Managerial Ideologies Dividing the Corporate Elite: A process study of the rise and fall of a counter-ideology

Ideology: Nature and structure

The nature of ‘ideology’ has been subject to multiple interpretations, with four main perspectives currently to be found in management research. The first is the (neo-)Marxist view of ideology as ‘a system of wrong, false, distorted or otherwise misguided beliefs’ (Van Dijk, 1998, p. 2) which prevents the proletariat from questioning the dominance of the ruling class (Purvis & Hunt, 1993). While this view of ideology as a distortion of the truth, and means by which the ruling class protects its self-interests, has persisted in critical management studies (Alvesson, 1985; Brunsson, 1982; Clegg & Dunkerley, 1980; Clegg, Courpasson, & Phillips, 2006; Filby & Willmott, 1988), it was amended by Gramsci (1971). He argued that ideological alliances can transcend class alliances, and that the dichotomy is not between (false) ideology and truth, but between good and bad ideology. Gramsci’s objection is therefore not to ideology as a system for maintaining dominance, but to limiting the concept to manipulative, oppressive and ‘false’ ideology disguised as ‘common sense’ (Böhm, Spicer, & Fleming, 2008; Laclau & Mouffe, 1985).

A third view of ideology seeks to eschew its (neo-)Marxist associations, and the link with domination and control, by conceiving it as a set of emotionally charged beliefs about cause and effect, as well as values and preferences for certain outcomes that are shared by a group (Beyer, 1981; Beyer et al., 1988; Meyer, 1982a, 1982b; Starbuck, 1982; Trice & Beyer, 1984; Weick, 1995). We would term this view culturalist, because it places emphasis on the role of ideology as a means to promote group cohesion and solidarity, rather than a means to advance the interests of the group (Trice, 1993). It emphasizes the subjectivity of ideologies as systems of belief which frame how ambiguous issues are collectively interpreted and acted upon (Barrett et al., 2013, p. 204). The more ‘neutral’ view of ideology (Parush, 2008) propagated by the above authors has been criticized by Weiss and Miller (1987, 1988) as almost synonymous with culture and thus theoretically redundant. Weick (1995) offers a potential way forward by defining both ideology and culture as different, although not mutually exclusive, resources for collective sensemaking. As sensemaking resources, both culture and ideology are not just passively internalized but actively produced and reproduced, thus co-creating reality – a process which Weick terms ‘enactment’. Ideology is distinct from culture in providing a collective vision to tame the unavoidable ambiguity of the future. While Weick’s work on sensemaking has been widely cited, the role of ideology as one of the main ‘raw materials’ for sensemaking and enactment has not.

A fourth, more recent version of ideology is advocated by scholars from more critical traditions (Chiapello & Fairclough, 2002; Fairclough, 2005; Laclau & Mouffe, 1985). They accept ideology as a resource for sensemaking, but emphasize its potential as a means of subordination and even oppression. They recognize the subjective nature of ideologies means that dominant ideologies do not exist in isolation, and are open to challenge from counter-ideologies. Similar to culturalist approaches, particularly Weick’s concept of enactment, they acknowledge how ideologies allow groups to act in conformity and promote a sense of shared identity and mission. From this pluralist standpoint, the structure of such ideologies encompasses ideas, values and beliefs about: (1) the identity of the group, who we are and what we do; (2) goals (i.e. the desired future state of affairs that drives both decision-making for the future and (re)evaluation of the past); (3) rules of the game (legitimate ways of achieving the desired state); and, (4) shared values and norms (Van Dijk, 1998, 2001, 2006, 2011). These group-level ideological allegiances are often accompanied by a deep sense of ‘us’ and ‘them’: who the group’s allies and opponents are; what must be done to create a better future; as well as what went wrong and what went well – and why – in evaluating the past. We will use this pluralist post-Marxist framework for the first part of our analysis, the identification of the competing ideologies in our case.

The pluralist conception of ideology and the way it can divide groups has been typically associated with political parties, activist groups etc.; but management and management theories also build on competing managerial ideologies (Barley & Kunda, 1992; Bendix, 1956; Czarniawska-Joerges, 1988; Guillén, 1994). Ensuring that ideology is shared is an effective way for elites – those who occupy positions of authority and are in charge of strategic decision-making (Dörrenbacher & Geppert, 2016; Pettigrew, 1992; Zald & Lounsbury, 2010) – to achieve coordinated action among those they lead. In the management literature, ideologies have therefore been described as powerful ‘systems of control’ (Barley & Kunda, 1992; Bendix, 1956; Czarniawska-Joerges, 1988). These studies of managerial ideology identify that, both in management theory and in practice, normative ideological control systems favouring autonomy and communalism are in tension with rational ideological control systems favouring authority and discipline. Czarniawska-Joerges (1988) links this tension to the paradoxical organizational need for both freedom and order, autonomy and authority.

How this tension plays out in practice, however, has not received much attention from management scholars in recent years. The relative absence of ideology in the management literature has been attributed to ‘the fashionization’ (Czarniawska, 2005; Parush, 2008) of alternative concepts such as ‘identity’, ‘culture’ and ‘institutional logic’ (Suddaby & Greenwood, 2005) that can also be classified as sensemaking resources (Weick, 1995). But ideology differentiates itself from other sensemaking resources in how it entails a strong sense of purpose and a shared moral compass in the face of ambiguities that can divide groups over what is ‘good’ and what is ‘bad’ when deciding about future action. This is why ideology deserves its place in the management literature.

A pluralist approach to ideological change processes

Most of the management literature on ideology has either concentrated on the way in which dominant ideologies have endured or have gradually evolved over time. Alvesson (1985), for example, critiques the dominance of the ideology of technocratic rationality, the one-dimensional man (Marcuse, 1964), in management literature and practice. Others have studied how ideologies adapt to new national contexts (Guillén, 1994) or to isomorphic pressures from external ideologies permeating the ‘walls’ of the organization (Pratt, 2000; Simons & Ingram, 1997). When discussing ideological change in this paper, our focus is not on these types of change, but on the more disruptive change from one ideology to its competing or counter-ideology.

