Gertner (2011) showed exponential growth in contributions, but also the lack of theoretical and conceptual frameworks in the field.
In the past five years, initial attempts have been made to increase place brand management. Hanna and Rowley (2015), for instance, discuss in their Place Brand Web conceptual model that places do consist of many different sub-brands. These place-related brands are often owned or communicated by important place stakeholders (such as companies that are closely related to the place or iconic places within the place), and place brand management should try to orchestrate this network of brands for a powerful joint place brand. This makes stakeholder management the most important task of place brand management. Other researchers focused especially on one specific stakeholder groupthe important role of citizens  and their participation in the branding of places Kavaratzis and Kalandides, 2015). Zenker and Erfgen (2014) discussed a place brand management model suggesting a management structure, where stakeholders (i.e. mostly residents, but other groups could be possible) can apply for funding and support.
To do so, they must show that their idea or initiative has value for the place brand (i.e. impact) and a high brand-fit. Besides these two management strategies, most other articles describe how to measure or conceptualize brands while staying vague with the question of what to do with them strategically. Therefore, this paper tries to offer a strategic approach in this regard while first discussing the brand definition in use.
The brand and the city Defining marketing concepts is not easy, as most of the time, there are no shared definitions but multiple viewpoints. In place branding, it has been highly criticized by some researchers that the brand definition according to the American Marketing Association (2015, no page) as a "name, term, design, symbol, or any other feature that identifies one seller's good or service as distinct from those of other sellers" is too narrow for use in the field of complex structures like places (Ashworth and Kavaratzis, 2009;Govers, 2013). Others have proposed that the essential feature of a brand is "nothing more and nothing less than the good name of something that's on offer to the public" (Anholt and Hildreth, 2005, p. 164). Brands are markers for offerings, ideally creating easier choices for customers by promising (a certain level of) quality, reducing risk and building trust between the consumer and the brand (Keller and Lehmann, 2006). According to Keller (1993;, this "good name" or reputation exists as a network of associations in the minds of the consumers as so-called brand knowledge. The knowledge about a brand is thereby built through brand awareness (the degree customers are aware of all features of a brand) and their brand image (defined as perceptions about a brand as reflected by the brand associations held in customer memory). In that way, customers evaluate those associations building a so-called overall brand attitude (Keller, 1993). Zenker and Braun (2010) transferred this definition of a brand as a network of associations to place brands in their highly cited conference paper.
In summary, it seems that the idea of place brands as place associations seems to have become a standard definition over the past years (Boisen et al., 2011;Kavaratzis, 2004;Zenker, 2011). However, some researchers seem to overemphasize the simple counting of associations while missing the network idea in this definition. As Kalandides (2015, p. 1371) state: "the problem is that associations cannot be 'added up' in a sum because they are not static [. . .] they are in a constant flux as they interact with each other". Thus, it is important to notice that the associations are influencing each other, leading to further associations and building a mental network in the mind of the place consumer. For instance, Zenker and Beckmann (2013) highlight that place brands consist of connected chains of associations that build a joint network, whereas the associations differ a lot between target groups. This also shows that place brands are highly subjective and rely on the relationship between the place and the perceiving subject (Kavaratzis and Kalandides, 2015;Zenker and Beckmann, 2013).
Another central aspect that is missed in current place brand definitions is the variety of importance of brand associations (Ajzen and Fishbein, 1980). While some associations may be dominant, they can show different relevance for buying, moving or visiting behaviormaking this aspect also highly situationally related (Schnittka et al., 2012). For measuring place brands, the importance of an association may not be relevant on first sight, as a place brand consists of all interconnected associations. However, from a managerial level, this information is crucial, as the management of brands (in terms of communicating and changing place brands) need some kind of hierarchy of those associations for an effective selection of communication topics. Here, it is sometimes not the associations with the most awareness but those which are strongly interconnected with other associations or touching a more relevant category (Schnittka et al., 2012). For instance, people may associate "Lübecker Questioning a "one size fits all" city brand Marzipan" (a well-known German sweet) with the city of Lübeck, Germany. The chances of potential residents moving to (or tourists visiting) that city for this reason however is very small, as you can buy this sweet even in New Zealand. The "historic city centre", on the other hand, is a well-known association for the place and shows relevance for residents and tourists alike for their residential/visiting decisionmaking those two associations, from a managerial perspective, different in their importance.
