Abstract
This paper analyses Hello Barbie as a commercial artefact to explore how big data practices are reshaping the enterprise of marketing. The doll uses voice recognition software to ‘listen’ to the child and ‘talk back’ by algorithmically selecting a response from 8000 predetermined lines of dialogue. As such, it is a useful example of how marketers use customer relationship management systems that rely on sophisticated data collection and analysis techniques to create a relationship between companies and customers in which both parties are positioned as active participants who are able to obtain what they wish from the interaction. I use dialogic analysis to see how Mattel ‘makes sense’ of the dialogue as a dialogic partner. I argue that, in spite of the rhetoric of instantaneity and personalization, in which the technology is positioned as an immediate response to a child’s imagination, Mattel’s dialogic communication is both asynchronous and carefully crafted to fit the child’s responses within predetermined consumer subjectivities that are crafted to encourage particular kinds of consumption. Although the dialogue spoken by Hello Barbie is able to situate Barbie as an active subject, the control exercised by the company in order to elicit data for customer relationship management purposes and steer the dialogue to brand-friendly messages relegates the child to a passive role. Accordingly, the doll fails to deliver the promises of customer relationship management.
Keywords
This article is a part of special theme on Big Data and Surveillance. To see a full list of all articles in this special theme, please click here: https://journals.sagepub.com/page/bds/collections/hypecommerciallogics
Since the advent of the World Wide Web, companies have sought to use interactive technologies to market to children as a unique demographic (Montgomery and Pasnik, 1996; Steeves, 2006). The incentives to do so are powerful. In 2014, children and pre-teens spent $40 billion directly and influenced more than 12 times that amount in family spending (CRM, 2014: 19). By 2020, the market in networked toys alone is expected to reach $11.3 billion (Smith and Shade, 2018: 2). Marketing to children is accordingly big business, and it is a business that is increasingly being shaped by the contemporaneous collection, analysis and operationalization of the data generated by the devices that children use as they go about their daily lives. Those data are typically fed into customer relationship management (CRM) systems that are designed to come to ‘know’ the child, by translating discrete pieces of information about them into actionable information that can then be used to engage the child in an ongoing dialogue in an attempt to shape their consumption (Smith and Shade, 2018).
Mattel’s Hello Barbie doll is an excellent exemplar of this marriage of marketing, data, technology and dialogue. The architecture of the doll is relatively simple. Sensors embedded in the doll record what the child says and then transmits it over wifi to a third party company called ToyTalk, where an algorithm analyses the child’s speech patterns and selects a response from a database of 8000 lines of dialogue. The response is then sent back over wifi and ‘spoken’ by the doll (Mattel, 2017a). It is this ability to ‘carry on a two-way dialogue’ (Techlicious, 2015) to generate more data that in turn feeds into the CRM loop that makes Hello Barbie so innovative (see also Mattel, 2017b). As the sales rep first demonstrating the doll at its launch at the New York Toy Fair in 2015 said, the doll can really respond to those things that I said to her, she understands them, and she can also remember them … she’s getting to know all of my likes, my dislikes and then she can actually have a conversation with me about those things.
From a marketing perspective, the ‘real prize’ in this approach is that the data can be used to understand how children ‘feel’ about the Barbie brand: ‘This is based on the belief that if one can sway emotions, one has a better chance of influencing cognition’ (McStay, 2016: 4, 5) and bypassing rational decision-making processes; since ‘reasoning leads to confusions’, companies that know how people feel will be able to ‘“help” give people what they truly wish for rather than what they say they want’ (5). Dialogic disclosures are also thought to deepen a consumer’s emotional connection or affection for a brand, and the effect of marketing messages is believed to be amplified when those communications are mediated through technology (Pang et al., 2018: 75). The aim is therefore to ‘collect it all’ through ‘always-on data collection’ to attain ‘the Holy Grail … real-time information about customer needs and emotions … [through] instant access to transactions and conversations’ (McStay, 2016: 4), and to use that data to ‘offer particular kinds of subjectivities to users’ (Graham, 2018: 8) that will generate profit for the company (see also 10).