Pluralist studies on managerial ideology provide us some clues for addressing the question of how counter-ideologies challenge the dominant ideology within an organization. Barley and Kunda (1992) found that management thought in the United States had, over a period of more than 100 years, oscillated between ideologies of ‘rational’ (technocratic systems of design) and ‘normative’ (humanistic systems of trust) control, arguing that over time the weaknesses of either ideology will be exposed and lead to a surge of the counter-ideology. Czarniawska-Joerges’ (1988) work on ideological control, which remains one of the most comprehensive empirical studies of ideology within organizations, found that normative and rational ideologies coexisted in the companies she studied. Her focus thereby is on ideology as a system of control: how ideologies facilitate collective action and therefore organizational control. What she does not explain is how a dominant ideology can be challenged or replaced by a competing or counter-ideology.

A more comprehensive starting point for understanding the dynamics of ideology that can explain such ideological challenges can be found in the writings of Teun van Dijk. His extensive body of work (1998, 2000, 2001, 2006, 2011), which advances the post-Marxist, pluralist view on ideology, has been highly influential in other areas of the social sciences. He argues that ideology is a ‘multidisciplinary’ and dynamic concept that can only be analysed and explained as the interaction of cognitive, discursive and social mechanisms or forces. Ideologies are discursively expressed and reproduced, but equally belong to the field of thoughts and beliefs, or what psychologists call ‘cognition’. They are social in the sense that they are related to group interests, to conflict, and used to legitimize or oppose power. Any explanation of ideological change which is based on mechanisms belonging to only one of these disciplines, he claims, is reductionist (Van Dijk, 1998, pp. 8–10). Van Dijk concurs with Weick (1995, p. 113) in conceptualizing ideology as a ‘sensemaking resource’. Sensemaking mediates between deep structures (ideology) and discourse, a ‘surface’ expression imbued with ideology. Discourse, however, seldom offers a direct articulation of ideology. To uncover the ideological beliefs, Van Dijk urges us ‘to look closely at their discursive manifestations’ (1998, p. 6), that is, the value-laden lexical expressions that group members share in their talk and the presuppositions they make in explaining cause-and-effect relationships.

Van Dijk has not examined ideological change within organizations, a highly structured context where ideological neutrality is often assumed (Czarniawska-Joerges, 1988). Applying his framework to the limited work that exists on ideological change in organizations reveals that studies typically concentrate on a single reproduction mechanism operating in isolation. Most recently, since the ‘linguistic turn’ (Alvesson & Kärreman, 2000), discourse has been a prime way of understanding how ideology is changed or maintained in the organization (Barrett et al., 2013; Riad, Vaara, & Zhang, 2012; Vaara & Tienari, 2008). Van Dijk’s framework would suggest that this reductionist approach risks understating the social and cognitive mechanisms at work in the process. Management research on ideologies prior to the linguistic turn was similarly reductionist, explaining ideological change by concentrating on either social or cognitive mechanisms. The starting point of cognitive explanations is that ideological commitments are maybe influenced, but not determined, by our backgrounds and experiences, and are open to change. Starbuck (1982, p. 6) finds that external crises can lead to an ‘ideological reorientation’; also, Brunsson (1982, p. 40) speaks of adherents’ loss of faith in the dominant ideology during crises. Simons and Ingram’s (1997) explanation of ideological change similarly relies on isomorphic pressures that lead to cognitive adaptation as well as cognitive resilience, the degree to which adherents are willing to hold on to the dominant ideology. Czarniawska-Joerges offers four triggers for how cognitive ideological change can occur at an individual level – either in leaders or followers: learning, triggered for example by new experiences or self-development; a change of perception as a result of persuasion and/or new encounters; striving for consistency triggered by the threat of cognitive dissonance; and functional change whereby new needs or motives arise (1988, p. 118).

Social mechanisms for ideological change within organizations have been infrequently studied. Czarniawska-Joerges (1988) highlights how a change of controllers can generate ideological change, a change that is often triggered by external disturbances or a crisis (Czarniawska-Joerges, 1988; Starbuck, 1982). During a crisis, ideology becomes ineffective in its influence (Czarniawska-Joerges 1988, p. 114), which allows opposing parties to challenge the credibility of the controllers (or what she calls the ‘ideology owners’). Within each ideological coalition or group, ‘adhering to a “right” ideology is a source of many social and psychological rewards: social acceptance, sense of belonging, reduced uncertainty. Deviation produces corresponding punishment: a feeling of isolation, cognitive dissonance, possible open criticism’ (Czarniawska-Joerges, 1988, p. 10). Despite these social pressures to subscribe to the dominant ideology, not everyone will believe in equal measure: Levels of commitment may vary and produce true missionaries, new converts, agnostics, sceptics and heretics (Czarniawska-Joerges, 1988, pp. 91–99); Van Dijk (2006, p. 119) speaks of varying ‘degrees of membership’.

While extant studies demonstrate that organizations are multi-ideological sites (Czarniawska-Joerges, 1988), not much is known about how managerial ideologies compete, in particular how and why a counter-ideology in an organization can rise and fall. In the following section, we will explain how we have used Van Dijk’s theoretical understanding of ideology for our analysis of how managerial ideologies compete. First, we use Van Dijk’s ideological framework to excavate the competing managerial ideologies from the interview transcripts. We then identify the ‘multidisciplinary’ process mechanisms driving the rise and fall of a managerial ideology, which challenged the dominant ideology and homogeneity within Unilever’s transnational elite. The case is able to offer a unique insight into the ideological tensions inside the corporate elite that arose when the company needed to interpret and act on the rising social demand for greater corporate social responsibility. The CSR debate challenges the dominant rationalist beliefs about the purpose of the firm, and is thus directly linked to ideology (Haase & Raufflet, 2017).

Step 1: From surface-level text to deep structures

Understanding the underlying belief systems which are drawn upon in the sensemaking process requires what Fairclough (2005, p. 916) has termed ‘analytical dualism’: an analysis of what happened at the surface level, and what transpires from the level of deep structures. Sensemaking is necessarily retrospective (Weick, 1995) and draws on deep underlying belief structures. These deep-seated belief structures are manifested at the surface level in the claims made by the narrator about past events. The historical narratives which comprised our dataset were therefore well-suited to the unmasking of underlying belief structures (Langley & Tsoukas, 2010; Maclean et al., 2014). The interviewees did not just report on the divisions which they experienced at the time of the events they narrated; they continued to be divided at the time of the interviews. As a result, their narratives were highly rhetorical, seeking to persuade the listener that their interpretation of the surface-level events was the correct one. Emotions continued to run high, despite the period of time which had elapsed – a period of time which also possibly made it easier for them to be more reflective about their experiences.