In summary, it becomes obvious that different customers will have different brand knowledge and a different brand attitude. A middle-class income family could evaluate the city of Munich as posh and too expensive, whereas someone else sees it as prestigious and desirable. Furthermore, brand knowledge (such as an association like "good job opportunities") can show different importance for, for example, a residential (high) and a visiting decision (low). To satisfy those varying needs and wants, customers are divided into different target groups, building more homogenous segments. This market segmentation (Smith, 1956) can be done by different variables like income, interest, geography, sociodemographics or lifestyle (Gonzalez and Bello, 2002).
Moving on from the marketers' comfort zone, it is a challenge for researchers from all disciplines to "translate" those contemporary branding insights to the context of places, as the special characteristics of places have to be taken into account. A first point is the great variety of place customers and their diverse needs and wants. The main target groups in place marketing and place branding are residents, businesses and visitors (Kotler et al., 1993). The latter, for example, seek activities like shopping malls or cultural offerings, whereas investors may be more interested in business opportunities. It is inevitable that there are conflicts and synergies between the needs and wants of those different groups, and therefore, communication of the place brand should be developed with these factors in mind. These main target groups can be further segmented, for example residents as internal residents and external potential new residents. Within these segments, smaller homogenous groups can be found, such as students or talents (Braun, 2008). Given the differences in needs and wants and different brand knowledge, it is quite obvious that those groups differ in their perceptions of a place (Zenker and Beckmann, 2013). As Lynch (1960) explains, we receive various signals from places through buildings, public spaces, streets, people and personal experiences. While some of these signals are planned as communication (e.g. landmark buildings) and can be to some extent controlled by the city, most of them are unintended and out of the power of city officials. However, all these factors communicate something about the place and contribute to associations formed in the minds of the target audiences, varying across target groups. The variety of communication of places leads to a dissimilarity in the way we perceive places in comparison to commercial brands (Kavaratzis, 2008), especially as the number of potential different communicators and stakeholder is mostly much higher than in the corporate context. Second, places are complex products. One's location cannot be seen separately from other locations within the placehence the place offers not a single location but a package of locations (Braun, 2008;Boisen et al., 2011). City customers choose their parts of the city for use and perceive the place as such through these location samples. Consequently, the product for tourists in a city, for instance, partly overlaps with the product for the city's residents, but may also have distinct locations solely used by one target group. This package of locations can vary in complexity by how many different locations are included, the haziness of each location (e.g. a poor and wealthy area at the same time) or the level of how organized the complexity is (Zenker and Petersen, 2014).
Third, the place is central to place branding and not the (brand of) responsible (territorial) governments, and these governments are not the exclusive brand owners. In practice, various public and private stakeholders are actively involved in place marketing and branding (Van den Berg and Braun, 1999;Braun, 2012). Furthermore, it is important to notice that places are "social constructions made up of a selection of spatial elements by humans" (Boisen et al., 2011, p. 137) and therefore, their geographical boundaries are more complex than administrational borders.
Finally, place branding cannot be separated from its political governance context. In the context of urban and regional policy, it is the subject of political decision-making and therefore deals with regional or municipal organization(s) and policy-making procedures (Braun, 2012). This setting cannot be compared to regular business practice, and any application of branding concepts needs to take these characteristics into account.
Dealing with place brand complexity Defining place brands and brand complexity Complexity can be defined as "a set of elements involving relationships among them" (Rapoport and Hawkes, 1970, p. 108). In addition, Zenker and Petersen (2014) focus on place complexity as "a quantitative measure and [. . .] a construct with different qualities", dividing complexity into "quantity" (i.e. by how many different elements are included) and its "qualities". These qualities are "ambiguity" (having, e.g., poor and wealthy residents at the same district) or the "degrees of entropy" (i.e. the level of how organized the complexity is).