In this paper, I analyse Hello Barbie as a commercial artefact to begin to unpack how the use of big data analytics is reshaping both the enterprise of marketing (Montgomery et al., 2017) and the dialogic opportunities offered to the children who are inserted into the socio-technical system upon which that marketing depends (see Lupton and Williamson, 2017; Smith and Shade, 2018). I begin by situating Hello Barbie in the CRM literature and tracing the doll’s development as a CRM tool. I then suggest that dialogic analysis can be a fruitful way to interrogate the marketing practices embedded within technologies that ‘speak’ directly to customers, and outline the particular method used to analyse Mattel’s construction of Hello Barbie as a dialogic CRM agent. The findings of our analysis are then presented as a response to four central questions: when Barbie is ‘speaking’, who is doing the talking, what is the speaker doing, who is being addressed and what kind of relationship is anticipated?
I argue that the dialogue situates the doll as a fully personalized subject who exercises a high level of control over the conversation in order to ensure that Mattel’s CRM goals are met. The child, on the other hand, is de-personalized and inserted into the dialogue as a reductive collection of likes and dislikes. Although the collection of these likes and dislikes furthers the CRM project, the child is ultimately positioned not as subject in an inter-subjective dialogue (as dialogic marketing was first conceived of by Kent and Taylor (1998)) but as an ‘inter-subject’ (or passive pipe) that enables Barbie’s ongoing dialogue with herself. The dialogic solicitation of personal information and the paucity of the subjectivities available to the child in the dialogue itself highlight the manipulative nature of big data surveillance of children and provide a new window into the ‘dark side of CRM’ (Nguyen et al., 2016).
Situating Barbie as a CRM project
Although there is no one accepted definition of CRM, marketing scholars universally describe it as a way to create and maintain relationships between companies and customers (Ngai, 2005: 583; O’Brien, 2011: 33), and typically give information technologies and communication a central place in how CRM is conceptualized. For example, Gronroos defines CRM as an information system that supports the creation of value through interaction and planned communications designed to strengthen the company–consumer relationship (cited in O’Brien, 2011: 33). Swift defines it as an ‘enterprise approach to understanding and influencing customer behaviour through meaningful communications in order to improve customer acquisition, customer retention, customer loyalty and customer profitability’ (cited in Ngai, 2005: 583). And Kincaid defines it as ‘the strategic use of information, processes, technology, and people to manage the customer’s relationship with your company’ (cited in Ngai, 2005: 583).
Prior to CRM’s development in the late 1980s, mass marketing methods only allowed companies to ‘broadcast’ unidirectional messages to predefined market segments of the population. Even when individual customers had formed emotional connections with their brands, companies like Mattel had no way to engage them in dialogue because customers were by definition passive message recipients who could not talk back (Beckett and Nayak, 2008: 304; O’Brien, 2011: 34). As a corrective, CRM focused on new ways to enable one-on-one, bidirectional dialogue to recast the passive consumer into an ‘active agent, keen to engage organizations in a dialogue about their needs, able to interpret their desires, and responsive to suggestions from their trusted trading partners’ (Beckett and Nayak, 2008: 300–301; see also O’Brien, 2011: 33).
Data mining technologies were crucial to this project, because they provided a way to easily access individual-level data when companies were communicating with individual customers (Shapiro et al., 2003: 88). However, the commercialization of the Internet significantly expanded CRM’s reach (Ngai, 2005) for two reasons. First, it enabled companies to collect vast amounts of the mundane data people generate as they surf. Second, it provided a number of tools, like email and instant messaging, that make it easier for companies to initiate dialogue with customers.
