Methodological proposals for the study of consumer experience

From dyads to ecosystems

Where an experience is evaluated holistically, it is reasonable to suppose that more than a single company will be involved (Addis and Holbrook, 2001). Tax et al. (2013, p. 457) proposed the idea of the service delivery network (SDN) as “two or more entities that, in the eyes of the customer, are responsible for the provision of a connected, overall service”. In their study of gap year travel, Baron and Harris (2010) expanded the domain of experience to include a network of organizations such as universities, accommodation, banks and tour companies, all of which form part of the experience.

Proposing a systemic perspective on consumer experience, Jaakkola et al. (2015) and McColl-Kennedy et al. (2015b) suggested looking beyond the dyadic interactions between provider and consumer to study the interactions between the consumer and multiple actors. McColl-Kennedy et al. (2015a, p. 251), for instance, highlighted the role of interactions with other actors (including “other customers, organizations, friends and family”) during the consumer experience. The S-D logic literature in particular has seen conceptual developments in relation to service systems and experience. For example, Akaka and Vargo (2015) showed how the context of experience has evolved from service encounters (i.e. dyadic interactions between customer and frontline employees) to ecosystems (defined as systems of resource-integrating actors connected through institutions and institutional arrangements) (Vargo and Lusch, 2016) at micro, meso and macro contextual levels (Akaka et al., 2015). The micro contextual level includes the consumer-company dyad whereas the meso contextual level includes those actors’ networks. The macro contextual level encompasses the shared networks and institutions that coordinate the interactions between actors (Akaka et al., 2015; Vargo and Lusch, 2016). Although some empirical studies have identified other actors as elements of the customer experience (Tax et al., 2013), many knowledge gaps remain regarding the influence of these actors and interactions with them. In addition to methods for identifying other actors, researchers need methods for investigating how the focal consumer’s experience is affected by them, their resources and the institutions that coordinate the interactions.

From provider-centric to consumer-centric journeys

As noted earlier, the main body of the consumer experience literature examines the customer journey, comprising points of direct and indirect interaction with the company (known as “touchpoints”). Although some researchers acknowledge that certain touchpoints may not be under the company’s control (Lemon and Verhoef, 2016), the journey remains provider-centric in the sense that it is centered on an object of consumption (buying and using a product). According to McColl-Kennedy et al. (2015b), this view is no longer adequate.

Recent studies have highlighted the importance of understanding the customer’s processes and activities. For example, according to Heinonen et al. (2010), the customer experience is actively formed by the customer, who selects companies and partners to integrate into their processes. As the customer experience is part of the customer’s ongoing life, Heinonen et al. (2010) claimed that service providers should offer value propositions that adapt to customers’ processes and to activities that may not always be immediately apparent. As noted earlier, this view of experience as process means that the customer is in control and chooses actors that will help them to achieve their goals. However, despite proposals to adopt a more consumer-centric approach to journeys, most such studies are largely conceptual (e.g., Heinonen et al., 2010).

First-hand description of experience

A phenomenological characterization of experience informs S-D logic (Vargo and Lusch, 2016), the Nordic school (Helkkula and Kelleher, 2010) and the original propositions of the experiential view (Holbrook and Hirschmann, 1982). Phenomenological research explores experiences from the participant’s (first-person) perspective as it emerges in context (Thompson et al., 1989). Shifting the scope of consumer experience from a focus on the provider to a consumer-centric perspective requires an equivalent shift in methods.

When consumers describe their experiences in their own words, they foreground events, activities and interactions with actors that seem most important from their point of view. By implication, these elements need not necessarily be linked to a single service provider. For example, Mickelsson (2013) identified a series of activities that formed part of the gambling experience (reading blogs and magazines, talking to friends) but were not connected to the gambling company. As investigation of the consumer experience must allow consumers to express themselves freely, surveys and interviews concerning a specific object of consumption are inadequate. Instead, truly consumer-centric methods must allow different actors, activities and events to emerge naturally in the consumer’s description.