As general (product) branding is often a process of reduction on core associations (Anholt, 2009;Keller, 2003), we often see complexity as a problem in practice and academia alike. Complexity may be understood as an unclear brand concept, meaning that the brand is fuzzy and incoherently positioned. Brand managers fear that consumers will become overloaded by information that is too complex (Zenker et al., 2017). Bettman et al. (1998) show, for example, that in cases of high product complexity, customers apply simpler heuristics and selective information processing, even though this reduces the effectiveness of their decisions. With a simple brand, consumers with little information can more accurately identify the brand, but they may also show more dissatisfaction with the brand and a higher need for more information (Scammon, 1977). More importantly, Zenker et al. (2017) show that for residents and place brands, the complexity of the brand positively influences the identification and commitment with the brand, whereas for tourists, the simpler brand had the same impact.
Thus, given the unique characteristics of places and place brands, a more enhanced definition of place brands is needed, that: better acknowledges place brand complexity; focuses on the differentiation between the general place brand and a communicated place brand; and delivers a stronger conceptualization of the process of building place brands.
As previously explained, it is not the brand, in terms of a name or symbol to identify and to differentiate, but brand knowledge that underpins our definition. On this basis, and by using the definition from Zenker and Braun (2010), we redefine the place brand in general as: A place brand is a network of associations in the place consumers' mind based on the visual, verbal, and behavioral expression of a place and its' stakeholders. These associations differ in their influence within the network and in importance for the place consumers' attitude and behavior.
Questioning a "one size fits all" city brand Place brand communication To influence those networks of associations, places try to communicate with their customers. In his place brand communication model, Kavaratzis (2008) included three types of place brand communication: first, the primary communication, including the architecture, urban design, infrastructure, museums and other real place offerings, but also the places' behavior from residents and other stakeholder groups. Second, the secondary communication, which includes the formal and intended communication through all forms of advertising, public relations, graphic design and the use of logos and slogans (promotion). Finally, the tertiary communication, which refers to word-of-mouth reinforced by the wide variety of city users and media, which is usually beyond the control of place marketers. Braun et al. (2014) adopted this content, putting the focus even more on the communicational aspect and empirically testing three types of Place Brand Communication: (1) The physical place: These are the "real" place offerings that could be of value for the target groups: on the one hand, the buildings, infrastructure, the design of the public space and the amenities (e.g. shops, theaters, museums, restaurants, sports facilities, schools and universities). This type of communication is mostly visual (physical buildings) and is mostly influenced by the city's politicians, urban planners and urban policy-makers, property owners and private developers. City branding professionals do communicate about these changes (intended communication), but they have little or no influence on the investment decisions in the physical place. On the other hand, it also includes the behavior of the local people and other relevant experiences with the place, including a behavioral component that is out of the control of city officials (unintended communication).
(2) Traditional: This concerns all forms of advertising or public relations via official channels. This is usually undertaken by communication professionals working in the city administration or in specialized marketing organizations operating at arm's length of the city administration. It is therefore mostly visual (e.g. logos, advertisement) and verbal communication (slogans) or a combination. (3) The place word-of-mouth: This "informal talk" between place stakeholders (between the different groups, e.g. between a tourist and a resident, or within the group) concerning the place is reinforced foremost by the media, the residents and visitors. This type of communication is mostly verbal, behavioral and increasingly visual (e.g. sharing pictures on social media platforms). As with the physical place communication, we could differentiate between intended and unintended communication here. As Braun et al. (2014) show, it is one of the strongest communication tools but also hard for place marketers to influence.

The City Branded House Strategy
First, we differentiate between the place brand in general (network of all associations in the mind of the consumer) and the communicated place brand (referring to those associations with high importance). While in the former the place and all its aspects are included (including negative or less relevant aspects), the latter is the part of the place brand used in the official communication (what is often misunderstood as the place brand and mostly includes as narrowed version of positive and influential aspects) (Figure 1). This differentiation is so important because it is not a logo or slogan, nor the (official) place brand communication, but overall place perceptions (in terms of brand knowledge) of the target group(s) that lead to so-called brand effects (attitude and behavior). For instance, researchers empirically showed that a more positive evaluation of a place (brand) leads to higher general satisfaction , identification with the place (Fleury-Bahi et al., 2008;Zenker et al., 2017), place attachment (Hernández et al., 2007) and results in an intention to stay at a place (Zenker and Gollan, 2010) (Figure 2).