Mattel was an early pioneer in exploiting these new capabilities. Its website everythinggirl.com contained a variety of online games that let children play with Barbie-branded content; Mattel then collected data from the children as they played, and used that data to craft and deliver customized marketing messages. For example, the site encouraged children to create ‘wish lists’ of products they wanted to buy and then emailed the lists directly to the children’s parents (Steeves, 2006). Mattel’s early efforts also used voice technology to extend their marketing reach. Recorded messages on the site invited girls to phone Barbie so she could tell them about the ‘wonderful wish [I’ve got] for you. I’d love to call you and tell you! Or just say Hello’. Children could also phone Barbie so she could read them pre-recorded bedtime stories or sing them Happy Birthday (178).
Prior to the development of big data analytics, the success of CRM projects like everythinggirl.com was limited by the fact that the data mining technologies of the day simply could not make actionable sense of detailed data about millions of customers in an affordable way (Shapiro, 2003: 88). Although the development of the Web facilitated the collection of individual-level customer data and communication for CRM purposes (Ngai, 2005), companies like Mattel are now turning to forms of algorithmic analysis that can both solicit data and use that data to create technologically mediated one-on-one dialogue with large numbers of customers in real time to pursue their CRM goals (Lager, 2017).
Using dialogic analysis to ‘read’ Hello Barbie’s utterances
Hello Barbie provides a unique opportunity to examine how marketers hope to do this, because Mattel has published a complete catalogue of the algorithm’s dialogic output. 1 Much of the literature that has engaged with networked toys like Hello Barbie has focused on the privacy implications of this type of algorithmic processing (Montgomery et al., 2017; Smith and Shade, 2018; Steeves, 2006). However, I suggest that dialogic analysis of the company’s output provides an additional way of interrogating the commercial practices at the heart of big data, by opening a window into the ways in which the marketing context (and the marketer’s goals and expectations within that context) shape the ‘doing’ of marketing as a big data practice.
My goal is not to interrogate how the child responds to Hello Barbie as a dialogic partner or to evaluate the impact of Mattel’s marketing practices on child’s play. The meaning of any media artefact is actively negotiated by the people who interact with it (Hall, 1980). Children accordingly make their own meaning of the technologically mediated conversations they have with networked toys, and that meaning can disrupt or resist the creator’s message in unexpected ways (Holloway and Green, 2016; Marx and Steeves, 2010; Rooney, 2012). 2 My focus is instead on the commercial practices built into Hello Barbie as a CRM device. I suggest that analysing the doll’s output can shed light on how Mattel is seeking to achieve its marketing aspirations as they have been concretized in the conversational practices embedded in the doll. Analysing the output also provides a way to map the particular kinds of ‘consumer subjectivities’ (Bauman, 2007; Graham, 2018; McRobbie, 2008) that Mattel makes available to children and the kinds of relationships the company is seeking to enact with them through their dialogue with the doll.
Dialogic analysis is a qualitative method that allows researchers to interpret the meaning of utterances (Matusov et al., 2019). An utterance is defined as ‘any communicative act (spoken, written or gestured) which is both a response and an initiation’ (Gillespie and Cornish, 2014: 436; see also Bakhtin, 1986; Linell, 2009; Mead, 1922). The lines spoken by Hello Barbie are accordingly utterances as they are spoken to initiate dialogue and to allow that dialogue to continue by responding to the child’s speech in ways that enable the dialogic partners to jointly come to understand the conversation.
From this perspective, the meaning of the utterance is not found in the grammatical structure of the sentences or the dictionary meaning of the words. Rather, meaning is ‘found in the relation between the utterance and the broad context, including the participants’ (436). Meaning is accordingly both contextual, in the sense that the words must be placed in context for the utterance to take on meaning, and ‘addressive in the sense that it always implies an audience’ (436). Accordingly, Hello Barbie’s utterances should be viewed not as evidence of any objective truth about CRM. Rather the utterances are a ‘resource’ for better understanding how Mattel as a speaker constructs the meaning of the dialogue between the doll and the child through the words and phrases that it uses. ‘[T]he task of the researchers [using dialogic analysis] is to offer interpretations of that resource’ to identify assumptions and other unchallenged ways of thinking that are embedded in the practices that they bring into being (Beckett and Nayak, 2008: 301).