Description of multiple relevant actors and institutions in the consumer’s ecosystem

The systemic nature of the experience also has methodological implications. Specifically, the method must enable the researcher to identify the actors, resources and institutions that form the consumer’s ecosystem at the micro, meso and macro contextual levels (Akaka et al., 2015). In many cases, the methods used in studies of consumer experience do not allow the respondent to refer to all relevant actors and influences; for example, interviews and surveys are confined to the dyadic relationship between company and consumer and questions about single service providers cannot adequately capture other relevant actors and institutions in the consumer’s ecosystem.

When collecting and analyzing data, the researcher must keep in mind that a systems perspective makes no distinction between producer and consumer (Vargo and Lusch, 2016) – in other words, every actor can supply resources to the consumer (who may also assume the role of producer) and so contribute to the consumer experience (Vargo and Lusch, 2016). Research methods must therefore enable the researcher to identify those actors, as well as the resources and mechanisms through which they influence the experience. In addition, the method must allow institutions and institutional arrangements to emerge in the consumer’s descriptions. Importantly, combining this requirement with first-hand description of the experience means that the ecosystem will be described from the consumer’s perspective.

Capture of processual nature of consumer experience

Consumer-centric and systemic perspectives further imply that experience has a processual nature that must also be captured. Interpreting the consumer experience as a process – seen here as a sequence of events over time – entails a number of assumptions (Pettigrew, 1990). First, any changes that occur during the consumer experience should be studied in their proper context at the micro, meso and macro levels of analysis proposed by Akaka et al. (2015). A second assumption relates to temporal interconnectedness: that past experiences shape present and future experiences, encompassing both the chronology of events and their underlying logics and structures. Jaakkola et al. (2015) highlighted the importance of these temporal patterns in understanding the consumer experience. A third assumption is that processes shape and are shaped by their context – that is, consumer experiences are shaped by contextual factors, but they can also change that context as they develop. Finally, explanations of change are not linear or causal but holistic and multifaceted (Pettigrew, 1990) – in other words, changes have multiple causes at different levels of analysis.

Because current methods such as cross-sectional surveys and interviews cannot adequately capture how an experience evolves over time, Jaakkola et al. (2015) and McColl-Kennedy et al. (2015b) have suggested the use of longitudinal research designs. However, because many process researchers use a combination of retrospective and real-time data collection methods, a study need not necessarily be longitudinal to capture the processual dimension (Bizzi and Langley, 2012; Halinen et al., 2013); the most important requirement is to capture the sequence of events over time and in context.

Phenomenological interviews

The first of the three selected methods is the phenomenological interview. Although Thompson et al. (1989) suggested the use of this method to study consumer experiences, only one of the 104 analyzed articles (Major and McLeay, 2013) explicitly mentioned its use as the main method of data collection. This suggests that phenomenological interviews have unused potential in this context. The phenomenological interview can be defined as an unstructured, open-ended, dialogical interview informed by the philosophical assumptions of phenomenology (Bevan, 2014; Stern et al., 1998; Thompson et al., 1989); its purpose is to obtain a first-hand, free-form description of a domain of experience, contextualized in the consumer’s lifeworld (Bevan, 2014; Stern et al., 1998; Thompson et al., 1989).

One broadly similar method is the narrative interview; both are phenomenological in nature and elicit descriptions of participants’ stories, so satisfying the first requirement. The key difference is that while the narrative interview is more closely related to story-telling, beginning from one main question (Roederer, 2012) with specific supplementary questions only when necessary (Juntunen, 2014), the phenomenological interview is more conversational.

Another data collection method that resembles the phenomenological interview in some respects is the unstructured interview. However, while it is safe to say that every phenomenological interview follows an unstructured script, not every unstructured interview is phenomenological at the level of philosophical assumptions. For instance, in a phenomenological interview, the participant must have lived the experience (Thompson et al., 1989), and the focus is on the contextualized description of that experience rather than rationalizations (Thompson et al., 1989), which may not be the case in an unstructured interview.

Guidelines for using phenomenological interviews

Because the phenomenological orientation is not homogeneous, the guidelines for this method follow general commitments within this tradition (Høffding and Martiny, 2016), which can be discussed under three headings:

1. selection of participants;

2. primacy of subjective experience; and

3. contextual factors.