However, for the place brand management, we try to influence this (general) place brand knowledge with our place brand communicationfocusing for instance on a few positive aspects that show importance for our target audiences. As mentioned before, this influence and the evaluation of city brand aspects can differ between target groups because of different brand knowledge levels of the target groups and the different demands for a place. Research shows that these differences are most striking between internal and external target groups. Zenker and Beckmann (2013) show that the city perceptions of an external target group are much more stereotypical, whereas the internal target group's city brand perception is more diverse and heterogeneous.
A "one size fits all" place brand communication for both target groups would disregard the complexity of a city and probably failespecially for target audiences with a profound place knowledge (Zenker et al., 2017). A stronger customer focus is needed with a differentiated brand architecture to match a specific target group with a specific city subbrand. This customer-focused view is not common in the public sector (Buurma, 2001) nor in city marketing practice. However, city marketers and other urban policy-makers could find parallels in the marketing and branding of diversified companies and learn how to deal with the complexity of the multiple target groups through their practice.
Obviously, brand management deals with the managerial part of brandingthus focusing on the place brand communication, keeping the general place brand in mind and the communicated brand in alignment. For a more complex brand management model, brand architecture (Aaker, 2004) is often used, showing hierarchical structures of brands (in the corporate context) with different strategies for multiple target groups. With the branded house approach, a brand architecture is built with related and more focused sub-brands for each product that are marked within the corporate umbrella brand (Petromilli et al., 2002). The aim is to build a strong overall company umbrella brand with the help of the target group-specific product sub-brands (e.g. "Nivea" as umbrella brand and "Nivea for men" as target group-specific sub-brand focus). This approach is not limited to product and company brands. It could also be extended to product or company brands that include a place brand (Uggla, 2006), or fully to the place branding context (Iversen and Hem, 2008;Kotler and Gertner, 2002). In contrast to our colleagues, we are not using the brand architecture in the  "one size fits all" city brand context of an umbrella country brand and regional or city sub-brands. Cities should develop a brand structure with a city umbrella brand and target group sub-brands as suggested by Kerr (2006). Very much like the modern organizational structures of marketing departments in companies (Homburg et al., 2000), the marketing structure of cities should be organized by their target groups (Braun, 2008). We call this strategy the City Branded House Strategy, including a branded house approach with target group-specific sub-brands and a city umbrella brand that is represented by the shared overall city brand perception by all target groups. This strategy has been developed with cities in mind, but it is also applicable to different spatial scales, such as districts, regions and even countries (Kotler and Gertner, 2002).
For our City Branded House Strategy (Figure 3), a city must be aware of the general city brand (network of associations in the mind of all place consumers) and the specific place perception of the specific target audiences (including importance of aspects). For this strategy, the focus of the official place brand communication should be on target groupspecific sub-brands (e.g. for tourists, residents, businesses and investors). Those target group-specific sub-brand foci can better concentrate on the particular needs and wants of the different city customerstaking the importance of different aspects of the city for different target groups into account. This leads to a more differentiated perception of the city by the target audiences, whereas some aspects will be equally important and dominant. These parts could be seen as a shared city umbrella brand (with shared associations between all different target groups with high importance for all target audiences).
Central in our strategy are traditional place brand communications as well as other means of city communication (i.e. physical place communication and place word-of-mouth), as these jointly build perceptions about the city brand in the mind of the city customers. While the traditional place (brand) communication can be (mostly) controlled by the city marketers and urban policy-makers, it is important to keep in mind that they could also show unintended communication effects, as a city's resident will, for example, also be affected to some extent by communication intended for tourists (so-called waste circulation). Additionally, sometimes, city customers fulfil different roles at the same time (e.g. to be a resident and an investor), and multiple sub-brand foci could be addressed to one individual. This is a challenge for managing the City Branded House Strategy, as the sub-brands cannot be seen as independent brands, but also may influence each other through the waste Figure 3. The city branded house strategy circulation and the multiple roles of target audiences. This is particularly challenging if the content of the communication differs or is even contradictoryan issue we will discuss later in the paper.