Because of the focus on dialogue, the coding team (which consisted of the author and four research assistants) first made a list of conversation topics that appear in the output. That list included Barbie’s personal history and inner life, fashion, daily activities (like paddle boarding and school), favourite things/likes, jobs, pets, special events/holidays, games/crafts and advice. The dataset was then coded by topic so all dialogic fragments available for a particular conversation could be bundled together.
After coding was completed, I applied Gillespie and Cornish’s (2014) interpretive device of asking sensitizing questions to understand the dialogic meaning of the utterances in each conversational bundle, as way of interpreting the data to produce novel, and useful, insights (449) into the dialogic meaning ascribed to the utterances by Mattel as the speaker. In particular, I asked:
Who is doing the talking? What is the speaker doing? … what prompted the utterance? … what is the speaker trying to set up? Who is being addressed? … What does the utterance assume about its audience? … how does the utterance position people? … what responses are enabled or constrained? (447) What kinds of subjectivities are anticipated by the company as speaker? Are these subjectivities constructed as active or passive? What kind of relationship between the company and the child is anticipated by the speaker?
3
The following sets out the findings of my analysis. Please note that the textual version of Barbie’s dialogic output includes written text and emotional cues or prompts which are capitalized and put in round parentheses. I have reproduced the fragments below as they were published by Mattel. In block quotes, each individual fragment begins on a new line. On occasion, I have added some information to link the fragment to the context in which it appears in the published output. This added information is included in square brackets.
Who is doing the talking: Barbie as active dialogic partner 4
As a conversational partner, Hello Barbie is programmed with an extensive back story which places the doll in a specific historical context. Indeed, much of what the doll tells children is in keeping with the Barbie storyline that Mattel created in 1959 when Barbie was first launched. For example: My full name is Barbara Millicent Roberts, but you can call me Barbie. My birthday is March 9th! I was born in Wisconsin, but I live in Malibu now. My mom’s name is Margaret, and my dad’s name is George. My dream house is in Malibu, California! I wouldn’t be surprised if Skipper, Chelsea and Stacie [three of Barbie’s seven siblings] are all hanging out together right now.
For example, the dataset includes frequent references to Barbie as a learner, who is especially interested in non-traditional fields for women like science: I’m a life long student. I just really like to learn new things. [Favourite subject at school] English, because I love to read… and Science, because I’ve always been really curious about how things work and what happens when you put different things together.
At the same time, Mattel’s ongoing attempts to ‘update tradition without completing discarding it’ continue to anchor the dialogue to the ‘neatly ordered patriarchal realms’ (12) that privilege Barbie as a fashion icon with traditionally feminine values. For example, even though Barbie says that she ‘was lucky enough to become an astronaut! Seeing the Earth from space gives you a whole new perspective!’, the dialogue also reinforces both the helping nature of Barbie’s work – ‘I think my favorite careers are the ones where I get to help people’ – and the importance of fashion and style: ‘I’m not a real princess, but when I imagine I’m a princess, I feel elegant and glamorous! I get to do awesome things to help people, plus wear lots of ball gowns – a double win!’. Accordingly by sending the message that, in the words of Mattel executive Chuck Scothon, any girl can ‘run for President and look good while she was doing it’ (cited at 11), Mattel is seeking to ‘balance nostalgic appeal with attracting new audiences’ (15) by both embracing feminism and rejecting it (see Crocker and Nevins, 2018). In this way, Mattel’s culture and history set the context for the dialogue.