To reconstruct the chronological sequence of events from the surface-level data, we coded the events that occurred based on dates and places mentioned by the participants, as well as in internal documents and press releases. For the unmasking of the deep structures, we followed Van Dijk’s (1998) method of ideological analysis and concentrated on the coding of two aspects of the interview transcripts. The first was the value-laden lexical choices that interviewees made, such as the use of metaphors and hyperbole: ‘Dove was in big shit! They started this fruit-salad approach and practically killed [the brand]’ (interview 21). The second focus for coding lay in the conflicting propositional statements, e.g. Was an increase in sales the result of the CFRB and thus proof of its success, or were other explanations given for the increase in sales? This included implicit ‘truth claims’ found in the narrative: ‘We had an entirely schizophrenic brand shouting out our campaign for real beauty on one hand, but had no idea what the hell that had to do with the products we sold on the other’ (interview 52). We would translate this into the presupposition that ‘good CSR strategies must create shared value’.

The above method enabled us to identify the deep structures underlying the conflicting narratives (see Table 1). Juxtaposing the two competing sets of beliefs revealed that one was typical of what Barley and Kunda (1992) called a rational managerial ideology, the other being what they called a normative managerial ideology. We could then sort interviewees on the basis of their ideological allegiances, with 32 aligning themselves to a normative and 22 a rational ideology. The remaining four could be classified as ‘conversion stories’, because participants had changed sides over the course of the 10-year period (‘I used to believe A, but now I can see that A was wrong and B is right’). In juxtaposing these sets of beliefs, our objective was not to determine which of the two represents ‘good management’ but concur with Czarniawska-Joerges (1988) that they illustrate the ambiguity of managing.

OPEN IN VIEWER

Table 1.

Structure of normative versus rational ideology.

Structural dimension	Normative	Rational
Managerial role: what role we should play in the organization	Transformational leadership is the preferred way to manage a global brand	Transactional leadership is the preferred way to manage a global brand
Organizational goals: desired state of affairs	Global alignment through shared vision and passion	Global alignment through well-designed systems and clear operating procedures
	An organization that can bring out the best in people	A unified organization that is run like a ‘well-oiled machine’
Rules of the game: legitimate ways of achieving the desired state for the organization	Becoming part of a political debate is positive; a sign that the brand is changing society	High performance culture: need to achieve individual targets, deliver quality, participate with diligence and discipline
	Risk-taking, experimentation, emotional involvement and following intuition are broadly legitimized practices	Minimize internal resistance and debate, maximize efficiency and standardization
	The pathway to success is experimentation, trial and error, i.e. learning	The pathway to success is ongoing centralization and specialization
	Involve as many people as possible in dialogue and experimentation	Being in control: becoming part of a political debate is negative; it cannot be controlled and unnecessarily exposes the brand to critique
Values and norms:	Empowerment, passion and creativity	Objectivity, analytical rigour and discipline

Step 2: Process analysis – change mechanisms driving the rise and fall of a counter-ideology

To reconstruct what mechanisms drove the rise and fall of a counter-ideology we first divided the 10-year sequence of events into three temporal brackets (Langley, 1999), which were ‘separated by identifiable discontinuities in the temporal flow’ (Langley, Smallman, Tsoukas, & Van de Ven, 2013, p. 7). Changes in leadership – at times in combination with a perceived crisis – triggered the discontinuities in the dominant ideology. As Langley et al. (2013) note, temporal bracketing allows for the identification of the mechanisms that explain the changes that were taking place in each era. To do so we had to question the data – or set out to find more data – to find the most plausible explanation for the processual patterns observed (Sayer, 1992, 2000). Van Dijk’s multidisciplinary conceptualization of ideology thereby urged us to look for not only discursive but also cognitive and social forces. In addition, we looked for contextual forces which supported or hindered the rise and fall of a counter-ideology in an organization.

Step 3: Presenting our findings – the construction of ‘composite narratives’

Every qualitative researcher faces the dilemma of how to present the rich, first-hand accounts of the study participants. In this paper, we have used the representational device of two contrasting ‘composite narratives’ (see Brown, 1998, for a similar approach) to convey the two ‘strong plots’ – the rational and the normative – we encountered in the stories of our participants. The narratives that we present are composite in that they are not derived from any single informant (Jarzabkowski, Bednarek, & Lê, 2014); rather, they have been constructed from multiple interviews (for more details on how the narratives were constructed, see the Appendix). Composite narratives as a means of presenting the results of our ideological analysis have a number of advantages. They allow us to present not just the chain of events – crucial to a process explanation – but also how each group made sense of these events. They also illustrate the stark differences between the groups, in terms of their word choices and interpretations of events. What the narratives do not capture is the diversity among participants who varied in their degree of ideological commitment. However, in this paper, our focus is on between-group differences, rather than differences within each ideological group.

Step 4: Verification and validation of findings

To verify our findings at the surface level – that is, the event sequences on which our composite narratives are based – we were able to cross-check across our multiple informants, as well as triangulate our interview data with press releases and other publicly available company reports. This allowed us to resolve any discrepancies in terms of dates, or to fill gaps in recollection. However, while it was possible to verify the chronology of events, we acknowledge that the ideological beliefs we excavated from the narratives – as well as the composite narratives we constructed for our representation of interviewees’ interpretations of events – could not be subjected to a similar process of ‘verification’. As Rhodes and Brown (2005, p. 480) caution, we cannot claim these narratives to be ‘“true” in any objective sense’. Rather, we have been transparent about how these narratives were constructed, retained an audit trail of the coding process and narrative construction, and carefully sought to attain verisimilitude in our representations (Van Maanen, 1988; see also Appendix).