The physical place brand communication also influences our perception and can only be to some extent controlled by city marketers and urban policy-makers (as product development is most of the time not influenced by place marketers). Buildings that are considered flagship buildings (Butler, 2007), or redevelopment projects (Paddison, 1993) can, for instance, communicate an equal and supportive message for a high-culture tourist brandand thus, place brand communication should react to physical changes. However, it should be noted that the intended perception of any change is not controllable. For example, residents could perceive investment in the redevelopment of their district as a form of gentrification, rather than as an improvement (Butler, 2007). Having made this argument here, it also shows the vital importance of involving the urban development department within city branding attempts, as it is one of the most powerful communication tools in the urban policy context . In addition, the behavior of local people (Freire, 2009) is part of this communication, which is even harder to influence through city officials.
Finally, the place word-of-mouth is influencing the city perceptiona process that cannot be controlled by city officials, even though city marketers and urban policy-makers intensively try, for instance, to influence the place's media image. In particular, contact within the target groups and with other target groups shapes our perception about a city, highlighting for instance the important role of residents in this process Eshuis et al., 2014) and making increasing positive place word-of-mouth one of the most important brand tasks of the place brand communication. On a managerial level, stakeholder management and creating incentives for stimulating positive place word-ofmouththat shows a fit to the place sub-brandsare crucial here (Zenker and Erfgen, 2014).
To further explain the main idea of the City Branded House Strategy, we will now use the City of Amsterdam, the capital of The Netherlands, as a hypothetical example. The city is relatively small compared to other European capitals such as London and Paris, with 821,752 residents (CBS Statistics Netherlands, 2016). Despite this, Amsterdam is a popular place to live and an internationally recognized business and tourist location. The city has welcomed more than 54,000 new residents in the past five years. Amsterdam is the 7th most popular tourist destination in Europe (Euromonitor International, 2016), and 41 per cent of the incoming investment into the Netherlands goes to the Amsterdam region (KPMG, 2015).
A key element in our City Branded House Strategy is to acknowledge that there are target group sub-brand perceptions, as we define brands as perceived by the (potential) place customers. For example, a young "backpacker" from the United Kingdom could see Amsterdam as a bohemian place and associate the city with "sex, drugs and rock n roll" (Kavaratzis and Ashworth, 2007). The city sub-brand perceptions of Amsterdam residents (internal target group) are different from those of visitors, as residents are usually better informed about the physical place and are concerned with very different city brand associations such as safety and livability. Amsterdam business owners (internal target group) develop a city sub-brand perception including, for instance, the business climate, or the local government attitudes toward business, as befits their business interests. These different city sub-brands are not independent; interactions between the city sub-brand perceptions are possible. There could be conflicts. For example, the deputy mayor of Amsterdam expects that more foreign businesses will locate in the Dutch capital because of BREXIT (Britain's proposed exit from the European Union): "Amsterdam has a good proposition for companies that are looking for Questioning a "one size fits all" city brand an establishment in Europe" (Het Parool, 2016, no page). The city's sub-brand perception as a business location could be negatively affected by the "sex, drugs and rock n roll" image that appeals to particular tourists. In another case, there could be synergies if this "bohemian reputation and character" of Amsterdam is valued positively by the many advertisement agencies located in the city. The challenge is to create synergies and avoid conflicting city sub-brand communication. The City Branded House Strategy makes this challenge explicit and encourages city marketers to manage it.
A strong influencing variable of the city perception is the physical place brand communication. This implies that city brand management also means developing the place to fulfil customers' demands (this goes far beyond the decision powers of city marketers and by definition links politics, urban planning and city marketing) and in a second step, to communicate a genuine picture of the place (Braun, 2012). Improving the physical place is not sufficient in itself. First, the message about the improvement(s) needs to reach the target groups; second, the target group has to be interested in learning more about the improved place, and third, the changes made to the place have to be seen as having made a real improvement.
Essentially, our strategy aims to influence the city (sub-)brand perceptions. The city subbrand communication and the city umbrella brand communication are intended to do just that. In 2003, Amsterdam launched the (communicated) city brand: I amsterdam. The brand is built up around three core valuescreativity, mercantilism and innovationand is developed for four main target groups: residents, visitors, companies and opinion makers (Kavaratzis and Ashworth, 2007)[2]. For a City Branded House Strategy, this I amsterdam brand is the communicated city umbrella brand (shared values in the place brand communication by all target groups). The ambition of the city of Amsterdam is that the I amsterdam brand features in the brand communication to specified target groupsor rather, in the terminology of our strategy; the city sub-brands communications refer and relate to the city umbrella brand communication. By building different target group subbrands, the needs and wants of those groups would be more effectively addressed. Again, there could be synergetic effects as well as conflicts caused by the city umbrella brand communication. Amsterdam's umbrella brand really strengthens the city sub-brand perception of Amsterdam tourists. For instance, the real life I Amsterdam logo in front of Amsterdam's most well-known museum (the Rijksmuseum) has become a tourist attraction in its own right. At the same time, there are residents that associate I Amsterdam with excessive tourism and the overcrowded historic inner city. Past year, these residents published an obituary of I Amsterdam (Het Parool, 2015) arguing that it has been too successful and that it creates a tourist mono cultureshowing a potential conflict here.