Utterances about Barbie’s current-day activities – her nightly dreams, her favourite things to do (like paddle boarding) and her enjoyment of special events and holidays – are also contextualized by Hello Barbie’s position within the history of the broader Barbie franchise. Certainly, the tone and interaction cues that accompany the utterances are designed to give Barbie a personality that is consistent with the personality that she inhabits in the 37 movies, 19 video games, and multiple YouTube TV episodes and vlogs in which she ‘stars’. For example, no matter what season is selected by the algorithm as contextually relevant to the dialogue, Barbie is programmed to ‘sound’ enthusiastic about it: Snowmen and sledding and giant sweaters and … (STOPPING HERSELF BECAUSE SHE COULD GO ON FOREVER …) yeah. Winter’s pretty awesome. Flowers and sundresses and bare feet and… (STOPPING HERSELF BECAUSE SHE COULD GO ON FOREVER …) yeah. Spring is pretty great. Swimming and lemonade and giant sunglasses and… (STOPPING HERSELF BECAUSE SHE COULD GO ON FOREVER …) yeah. Summer’s pretty amazing. Apple picking and leaves changing and pumpkin pie making and … (STOPPING HERSELF BECAUSE SHE COULD GO ON FOREVER …) Fall is so awesome.
As noted above, Mattel has used this kind of computer-mediated interactivity on its website in the past to deepen a child’s relationship with the brand (Steeves, 2006). However, the dialogic nature of Hello Barbie – the fact that children can ‘speak’ to her and ‘be spoken to’ by her – provides new opportunities to reinforce the doll’s status as a person with her own subjectivity. Throughout the utterances, Barbie is presented as someone who is knowable, both by child – ‘(CUTE/WITH A WINK) You know me … I love trying out new careers!’ – and by itself – ‘I’m most myself when I’m around my friends, just playing games and imagining what we’ll be when we grow up!’. The spoken works and verbal cues combine in the dialogue to reinforce the doll’s status as a real person, and Barbie’s status as a dialogic partner is affirmed by her own sense of reflexivity. In Barbie’s words, ‘If I seem real, I am real, right? And what’s really real …?’.
Barbie is accordingly given a fully articulated subjective position in the dialogue as a speaker in her own right. Although she is acting as Mattel’s proxy in the conversation, the company’s presence is hidden behind Barbie’s fictional identity as an anthropomorphized object, which works to both facilitate and legitimize the interactions between the algorithm and the child (Fink, 2012). This positions Barbie to ‘act’ through the dialogue, to ‘do’ the work of CRM.
What is the speaker doing: Collecting data
Hello Barbie makes use of the doll’s status as a dialogic actor to solicit reams of information about the child and his/her family. Responses to questions about favourite subjects at school or animals and likes or dislikes are well-positioned to attract the kind of data that drives CRM because it helps the company come to know the child and her family members. For example, in one ‘game’, Barbie asks the child to describe which stores in an imaginary town his/her family members would own or operate: (IMAGINE PRIOR LINE IS “LET’S TALK ABOUT FAMILY”) … I’d love to learn more about you (BRAINSTORM) Oh, I know! Let’s make a game of it. The game’s called Family Town! We’re gonna pretend all of your family members run different shops in a make believe town! Okay, so every member of your family gets their own shop. One per person! I’ll visit each shop, and you’ll tell me who runs it! Got it? Hm? Which of your family members would run the movie theatre? Your stepmom? That’s cool! What kind of movies does she like to watch? Hm? Which of your family members would run the pet shop? Your dad? Awesome! Which animals are his favorite? Hm? Which of your family members would run the arcade? Cool! So your sister must really dig video games, right? Oh! You know what’s good? A dream diary! Have you ever written your dreams down in a notebook? Well, a dream diary is a little notebook that you keep next to your bed. And right when you wake up, you write down what you were dreaming before you forget! Isn’t that interesting? What’s the most interesting dream you’ve ever written about? Ooh! Tell me about it!