Checking whether we achieved verisimilitude took the form of presenting our findings to three of our key participants: one representing the normative account, one representing the rational account, and one consultant who had been closely involved in the brand strategy from 2003 onwards but had changed sides when the rational ideology had been re-established. All confirmed that the explanations were both plausible and illuminating, although at first confronting. We will return to this reaction in the conclusion, but first we present the results of our dual-layered analysis, going from the surface level to the underlying ideological structures and dynamics.

Analysis of ideological structures

Our analysis of the lexical choices and propositional statements used in the individual narratives clearly highlights the conflicting belief structures underlying the two accounts. Analysing the nature of these (see Table 1) leads us to conclude that they are consistent with what Barley and Kunda (1992) called a rational and a normative managerial ideology.

Process analysis of ideological contestation and change

As mentioned above, we divide the process of ideological change in three distinct eras. For each era, we will first present our data in the form of two conflicting composite narratives offering two interpretations of the same era in the history of the brand. These competing stories reflect the key events in each era (marked as E1, E2, etc.); they illustrate the typical lexical choices (underlined in the text) and causal presuppositions made by each group (verbatim explanations in italics in the text), and provide the basis for a process understanding of ideological contestation.

Era 1: The rise and diffusion of a normative ideology (2001–2005)

This era starts when Dove’s first global brand leader was appointed (E1). She was not given any formal authority, but she had the full support of the global chief marketing officer, who had always been ‘something of a maverick’ (interview 32) and gave her considerable autonomy. Although the decision-making power lay with the regional Dove managers, and local Dove managers still had considerable influence, they now all formed part of the Global Dove Team.

The normative narrative:

Initially we were looking for a unique position for Dove that would help us turn the brand into an iconic brand (E1). We realized that the only recipe to manage a global brand was to have big ideas. We also wanted to make sure that everyone around the world spoke the same language. If you just say ‘the principle is that we have to be global’ you go nowhere. You will always have somebody against you. So we started working on a White Paper, and every time we looked at anything, Susie Orbach’s [a clinical psychologist highly critical of the beauty industry] name came up, so we called her. She was thrilled that we called and was interested in helping out (E2). She made us realize how hard it was for women to live in a world with such a limited notion of beauty. Some of the women in the team could relate this to their own personal experience and there was a strong sense that we should do something about it. Soon it became a very personal mission for everybody involved.

At the end of 2002 the core team went on a leadership course (E3). The course was designed to make you look at things in a very different way, with a huge emphasis on making a difference in the community. It was all about the impact you can create; that you are not a kind of a victim within a big bureaucratic organization. It led to this change of the paradigm in terms of, no, I’m actually at the centre of this and can have a massive impact on the people around me, the world as a whole (E4).

During the leadership course we decided to write the Dove manifesto (E5). There was a sense of hey, let’s experiment, let’s do something meaningful, we have nothing to lose. It all helped to create an environment in which something new and brilliant could come out. Afterwards we shared the manifesto in big 200-people meetings all over the world (E6). All these discussions helped in the end to shape the campaign. Testing was not a concern. We just knew it was right. Just do it. In the process, we became a passionate team; a strong global Dove community (E6). What really drove us was that incredible personal passion.

And when we first launched the new campaign with real women (E7) the PR exploded, sales improved, market shares improved, it was a machine, really; products were flying off the shelves (E8). It was not something that we had foreseen, it just happened.

The new campaign offered a distinctive point of view on beauty. Dove was now such a loved brand; we became so famous that we had famous people talking about the brand, such as Oprah Winfrey. When we first came out with the billboards showing real women and not models, it was just like electricity being turned on. You could just feel it. Women were talking about it everywhere. The campaign started to change society (E8); we became part of politics.

We were on a shared mission, a journey to improve the way women felt about themselves. We did get negative reactions, but we were comfortable with that because we wanted to have a debate. We were not afraid of that. Was it out of control? In a way, yes, but as far as it did, getting out of control was very exciting’.

The rational narrative:

The time we launched the campaign for real beauty (CFRB) was an exciting time (E7). We were winning awards and created a lot of positive brand awareness that gave Dove an ability to charge a premium in the marketplace over commodities. But an increase in sales is always the outcome of a complex set of factors; the things that make a brand work are very multi-levelled. Dove did grow at the time, but we were also expanding into new categories and new countries; we introduced some innovations that were very successful (E8), so it is very difficult to isolate the impact of the CFRB. In the end, the campaign proved to have a number of problems. First of all, we were not able with that campaign to grow our core business. The social mission was not linked to the products. What we found as a result was what we often refer to as the ‘love Dove, buy Olay syndrome’, people loved the brand but didn’t necessarily buy it. They will say thank you for that Dove, but where is my Pantene, where is my Olay?

Although the CFRB produced some brilliant advertising that created a lot of goodwill for the brand (E8), it became evident that the increasingly aggressive stance the brand was taking towards its own industry was always going to render it deeply uncompetitive in the beauty industry. The brand was promoting a social mission that legislated itself out of existence; we were attacking an industry we were part of. If the brand says you are great the way you are, then why sell beauty products? We opened the door to criticism because we were critiquing. When we launched an ad that directly attacked the beauty industry that is probably where we went a bit too far. It did get some negative PR (E9)… it is a bad example of us going too far.

Another concern was that we were leading a global brand with a global footprint with the majority of your growth opportunities probably sitting in Asia, and we were positioning the brand in a way that obviously didn’t work out at all in Central and Eastern Europe, Russia, not even mentioning Asia (E10). Yet at the time, to say it did not work in Asia was heresy.

The key events driving ideological change in the first era are summarized in Table 2, along with our analytical categorization of each event. The era marks the rise of a normative ideology in a predominantly rational environment. The era starts with a major structural change that took place in 2001 and let to the appointment of Dove’s first global brand leader (E1). At the time, Dove was marketed in more than 80 countries around the world with diminishing strategic alignment. The global leader was challenged to realign the dispersed brand team behind a shared brand strategy. In search for an inspiring global positioning for the brand, she – together with her key advisors – turned to experts, ‘new voices’ whose very different way of making sense of managing a brand triggered a cognitive process of reflection and ideological change (E2). This cognitive change was further solidified and socialized when the global core team – including all the regional brand leaders – jointly attended a leadership development course (E3) that was later described as quite controversial within Unilever. The intensive and deeply personal course represented a ‘new voice’, a new expression of making sense of what it means to manage, and led to a powerful cognitive transformation (E4) or ‘ideological conversion’ for most of the core team members. They started to see their role as managers as more than increasing Dove’s market share; it also placed them in a powerful position to make a difference in the world. But the ideological transformation went further than that. They also aligned on how to achieve this new purpose: to experiment, to avoid bureaucracy and to follow their passion. With the intention to share their new-found beliefs with the rest of the company (to socialize the normative ideology), they discursively crafted and proposed a ‘social mission’ for the brand in a document they called a ‘manifesto’ (E5), not a strategy.