An advantage of working with target group sub-brands would be the already established organizational structure in city governments (e.g. the separation of the tourism office and business development). By using this strategy, branding policies can be more effective through established specialized agentsdefining the task of the general place brand manager to a cooperative unit that is in charge of balancing and improving the sub-brand communication of the specific units by aligning them to the general city brand (network of associations) and handling conflicts and creating synergies between those units.

Discussion and implications for city branding
The City Branded House Strategy reveals several challenges: For physical place brand communication, it has to be noted that the real physical place cannot be controlled by the city marketing professionals but is the domain of urban planners (i.e. urban development departments) and other urban policy-makers (e.g. mayors and aldermen), or the place stakeholders themselves. Those professionals have a different scope, and the city brand perception is usually not a priority for them. The challenge for city marketers is to raise awareness that the decisions of planners and other urban policy-makers also affect the perceptions of (potential) place customers and promote the idea that city branding considerations should be part of planning decisions. Another group with a strong influence is property owners and private developers. This group is concerned with the perception of their property and its surroundings (the owners) or with the perception of their development projects (developers) but do not necessarily take an interest in the overall city brand. In particular cases, city marketers could explain the impact of their behavior on the city brand. City marketers should also aim to influence the decisions of developers when it is relevant for the city brand. For both, the urban professionals as well as the property owners and developers, this is easier said than done; the number of city marketers is usually small, urban planners, and city marketers do not necessarily think alike, and the influence on private owners and developers is limited.
A second challenge is to motivate the city's stakeholders to support the city sub-brand strategy (within the city government, e.g. aligning tourism and business development departments and outside, i.e. between other city stakeholders) and make them understand that they jointly influence the city brand perception as such. The launch of the I Amsterdam brand was met with criticism from residents and the local media. At least two counter branding initiatives, "Ai Amsterdam" and "Amsterdamned", were developed after the launch to express the disapproval of the new Amsterdam brand. Organizing wider support is crucial to increase the chances of success, but full support is very unlikely, as there will always be opposition from stakeholders that do not feel represented by the brand or dislike the use of marketing and branding altogether.
For the third challenge, we identified the potential overlap and influence of the subbrands with each other. By definition, the sub-brands cannot be seen as independent from each other and may influence each other through the waste circulation and the multiple roles of target audiences. While nonconflicting sub-brand parts are no problem, it is particularly challenging if the content of the communication differs too much or is even contradictory. This could create confusion or mistrust with the target audiences and lower identification with the place brand. Here, it is the task of the place brand manager to carefully monitor and mediate in case of conflict. However, no other current management model (e.g. a "one-sizefits-all" brand) is solving that issue, and in our opinion, the benefits of our strategy prevail over the potential risk.
This leads to the final challenge of the City Branded House Strategy, concerning the coordination, monitoring and safeguarding of the communication between the sub-brand units. Even if the brand structure is implemented, brand management needs ongoing coordination of partners and changes. As city branding deals with numerous stakeholders, the sub-brands and the potential joint city umbrella brand communication will be constantly questioned and influenced by them. Changes in the sub-brands will affect the city brand perceptions, and therefore need to be monitored with care.

Practical issues
Fortunately, most of the planners and other policy makers are aware of the complex urban environment in which they work. Our City Branded House Strategy shows that this complexity should also be acknowledged in the branding strategies deployed by cities. If brandingas a process of reductiontries to simplify messages to the consumers (Anholt, 2009), it may not be as useful as one would think in such a case of high complexity. It could even be that communication which is too simple and merely focuses on an external target Questioning a "one size fits all" city brand group can prove very dangerous for the support of city branding among residents, and could also reinforce the idea among planners and other urban policy-makers that city branding is merely the "selling" of oversimplified and often stereotypical images (Zenker et al., 2017). By using our strategy, marketers, planners and urban policy-makers will have a better understanding of place branding and its complexity. Dealing with complex city brands calls for governance strategies moderating the different interest groups (sub-brands) and mitigating against conflicts of interest.