Who is being addressed: Child as inter-subject
This enhanced control works to constrain the child’s opportunities as a dialogic partner because it makes it more difficult to ‘challenge the inferences and predictions that are made by [the] algorithmic calculations’ (Lupton and Williamson, 2017: 786) that drive the dialogue. In other words, Mattel uses the algorithm to ensure that the corporation has a privileged place in the dialogue as a one-sided conversationalist who controls the flow of the conversation. Although the algorithm asks for input from the child (especially with respect to simple questions that have relatively simple answers like ‘what’s your favourite food/movie/colour’), the algorithm is designed to serve back responses that focus the ongoing conversation on Barbie’s stories, thoughts and pre-programmed games. In this way, the child’s dialogic input is slotted into predetermined story lines, with no impact on the direction of the dialogue or its meaning. If a child steps outside of the prescribed limits, the conversation quickly becomes nonsensical (e.g, see YouTube, 2015); the child is only given the opportunity to have a conversation with meaning if they follow the script. In this way, Mattel seeks a heightened level of control over ‘the constitutive power’ of the dialogue as a form of social interaction (Gillespie and Cornish, 2014:).
The following dialogue fragment illustrates how this works. The doll initiates a conversation by saying: Hey! It’s so good to see you! I had a stand up paddle board lesson today and I can’t wait to tell you about it! Have you ever gone paddle boarding before? Isn’t it so fun?? Well, it’s not for everyone, but I really like it Oh! You should try it some time! Anyway, today, while we were in the water, some baby dolphins came RIGHT up to us! They were SO close! Have you ever seen a dolphin up close before? They’re so beautiful! AND they’re SUPER smart. It was really amazing. So what about you? What was the most exciting part of YOUR day? What about you… what was the most amazing part of YOUR day? How was your day today?
Barbie’s position as the driver of the dialogue is also reflected in her discussion of her internal thoughts/memories, such as: (NON DESCRIPT HUMMING TO HERSELF, SEE SONG AT END) Hmm, hmm, hmm …. (STARTLED) Oh Hi! (LAUGHS) (LAUGHING) Sorry, I was just trying to remember this song … (HUMMING AGAIN) hmm, hmm, hmm … My dream was pretty lovely! (WISTFULLY) It was a beautiful day, I was soaring high in the air above Malibu … I could see my house, the ocean … doesn’t that sound peaceful?
This raises important questions about the potential impact of algorithmic dialogue nested within a CRM project that is masked as an empowering toy. Unless the child transgresses the scenarios the doll offers, he/she will be unlikely to steer his/her interactions with the doll or engage in imaginative play (Moller, 2015) precisely because transgressive dialogue works against Mattel’s expectations of the doll as a CRM project.
The child’s status as inter-subject is also revealed in Hello Barbie’s utterances on religion. The doll is programmed to talk about Diwali, Thanksgiving, Kwanzaa, Halloween, Christmas and Hanukkah. However, with respect to Diwali, Kwanzaa and Hanukkah, the comments are pedantic and focus on explaining the cultural or spiritual meaning of the festival to the child. For example: (LIKE YOU MIGHT SHOUT “SURPRISE!”) HAPPY DIWALI! Diwali is a five day Indian festival that happens at the same time as the Hindu New Year … so many reasons to celebrate, right? Today’s Kwanzaa principle is Umoja (oo MO jah) It means “unity”! Oh ok! Would you like to learn about it? Great! No worries! Sorry, I didn’t catch that. Was that a yes or a no to talking about Kwanzaa?