OPEN IN VIEWER

Table 2.

Key events for Era 1 (2001–2005).

ERA 1
Event #	Event description	Analytical categorization of event
E1	First global brand team appointed as part of a corporate restructuring towards increased globalization (2001)	Internal trigger: structural change
E2	Psychologist contacted for advice (not supportive of rational managerial ideology) (2002)	External trigger: new voices
E3	Core team attended a leadership course that later proved to be quite controversial (not supportive of rational managerial ideology) (2002–2003)	External trigger: new voices
E4	Team decided to embark on a social mission (2003)	Cognitive transformation: type 1 change = cognitive shift
E5	Core team wrote manifesto (2003)	Team leadership proposing new ideology
E6	Core team built a strong global ‘Dove Community’ (2003 onwards)	Socializing of normative ideology
E7	CFRB launched (2003–2004)	Ideological enactment (normative)
E8	Profit, market shares and PR up (market success) (2003 onwards)	Positive market feedback (reinforcing socializing of normative ideology)
E9	New ads attacking the beauty industry (talk to your daughter before the beauty industry does) (2005)	Ideological enactment (normative)
E10	Internal disquiet over market performance in Russia and Asia (2005 onwards)	Negative market feedback (weakening normative ideology)

Cognitive, discursive and social mechanisms were all intertwined in the rise and socialization of this new (counter) ideology. To socialize the new ideology within the larger global team, large inspirational meetings were organized around the world (E6). Without structural power, the global leader relied on discursive rhetorical acts of persuasion and motivation to socialize the vision and beliefs underlying the manifesto. The outcome was a growing coalition of highly motivated ‘converts’ around the world. They started to make sense of their role in the world around them based on a normative rather than rational management ideology, and acted accordingly (i.e. enactment of the new ideology). They took part in national political debates on women’s self-esteem, used brand budget to support NGOs that addressed eating disorders, and together developed a new campaign that translated the manifesto into a branded message. Once this new campaign (CFRB) was launched (E7), positive publicity and a growth in sales (positive feedback) strengthened the coalition (E8). In their stories, this positive feedback confirmed the ‘rightness’ of the new ideology, and reinforced its socialization among a growing group of managers.

Although a majority of regional and local brand managers converted, not everyone did. A group of sceptics and agnostics formed but did not feel their voice would be heard. They would be dismissed as ‘heretics’. As can be seen from the sensemaking in their narrative, they attributed the brand’s success to other factors than the CFRB. While the ‘ruling coalition’ interpreted some of the negative feedback (E9) as an unavoidable consequence of being part of a social debate, the opposition interpreted it as a sign of the brand ‘going too far’. They were seriously concerned about the damage this would do to the brand in the longer run. They also point out that the strategy was not sufficiently global (E10). Public response in countries such as Russia and Japan was not as positive as it was in the western world, something that was interpreted as part of the learning curve in the normative account, but was brought forward as evidence of a fallible brand strategy in the rational account. But, at this point in time, the positive feedback dominated and immunized the ruling party from external opposition and internal doubt.

Era 2: The brand in crisis? (2006–2007) – ideological struggle

This era starts when the global brand leader was promoted and a new leader was appointed (E11). The promotion was part of Unilever’s standard operating procedures whereby successful senior managers were promoted every 2-4 years. Around the same time, Unilever moved from a regional to a global structure. In this new structure the global brand leader obtained formal power, which meant that his influence no longer solely depended on persuasion and motivation, as had been the case for his predecessor. Not long after these structural changes, Dove sales started to slow down.

The normative narrative:

In 2007 sales were disappointing (E12). This caused a lot of internal debate and anxiety (E14) and suddenly there were voices saying that what the Dove team did was wrong, like total rubbish; that the CFRB did not work; that the brand became angry and feminist and had gone too far. We didn’t go too far at all. The problem was that the original Dove team had been broken up (E11). It could have been a different story if the original team could have continued working on the brand. Key players were replaced by people with a much more aggressive ‘I tell you, you bloody do it’ sort of a management style. With the old team we were on to something. We always knew this was as much about the people, who were involved in something they really believed in, as it was about the strategy and consumer insight, so the moment you start to take these people away you lose something. Often the most iconic, the most successful brands that embark on unusual, challenging endeavours are driven by a group of committed, passionate individuals and the thing basically works for as long as these people are in place but how do you create such a model in a multinational where inevitably people tend to rotate? With this new group of people, I thought I’d jump ship.

The reason for the disappointing sales was that we had an innovation strategy that did not work. Structurally, Dove was not part of the category teams that controlled the R&D agenda (E13). Especially in skin care we had not been able to build strong links with the category team. For example, we wanted to enter face care; but they had no intention at that time to enter face care. We launched the face care products anyway (E15), but arguably the quality was not as good as we would have liked it to be. The products didn’t sell and our credibility suffered (E16). As a result we were not able to get approval for another innovation we developed, Dove for Men (E17), a range that later proved to be very successful.

The big learning was Asia. Because, to be honest, the campaign did not work really well in Asia (E16). The CFRB was developed when we still worked as regional teams. In 2005 we became a global team. We rolled the CFRB out in places like Russia and Japan with very little testing. The success in Europe and North America had been so overwhelming that it took about a year for the global team to realize that it really was not working everywhere. Asian countries were persuaded to launch as well. It was not a disaster; it was just that women were not jumping on their chairs about it.