First, because the goal of this type of interview is to obtain a first-hand description of a domain of experience (Thompson et al., 1989), interviewees must have lived the phenomenon under investigation (Thompson et al., 1989). For example, to investigate how a consumer experiences a journey to a healthier lifestyle, the researcher must select participants who have undergone or are undergoing this journey, as this philosophical position takes the subjective experience to be the reality (Thompson et al., 1989).

Second, the phenomenological tradition asserts the primacy of subjective experience over theoretical assumptions (Horrigan-Kelly et al., 2016). The researcher must therefore take care to set aside any theoretical assumptions, hypotheses or frameworks when formulating the interview script (Thompson et al., 1989) or at least be reflexive about them, assigning primacy to the subjective experience (Horrigan-Kelly et al., 2016). Either way, the researcher must remain open to emerging themes (Horrigan-Kelly et al., 2016; Thompson et al., 1989). Consequently, the structure of the interview is very open and is dictated by the respondent (Thompson et al., 1989); the interviewer’s role is to ask open and unstructured questions, resulting in a conversation rather than the questions and answers typical of an interview (Stern et al., 1998; Thompson et al., 1989). This conversational approach can be achieved if the researcher really listens to the participant, reacting to their descriptions and adding or deleting questions as the conversation proceeds, without being restricted by the interview script or prior assumptions.

Finally, because the goal of the interview is to obtain a first-hand description of an experience in context (Bevan, 2014; Stern et al., 1998; Thompson et al., 1989), the researcher must strive to understand what the experience means in the consumer’s lifeworld. Encouraging detailed description of the experience typically foregrounds these contextual factors during the interview.

Granted the need for an unstructured approach, Bevan (2014) developed a method that maintained the phenomenological interview’s conversational style while organizing it in three parts: contextualization, apprehending the phenomenon and clarifying the phenomenon. Contextualization means taking account of the context in which the experience gains its meaning (Bevan, 2014). The interviewer’s role is to set a meaningful context within which the respondent can describe their experience (Thompson et al., 1989), and this should provide a point of departure for the interview (Bevan, 2014). For instance, in relation to a journey to a healthier lifestyle, the researcher might begin by asking about the participant’s background and any unhealthy habits they previously had.

Apprehending the phenomenon means focusing on the experience of interest to obtain a clear description of events and activities. This can be achieved by asking descriptive questions, followed by structural questions to clarify the phenomenon (“Could you please describe what you mean by […] ?”) (Bevan, 2014). Why questions should be avoided because they change the focus from description to rationalization, which is not the goal of the phenomenological interview (Thompson et al., 1989). In addition, the interviewer should not ask questions that are too abstract, focusing instead on specific events to elicit rich descriptions of experiences (Thompson et al., 1989).

Clarifying the phenomenon involves imaginative variation, adding questions that vary in structure as a means of clarifying the phenomenon (Bevan, 2014). During the interview, the interviewer should formulate questions based on the consumer’s reflections, asking for further descriptions (Bevan, 2014; Stern et al., 1998). For example, if the interviewee reports that it is important for them to go to the gym with a partner, the researcher might ask them to imagine what would happen if they went alone.

Another important point here is that interviewer and interviewee are equal in status; the interviewee is the expert in relation to their own experience, and the researcher should not try to direct the interview or impose theoretical assumptions. If this happens, the data should be excluded (Thompson et al., 1989). For that reason, data analysis is as important as the interview itself when using a phenomenological approach. Following the first interview and before proceeding to another, researchers should listen to the interview and transcribe it to check whether they are adopting a conversational approach or imposing their own theoretical assumptions.

Satisfying the methodological requirements

Given that the phenomenological interview must allow themes to emerge, assigning primacy to participants’ subjective experience rather than to theoretical assumptions (Horrigan-Kelly et al., 2016; Thompson et al., 1989), this method fully satisfies the first requirement: a first-hand description of experience. Furthermore, according to Stern et al. (1998, p. 198), the phenomenological interview produces a “narrative unit encompassing past, present, and future”. By analyzing these narratives, the researcher can better understand both the consumer’s experience and their relationships with others (Stern et al., 1998), shedding light on the systemic structure of the consumer experience. It is assumed that relevant institutions may also emerge in the descriptions of interactions with other actors. Although the interview should follow a largely unstructured script, the researcher can elicit descriptions of any institutions that seem relevant for the consumer. This fulfills the second requirement: the description of multiple relevant actors and institutions in the consumer’s ecosystem. Finally, because a narrative can be viewed as a process, this method can capture the processual dimension of consumer experience.