However, what is the optimal number of sub-brands. Is there a limit? Cities serve a wide range of target groups that can be divided into numerous market segments, making coordinated customer management very difficultif not impossible. In some cases, it could be wise to deploy separate city sub-brands when a lot of heterogeneity exists within one particular target group, such as tourists, as preferences of (prospective) visitors vary according to their activities, interests and opinions (Gonzalez and Bello, 2002). Let us consider the Amsterdam example again. The young "backpacker" from the United Kingdom could see Amsterdam as a bohemian place and associate the city with "sex, drugs and rock "n" roll", whereas an American pensioner considering a trip to Europe could associate Amsterdam with heritage, culture, the typical canals, monuments, museums and the Anne Frank House. Their city sub-brand perceptions are so different that it is useful to communicate two separate sub-brandsa bohemian and a high cultural one. Of course, there is also interference between these separate sub-brands: if the American pensioner is a regular viewer of the O'Reilly Factor on Fox News,[3] he might be confronted with a negative view on bohemian Amsterdam, where Amsterdam is portrayed "as a cesspool of corruption, crime", where "everything is out of control", and children are taught "to have safe sex and smoke grass". Two separate sub-brands could, in this case, be helpful for lowering such a spilloveror at least be a more efficient reaction to such a perception problem. Even though further segmentation could be valuable, more brands also lead to more complexity, and the question remains as to how many city sub-brands are still manageable.

Theoretical implications and future research
The City Branded House Strategy and the described problem of city brand complexity should also hopefully stimulate further empirical testing in the laboratory and in the field. The concept of brand architecture included in the proposed approach has been studied in marketing, but very little research has tested how individuals react to the increasing complexity of a city brand (an exception is Zenker et al., 2017).
In addition, a comparison of effectiveness between existing place brand management models could be of interest. The participatory place brand management model of Zenker and Erfgen (2014) could be easily incorporated in the City Branded House Strategy, by adjusting that the different stakeholders applying with their ideas for funding must show impact and brand-fit with either the umbrella brand or a specific sub-brand. The model of Hanna and Rowley (2015, p. 110) on the other hand may be "useful for practitioners to undertake a mapping exercise to evaluate the relative impact of the places' different brands (including the one that they are responsible for) on the overall coherence of perceptions of THE Place Brand", but lacks in contribution for what to do as a place brand manager, besides identifying and managing stakeholder brands. Therefore, we regard the City Branded House Strategy as a valuable improvement for place branding. It could be argued that the Place Brand Web by Hanna and Rowley (2015) is more a grassroots approach, identifying existing parts of the place brand, whereas the City Branded House Strategy can be seen as more of a top-down approach, strategically selecting target audiences and the communicated place brand. Having said this, the City Branded House Strategy is open to more participatory approaches and is based on a definition of place brands that highlights the importance of all place customers as part and source for the place brand. Nevertheless, place branding is often a normative and selective process (as such our strategy can be used here). This has been rightly criticized by many researchers (Bennett and Savani, 2003;Boisen et al., 2011;Braun et al., 2013;Kavaratzis and Kalandides, 2015;Philo and Kearns, 1993;Zenker and Erfgen, 2014), and this warning should be taken into account by place brand managers using our strategy.
However, in our opinion, the City Branded House Strategy (and our comprehensive place brand definition) fulfils all criteria for a good translation of marketing knowledge to the context of cities and urban governance. It incorporates contemporary branding insights and appreciates the public context. The strategy will be helpful for city marketers and urban policy-makers alike dealing with diverse target groups, and is likely to improve the target group-specific communication. By following our approach, city sub-brand managers could concentrate more on the specific demands of their target group, and identify their place competitors more easily.
Notes 1. Please keep in mind that the communicated brand is not the logo/slogan (Govers, 2013), but the values that fit existing brand knowledge, or that should be implemented in the brand knowledge of the target groups.