The rational narrative:

The debate on ‘real beauty’ was more and more taken over by political pressure groups in the US. Dove had become more and more angry and feminist and had gone too far; it was becoming the anti-beauty brand (E14). It was too politically aggressive and needed to be cleaned and tightened up. The brand did not have a solution for the much bigger challenge we are facing: to achieve both brand integrity and business sustainability. 2007 was a disastrous year (E12). The sales of Dove’s core products stopped growing because we had put all the money and energy behind new products and communication of the social mission which wasn’t linked to the core products (E16). In fact it was a huge mess, we had to simplify things, make them much more logical.

Also in terms of discipline , the brand needed greater control . If you want to succeed in consistently managing a global brand you must have clear processes and a culture of adherence. It’s called a performance culture; targets are targets and you have to achieve them. The problem was that local brand managers still often did whatever they wanted, which is a waste of resources and not necessarily the best for the brand. With clear and well-designed processes you can massively reduce the risk of maverick individuals and of non-compliance.

In our analysis of what caused the normative ideology to weaken and the ideological struggle to intensify, we found that the process of ideological change in this era was triggered by: (1) internal structural change; and (2) a crisis in the form of negative market feedback (Table 3). Structurally, the promotion of the global brand leader (E11) to a position outside the Dove team disrupted the cohesion in the team and the confidence they had in the way forward. The new global brand leader supported his predecessor’s brand strategy, but was not as fervent a believer in the normative ideology. The mixed messages he sent out to the team failed to strengthen the coalition, and the global Dove team became ideologically divided. Although the most fervent advocates remained loyal to the cause – ‘for us it was a movement, not a commercial thing’ (interview 29) – some of the less entrenched followers of the new normative ideology indicated that, at this point in time, their belief in the ‘new way’ started to waver. In their retrospective interviews, they reported this cognitive shift: for a while there had been ‘a temptation to a certain level to eliminate objectivity’ (interview 36), and although they had loved that time, they felt they had to get back to their senses (i.e. to return to a rational ideology).

OPEN IN VIEWER

Table 3.

Key events for Era 2 (2006–2007).

ERA 2
Event #	Event description	Analytical categorization of event
E11	Global brand leader promoted and team changed (2006)	Internal trigger: structural change
E12	Global market shares stagnated (2007)	Negative market feedback (weakening normative ideology)
E13	Separate global brand division created for Dove (2005)	Internal trigger: structural change impeding broader coalition-building
E14	Internal anxiety increased and opposition voiced (2007)	Socialization: weakening of coalition building (weakening normative ideology)
E15	Products launched despite lack of R&D support (2006–2007)	Ideological enactment (normative)
E16	Slide in market share continued despite launch of new products (2007–2008)	Negative market feedback (weakening normative ideology)
E17	Internal support withdrawn (2007 onwards)	Internal feedback: weakening of normative coalition

Although the new global structure introduced in this era was all around setting up global category divisions (e.g. deodorants or skin care), an exception was made for Dove, a brand operating in three different categories. To preserve its success, Dove became its own global brand division (E13). This unique structure unintentionally had the effect of isolating the team, thus limiting further socialization of the ideology. In particular, a deteriorating relationship with the category and R&D teams who were now part of the global category divisions negatively affected the necessary support for enacting the Dove strategy (new product development; E15).

The process of wavering ideological cohesion was exacerbated by disappointing market feedback (E12). This had both social and cognitive repercussions (E16). A slowdown in sales created anxiety (E14) and needed explanation. As in the previous era, the different ideologies led to fundamentally different sensemaking and, thus, for conflicting explanations about the slowdown in sales. In the normative account, the slowdown was attributed not only to a learning process but also to a lack of support from R&D in the new structure, an issue they felt could have been solved, had they been given the opportunity. According to the narrative of the opposition, the slowdown in sales indicated that the ruling coalition was not in control, and that the brand needed rescuing. What we observe here is that the ideological controversy had intensified, an illustration of what Van Dijk (1998) describes as the growing polarization between ‘us’ and ‘them’ over time. Negative feedback is downplayed in the stories of the ruling normative party, but amplified in the stories of the rational opposition, just as the success in the previous era was amplified in the normative account but downplayed in the rational account.

As a result, the position of the ruling coalition came under threat. They were not able to regain the support of Unilever’s top executives to enact their strategy (i.e. to launch a new range of products that they claim could have turned the situation around (E17)). At a discursive level, the opposition was no longer muted, and the rhetoric on both sides became increasingly divisive and emotionally charged, as the examples in the composite narratives above demonstrate.

Era 3: Redirecting the brand (2008–2011) – reinstatement of rational ideology

The weakening of the ruling coalition and the strengthening of the opposition of the previous era eventually led to a deliberate structural change of leadership (Table 4, E19), and the dismantling of the Dove division (E18). The global team was split into three, each now part of their respective category divisions. The newly appointed global leader for Dove Skin Care was made responsible for the overall brand strategy (E19). Ongoing negative feedback and weakening of the normative ideology had given the opposition sufficient strength to challenge the leadership of the brand. The new global brand leader was part of the ideological opposition and had been concerned about the direction Dove had taken in the last two eras. He brought in a number of managers whom he trusted to be part of his core team. As a result, the leadership team that had led the brand between 2001 and 2007 ceased to exist, and was almost completely replaced by a new leadership team which subscribed to a rationalist managerial ideology. Managers subscribing to a normative ideology became a minority.

OPEN IN VIEWER

Table 4.

Key events for Era 3 (2008–2011).

ERA 3
Event #	Event description	Analytical categorization of event
E18	Global Dove team restructured: global Dove division dissolved (2008–2009)	Internal trigger: structural change
E19	A new global brand leader appointed and team changed (2009)	Internal trigger: structural change (new team leadership)
E20	New rhetoric and new structure of communication introduced (2009–2010)	Socializing of rational ideology: coalition-building; legitimizing rational narrative/delegitimizing of normative narrative
E21	Clear operating systems and processes introduced (2009–2010)	Ideological enactment (rational)
E22	Products of functionally superior quality introduced (2010 onwards)	Ideological enactment (rational)
E23	New Masterbrand strategy and campaign; redefinition of social mission linked to sales (2010)	Ideological enactment (rational)
E24	Market success: profit and market shares up (2011)	Positive feedback: strengthening rational ideology

The normative narrative:

Arguably, in our enthusiasm of ‘we can conquer the world’ which was fantastic and we had a lot of fun, we had become a bit overconfident and had made some mistakes. The new team that was appointed (E19) fixed the product quality (E22) and that was good. But in terms of culture we have become very risk-averse, we have become engineers. It’s a return to rationality. The new system is stifling, everything has to be measured, under control, no risk! (E21); but why should we have to calculate return on investment if something is obviously the right thing to do? We might be little by little killing the fire of this company. The people who have made this company’s brands are people with a passion.