What kind of relationship is anticipated? The limits of CRM as dialogue
CRM theory posits that dialogic communication creates affordances for marketers because any feedback on the part of the customer can be seen from ‘a behavioral confirmation perspective, in which an individual’s impression of a communication partner shapes his behavior towards that partner, which in turn influences that partner’s attitude and behavior toward him’ (Pang et al., 2018: 72). Pang et al. (2018) suggest that the company’s capacity to meet its CRM needs is amplified in computer-mediated dialogic communication precisely because the dialogue is ‘asynchronous’; the technical lag means that ‘practitioners can craft messages carefully in a process of writing and editing before sending’ (72). This asynchronicity is key to understanding Hello Barbie as a big data device; in spite of the rhetoric of instantaneity and personalization, in which the technology is positioned as an immediate response to a child’s imagination, Mattel’s dialogic communication is used to fit the child’s responses within predetermined consumer subjectivities because they advance the corporation’s interests (Punj and Stewart, 1983). The fact that Barbie’s utterances have been carefully crafted in advance for a marketing purpose is hidden behind the speed at which the algorithm can analyse the child’s voice patterns and select a (carefully crafted and predetermined) response.
Accordingly, although the purported goal of CRM is to collect data to build bidirectional relationships where each party in the relationship gets what it wants (O’Brien, 2011), the unidirectional design of the dialogue ensures that the child is continuously steered to topics of Mattel’s choosing. For example, the dataset is replete with fragments that prompt the child into pre-established dialogic interactions, like talking, playing one of Barbie’s games or doing an activity (like making a card for a parent). Although Barbie will take no for an answer, there are also a number of utterances to wheedle the child to encourage him/her to follow the doll’s lead: Aw, come on … give it a try! I promise it’s fun! Ah pleeeease? It’ll be fun, I promise! C’mon … you want to know why? Huh?
This strategy is used both to encourage the child to disclose information and to deliver brand-friendly messages that encourage consumption. The latter occurs in one of two ways. First, many of the pre-established storylines are built around allusions to specific Mattel products. For example, a significant portion of the dialogue, like the following fragment, is devoted to animals, caring for animals and being a veterinarian: Ooh! What about becoming a veterinarian. How about that? A veterinarian is someone who heals sick animals. A vet. Is that something you might like to do? You said you wanted to be a veterinarian when you grow up; why don’t we talk about animals! (GIDDY/EVOKING FUNNY MEMORY) Remember when we played that veterinarian game? And you cured the imaginary puppy?
Second, the utterances work to both reinforce the brand identity and encourage the child to adopt that identity for his/her own. The most obvious examples of this are the multiple prompts to re-direct the conversation to fashion: So now that we’ve been used our imagination and played games, (MOCK SERIOUS) let’s get serious and talk about something really important… FASHION! We’ve been talking a bit about school, why don’t we try on something else for size… (LAUGHS) let’s chat about fashion! One of my favorite things to do with my friends is talk about fashion! With three sisters in my family, we spend a lot of time talking about clothes and fashion! I remember you were interested in fashion design, wanna talk about fashion now? Red’s a rad color! What do you like about it? Orange is outstanding! Why do you like it? I say yes to yellow! What do you like about it? Oh, green is great! What do you like about it? (DECLARATIVE) I love pink! Pink is positively perfect! And it’s my favorite too, by the way. Why do you like pink? I like pink. You’re speaking my language! I love any shade of pink for my fingers and toes! (SERIOUS) You won’t believe this, but I usually get pink … (LAUGHS) Okay, maybe that’s obvious. It’s just my favorite! My grandma always chooses pink! Hm … Kinda like her granddaughter! Well, tell them to get that pink polish ready for me! Also a pink paint job [for a car] doesn’t hurt. Our favorite color! I still think it’s so awesome that we both love pink, by the way. Pretty in pink! What? Who knew you were such a punk rocker? I went through a punk phase myself. You know, pink hair, studded bracelets…. ahh. I miss the 80s. I love hanging out with you! (SIGH) (IN A BURST) I’m so happy you’re here because you’re smart and funny and you’re super nice and you’re just the best! You know it’s all true, right? I ALSO just learned what the word “Hanukkah” means in Hebrew. Do you know? It means “dedication”. (THOUGHTFUL) I love that. It makes me think of all the things that I try to really devote myself to like my friendship with you! Well, if you ever need someone else to cheer for you, you know who to call… Me! (GENTLY) I’m so sorry to hear that. (ENCOURAGING) I think you’re exceptionally beautiful. And I’m so thankful to have you as my friend! (GRACIOUSLY, GENEROUSLY) I have to say, my good thing is talking with you! Now my day’s gotten a million times better! We could make a special card for your Dads to show them how much you love them. Let’s think of something now. How would you tell someone that you love them? Maybe you should try something like “I love you so much.” “How does that sound?”. Oh, you think? Well, anything can be hard when you first start learning it. But it’s the best feeling in the world to finally master a tricky problem. There are a lot of things I thought were hard at the beginning. But I love that feeling of accomplishment after working hard to learn new skills!