Event-based approaches

The second proposed method of data collection is the event-based approach, referring here to methods that focus on events or incidents as a means of understanding the consumer experience. Interviews are most often used to collect data (Åkesson et al., 2014; Helkkula and Pihlström, 2010); of the 104 articles analyzed here, two used sequential incident technique (Stein and Ramaseshan, 2016); one used critical incident technique (CIT) (Grove and Fisk, 1997), and one used an experience-based event technique (Åkesson et al., 2014). However, these articles explored events or incidents that related to an object of consumption (the journey with a company or the use of a service) rather than the journey the consumer undertakes to achieve a goal (i.e. the consumer journey), as it is proposed here.

Guidelines for using event-based approaches

The guidelines for using these approaches can be divided into three major groups:

1. delimitation of events;

2. temporal aspects; and

3. elicitation of events during the interview.

To begin, the researcher must define the central process under investigation (Halinen et al., 2013) to determine what events to look for (van de Ven, 1992). Generally speaking, events can be defined as outcomes of human action or changes caused by nature (Hedaa and Törnroos, 2008). To study the consumer experience as defined here, it is suggested that the researcher should focus on events or incidents that relate to the consumer’s pursuit of their goal in their lifeworld – in other words, the consumer journey is the process. Revisiting the example of a journey to a healthier lifestyle, the event might be defined in terms of its contribution to the goal, allowing the participant to describe the most relevant events from their own perspective. In contrast to existing studies that focus on events associated with using a specific service (Åkesson et al., 2014), this makes it possible to obtain a description of events that are not necessarily related to a single company but include other actors and contexts.

Second, as event-based methods require in-depth description and the participant needs to be able to recall the event, researchers should recruit participants who have lived the experience recently to guard against faulty memory and rationalization effects (Flanagan, 1954). To this end, the researcher might, for example, establish a temporal limit or recruit participants currently living the experience under investigation. Ideally, a combination of real-time and retrospective data collection methods could be used to capture events over time (Bizzi and Langley, 2012; Halinen et al., 2013). Scheduling follow-up interviews with participants currently living the experience would facilitate collection of real-time data, maximizing the potential of this method to capture the processual dimension of the consumer experience (Bizzi and Langley, 2012).

Finally, researchers should elicit detailed descriptions of the events. Based on an extensive review of the literature on CIT, Butterfield et al. (2005) noted that the method has evolved to illuminate the context and to capture the meaning of events, in line with a contextualist phenomenological approach. By connecting events and incidents to the consumer’s goal, participants can describe events in relation to multiple actors and different contextual levels. To that end, researchers should explore events in depth during the interview, with an emphasis on context (Hedaa and Törnroos, 2008; van de Ven, 1992). For instance, Gremler (2004) suggested that researchers should look beyond cognition, collecting data on emotional aspects of the events and exploring why those specific events occurred. By remaining flexible when collecting data on events, further events can also be explored (Halinen et al., 2013).

Satisfying the methodological requirements

According to Gremler (2004), event-based approaches involve the collection of descriptions of events from the participants’ perspective without reference to prior hypotheses. This method therefore satisfies the first requirement: obtaining a first-hand description of the experience. In addition, the event-based method can satisfy the second criterion: description of multiple relevant actors and institutions in the consumer’s ecosystem. According to Hedaa and Törnroos (2008), a contextualized description of events allows the description of actors and their subsequent reactions to events (i.e. acts and activities). Process researchers have proposed guidelines for using event-based approaches to identify how a phenomenon develops in a multi-layered context at micro, meso and macro contextual levels (Halinen et al., 2013; Makkonen et al., 2012). However, although the participant can be expected to describe interactions with a multiplicity of actors, the researcher must keep in mind that this technique may not provide a holistic perspective; for example, it may be challenging to obtain descriptions of institutional logics.