They sneaked away from the principles for the democratization of beauty that we had set and that led to these great campaigns, and what you get is these more mainstream ads, really very boring but arguably with better product cues in them (E23). We were brand evangelists but Dove is no longer changing the world. In terms of sales they did a good job (E24), although it sure helped that we gave them Dove for Men on a plate.

Five years ago we were flying high because of the Dove vision and how it all came together, the way it all felt, honestly you could have given me any job from the outside, I would have said no. Just because it was so super exciting. But you put new people in charge and all of a sudden they find all kinds of reasons of what is wrong with it. How do you preserve something so unique in times that sales are not going through the roof? How do you create such a model in a multinational where inevitably people tend to rotate? There is obviously an important people dimension here.

The rational narrative:

By 2008 it was obvious that Dove was in trouble. A new brand leader was appointed (E19); very clever, very hardworking, really diligent, someone who can actually make things happen. The Dove division was dismantled and split up over the three category teams (E18). As a new global Dove team we found we had to work in a very different way to better leverage our scale and our knowledge around the world. Structurally and procedurally we needed to create and communicate a logic that would enable us to do things in a much more orchestrated way (E20). So for Dove bars, for example, we now have a very clear and standardized repeatable model that we implement depending on the characteristics of a country. To design this logic on a global scale is a stimulating intellectual challenge. You have to work in this centralized and disciplined way if you want to run a global brand (E21). Initially we focused on improving product quality (E22) and communicating this to our consumers without being distracted by politics and social activism (E23). It doesn’t matter if it isn’t always the best message in the world, as long as it works and is consistent. We did a rigorous re-evaluation of the Masterbrand strategy (E23). The first step was to go back to the consumers and do market research to gain deeper consumer understanding (E21). As a result we adjusted the social mission and it is now much more integrated with the product (E23). It may be less glamorous but it means we are no longer faced with the difficult decision of do I spend money on the social mission or on promoting the product? We are aware that the more diehard converts to the 2005–2007 version of the mission feel that something has been lost, but there was a very good reason for that and the sales figures prove it (E24).

As can be seen in the two accounts, the way they make sense of what had happened and what needed to be done to fix the situation is built on conflicting managerial ideologies. The normative account builds on their strong underlying belief in trial and error, passion and autonomy; the rational account builds on a strong underlying belief in order and design. Although the new brand leader was not consciously selected on the basis of managerial ideological preferences, he was considered by the more rationally oriented senior executive at the time to be the right candidate to ‘save the brand’ (interview 22). In political terms, the restructuring meant a takeover by the opposition.

The socialization of the rational ideology was achieved by bringing in new people (E19) and by introducing new norms for managerial behaviour, such as analytical thinking, measurement and disciplined adherence to clear processes and procedures (E21). This social mechanism of legitimizing and de-legitimizing ‘rules of the game’ was bolstered at the discursive level by a deliberate change in rhetoric (E20). As one of the brand leaders reflected: ‘by retroactively imposing a narrative on [past events], you are actually able to lead your company in the … [desired] direction’ (interview 52). As in Era 1, the new ideology was disseminated globally through (re)telling of this narrative in global meetings. The rhetoric included the de-legitimation of some past practices, such as intuitive decision-making without extensive testing, or promoting a social mission that did not simultaneously support the sale of product. Once socialized and internalized by the wider team, it started to reshape team sensemaking and enactment. A new Masterbrand strategy was introduced and new advertising campaigns (E23) focused on the communication of Dove’s superior product quality (E22). The support for women’s self-esteem was not abandoned, but Dove was no longer taking a political stance and blaming the beauty industry for the crisis. Following rationalist ideological sensemaking, that was seen as a poor strategy.

The social cost of not adjusting to the dominant ideology was higher in this era than in the first era, because the rationalist ideology was also the dominant ideology in the broader corporate context. Some of the most fervent advocates of the normative ideology, rather than adapt, found roles outside the company; others stayed and now formed part of the opposition, their frustration often expressed in their narratives. But, at the time of the interviews, the brand was doing well, and any opposition was silenced by the positive market feedback (E24). For others, the convincing discourse of the new regime and the social benefit of belonging had helped them make the cognitive change and to embrace the rational ideology. In their conversion narratives, they remember the days of working on the CFRB with fondness, sometimes even declaring it was the best time in their working lives, but then conclude that it could not last, and bring to the fore the rational arguments of the opposition: ‘We needed to put the house back in order. … The whole CFRB as we had it before was kind of too angry … the product was not embedded in the story’ (interview 10).

While the rational ideology was re-introduced and the normative ideology delegitimized, it was not a straightforward return to the past. The times had changed and some of the sensemaking regarding the social role of the brand had evolved. Creating shared value – a rationalist sensemaking of CSR popularized by Porter and Kramer (2006, 2011) – had become the norm, and all brands were to have a ‘social mission’. Although Dove had been one of the first brands with a clear social mission in the company, its social mission was now rephrased to submit it to a rationalist logic. It had to be linked to selling product. Rationalist values (see Table 1) were thereby not changed, but in order to achieve the ideal image of a perfectly engineered system of value creation a new rule was added, namely, that commercial value creation should go hand in hand with creating shared value.

Proposing

In this case the counter-ideology was proposed and diffused by the newly appointed leader, the ‘ideology owner’ (Czarniawska-Joerges, 1988, p. 91). Leadership involves providing direction and setting out a strategy to achieve the goal or, as Brunsson said, ‘to choose the right thing to do and to get it done’ (1982, p. 37). Both are ideological acts, in that they involve sensemaking and thus assumptions about the desired state of affairs and the ways of achieving them. Leaders can propose ideological ideas that align with the dominant beliefs of the group (Era 3), or propose ideas that represent a new cognitive map that challenges the dominant beliefs, as was the case in Era 1.