Conclusion
Hello Barbie is the latest in a line of Mattel products that seek to use the interactivity of networked technologies for CRM purposes. My analysis suggests that Hello Barbie’s output is designed to use computer-mediated dialogue to collect actionable data that can then be algorithmically analysed to fuel a hyperpersonalization (Pang et al., 2018) that can deepen a child’s emotive connections to the brand. However, it is the doll, and not the child, who is hyperpersonalized and able to fully inhabit the role of dialogic actor. The high degree of control the doll then exercises over the dialogue detracts from the subjectivities available to the child, who can disrupt the planned conversation unless he or she is nudged into pre-defined storylines. The child is accordingly reduced to a list of likes/dislikes and is unable to influence the dialogue in a meaningful way. If the child does transgress the given storylines to take an active role, the dialogue itself becomes meaningless and the dialogic interaction fails. This perhaps explains why sales of the doll remain well below what Mattel projected and its intended rollout outside the US was cancelled within 12 months of the doll’s launch. 7
Commercial success or not, Hello Barbie provides a new window into the ‘dark side of CRM’, in three ways. First, it is an excellent exemplar of the ‘potential costs and ramifications of poor, mismanaged or dysfunctional relationships’ (Funches, 2016: 74) created through big data CRM practices. Second, it illustrates the limits of consumer subjectivities that are conceptualized solely ‘through the notion of needs and the relevance of those needs to the producer’ (Beckett and Nayak, 2008: 305). Third, the seamless collection of the child’s data and the level of control exercised over the doll’s dialogue raise concerns about unfair or deceptive marketing practices. As N’Goala (2016) argues, big data marketing solutions tend to make the customer more transparent to companies, creating an information asymmetry that tips the power scales in favour of the company. ‘The temptation for firms to exploit this asymmetry and to behave opportunistically is, therefore, very high’ (123). The appearance of manipulation, i.e. uses that ‘undermin[e] what the marketer believes is his/her audience’s normal decision making process … by playing on a vulnerability’ (131), is even more marked when companies are interacting with populations like children who are vulnerable by definition (137).
Certainly, significant concerns were raised about Hello Barbie as soon as the doll was released. The Campaign for a Commercial Free Childhood initiated the Hell No Barbie campaign, attacking the doll because of its ‘creepy’ information practices. Executive Director Susan Linn argued that, ‘Kids using “Hello Barbie” aren’t only talking to a doll, they are talking directly to a toy conglomerate whose only interest in them is financial’ (Lobosco, 2015). From the company’s perspective, this is precisely the intent of Barbie as a CRM device; however, the marketing goals embedded in the doll are precisely what parents were most concerned about.
Dialogic analysis is accordingly helpful in unpacking and critically evaluating the marketing fantasy of using big data to create ‘precise forms of advertising that cannot be resisted’ (Hilu, 2016: 2). In this way, it can contribute to increasing our ‘understandings of how new technologies impact consumption markets and cultures [by] scrutinizing the meaning-making capacities of cultural practices of representation’ (16) that are contained within our interactive devices.
Footnotes
Acknowledgement
The author would like to thank the peer reviewers for their insightful comments on earlier drafts of this article, and the Social Sciences and Humanities Research Council of Canada for their support.
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