Of the three types of ideological change we identified in the case, two led to the proposal of a counter-ideology; one to an evolutionary change within the dominant rational ideology. The types of change that triggered the rise of a counter-ideology were: (1) the appointment of a new leader who challenged the ideological orientation supported by their predecessor (type 1 change); and, (2) the exposure of influential members of the team to ‘new voices’ offering a new rhetoric that led to cognitive dissonance and change (type 2 change). The extant literature discusses cognitive change as a trigger (Czarniawska-Joerges, 1988; Pratt, 2000; Simons & Ingram, 1997). We add three enabling conditions: leaders willing to challenge their own thinking; a relationship of trust with the deliverers of new ideas; and a supportive organizational environment that gives the new leader the autonomy to form their own ideas and lead as they see fit. At the start of Era 1, all of these conditions were in place: ‘It was an alignment of the stars’ (interview 30).

The third type of change, evolutionary change, is also discussed by Kwon and Constantinides (2017), who argue that critical events can force actors to adapt some of their ‘rules of the game’ without changing the core ideology. In this case, the increased demand for CSR shifted the rationalist rule of the game that demanded working towards ‘total shareholder return’ to ‘creating shared value’ (Porter & Kramer, 2006, 2011). We argue that this evolutionary change in the rational ideology also demonstrates that, even though the rule of the normative ideology did not last long, it did leave lasting traces behind.

Socializing

Achieving ideological change depends on successful socialization of the ideology, the cognitive process of capturing the hearts and minds of the key decision-makers in the organization, the corporate elite. This process of socialization, or coalition building, can be discursively achieved through persuasive and motivational rhetoric that offers attractive (alternative) goals and convincing reinterpretations of the past (Barrett et al., 2013). These influence tactics are a form of politicking that walks a fine balance between brainwashing and enlightenment (Czarniawska-Joerges 1988, pp. 10–11). They can be self-serving – as suggested by Marxist thinkers – or arise from a deep conviction that ideological reorientation is necessary for the common good. The global Dove leader in Era 1 successfully used discursive means to socialize a normative ideology; the global leader in Era 3 used both discursive means and (de)legitimation to successfully socialize a rationalist ideology. Both were convinced that their sensemaking was right and necessary for future success. Paradoxically, we note that in Era 3, when the global leader had formal power to control global decision-making and decision-making criteria, ideological ideas were quickly dispersed but not fully socialized. In contrast, the leader in Era 1 was able to build a very strong coalition despite – or because of – the fact that she depended on voluntary acceptance. Formal power can therefore be a two-edged sword in the socialization process.

Successful socialization does not necessarily mean that all opposition disappears. The opposition may lose its less entrenched members (the ‘floating votes’) and be temporarily silenced, but in our case never disappeared completely. As can be seen in the model, ideological cohesion is maintained and strengthened by rhetorical acts from the leadership team and positive feedback. Era 2 shows that a lack of ideological consistency and discursive promotion from the leadership team increased uncertainty and conflict and jeopardized organizational action, a pattern earlier noted by Brunsson (1982). Studying the adoption of ideologies in different cultural settings, Guillén (1994) found that the adoption of ideology also depends on the extent to which it resonates with pre-existing beliefs. Given the predominance of the rational ideology, not only in companies such as Unilever, but also in the management literature (Czarniawska, 2003; Ghoshal, 2005), socialization of the normative (counter) ideology was always going to be a greater challenge. Frequent structural changes, replacements and promotions of team members, and the structural isolation of the team in Era 2, also disturbed the ideological socialization process.

Enacting

Ideology is a powerless system of control unless it guides collective sensemaking and action. An ideology that is successfully socialized directs group sensemaking, decision-making and thus action without constant reinforcement and debate. What was evident in the case was that the short era of normative control did lead to a different type of decision-making. In Era 3, when a rational ideology was successfully resocialized, this type of decision-making was no longer legitimized. For example, launching without extensive market testing was now delegitimized, as were social missions that had no link to profit. Legitimate decisions and actions in this era were disciplined, well-researched, following best-practice protocols, and always aimed at increasing business results.

Feedback

Managers have control over decision-making and action, but not over outcomes (Weick, 1979). The complexity of social reality leads to ambiguous outcomes or ‘feedback’ that then needs to be explained retrospectively. As the conflicting plots in our findings demonstrate, the attributions managers make to explain (or make sense of) an outcome strongly depends on their ideological orientation. It is hard to tell if the disappointing sales figures in 2007 were attributable to Dove’s focus on the CFRB: in the normative narrative, they were not; in the rational narrative, they were. It must be noted, though, that the opposing ideologies did share the assumption that growing profit and market share is positive, making this seemingly beyond doubt in the corporate context.

Although a counter-ideology can rise to the fore, its dominance will be short-lived unless a major ideological shift takes place in the organization at large. As pointed out above, the interpretation of outcomes is not ideologically neutral. When profits were disappointing while the proponents of the counter-ideology were at the helm, it triggered a degenerative cycle (see Figure 1): temporarily silenced opponents regained confidence to speak up and explained the result as the outcome of poor management. This in turn triggered ‘faith’ to weaken among the less entrenched converts and a new leader to be appoined. When a dominant ideology is in control, proponents of this ideology are less likely to find themselves surrounded by counter-ideologues who may attribute negative outcomes to their way of managing, so their vulnerability in the face of negative feedback is not as great.

In conclusion, what our model shows is that the managerial ideologies that guide group sensemaking and enactment are dynamically maintained and not immune to change and disruption. To be stable they need ongoing discursive reinforcement from the leadership team, and positive feedback. External voices, negative feedback, structural changes and changes in leadership can disrupt this balance. The trend of ongoing centralization and rationalization that is taking place in companies such as Unilever is possibly making it less likely for counter-ideologies to rise to the fore, let alone to survive for more than a short period of time. However, our model leads us to suggest that the more a management team is shielded from internal and external counter-ideologues, the more it will find itself in a self-reinforcing cycle that can lead to losing sight of the subjective nature of their beliefs. Ideology becomes knowledge, while counter-ideology becomes lack of knowledge or ‘forbidden knowledge’ (Czarniawska, 2003).