When describing events in sequence, the event-based approach helps the researcher to understand the processual and dynamic aspects of the consumer experience. As noted before, process researchers define process in terms of events (Pettigrew, 1990), which satisfies the third requirement for investigating the broader aspects of consumer experience. Using multiple event-based interviews over time maximizes this method’s capacity to fulfill the third requirement (although this is not essential). Finally, an additional advantage of event-based approaches is that the researcher can address relevant issues by focusing exclusively on specific events (Åkesson et al., 2014), saving time during interviews and when analyzing the data (Helkkula and Pihlström, 2010).

Diary methods

The third proposed data collection method is the diary. Carú and Cova (2008) suggest the use of personal introspection to study the consumer experience, where consumers analyze and report their individual experiences. To date, however, the diary method has attracted little interest in the marketing literature (Patterson, 2005); of the 104 analyzed articles, none used a longitudinal diary as the main data collection technique. In general, participants are invited to write about their experiences at a point in time. For instance, Ryynänen et al. (2016) collected data online by asking participants in a data collection community to write about a personally meaningful consumer experience involving packaging.

In a non-research context, a diary can be defined as a personal record of events, thoughts and observations (Patterson, 2005); in a research context, it is an instrument for self-reporting events, thoughts and observations (i.e. ongoing experiences) (Bolger et al., 2003). In research terms, the record of these ongoing experiences constitutes a narrative (Livholts and Tamboukou, 2015), providing an account of a sequence of significant events (Kenten, 2010). Although Hurmerinta and Paavilainen-Mäntymäki (2013) contended that this method is closely linked to ethnography, the diary is discussed here as a data collection method rather than as a methodological approach.

Guidelines for using diary methods

The guidelines for using diary methods can be divided into four parts:

1. choosing the diary design,

2. choosing the means of reporting,

3. giving instructions; and

4. maintaining the participant’s motivation.

First, the researcher has to select the appropriate type of diary for collecting the data in light of the research goals, choosing between solicited (Kenten, 2010) and unsolicited diaries (Livholts and Tamboukou, 2015). Solicited diaries ask participants to report their experiences for the purposes of the research whereas unsolicited diaries are not produced in the research context. Researchers can make use of several forms of online diary or other tools for personal introspection available online – for example, Helkkula and Kelleher (2010) studied narratives found in blogs. While this approach can provide more naturalistic data, it is important to keep in mind that the participant may be presenting an ideal self online (Zhao et al., 2008) (although, of course, this may also arise in the case of non-naturalistic data). Granted the benefits of the unsolicited approach, the following guidelines apply to solicited diaries.

There are at least three types of diary designs: interval-, signal- and event-contingent (Bolger et al., 2003). The interval-contingent design requires participants to report in their diaries at specified regular intervals. In the signal-contingent design, researchers emit a signal that prompts participants to report in their diaries, and the event-contingent design requires participants to report in their diaries on each occurrence of the event being studied.

In addition to choosing a diary design, the researcher must choose the most appropriate means of reporting the experiences. Participants may not be used to writing every day or may find this inconvenient. As the introspective narrative can take the form of text diary, audio diary or video diary (Bolger et al., 2003; Carú and Cova, 2008), researchers can use devices such as audio recorders to capture “real-time” descriptions (Bolger et al., 2003; Hurmerinta and Paavilainen-Mäntymäki, 2013).

As a third guideline, it is important to give clear instructions to participants (Carú and Cova, 2008; Kenten, 2010); for instance, the event-contingent design requires researchers to provide a clear definition of the focal event (Bolger et al., 2003). On the downside, Patterson (2005) noted that instructions that are too restrictive can stifle descriptions and cause participants to lose interest in reporting their experiences. Hurmerinta and Paavilainen-Mäntymäki (2013, p. 201) provided an example of good instructions, asking entrepreneurs to “write freely, either in the form of a diary or as short notes or log entries, about how you search, create, observe, analyze and exploit business and internationalization opportunities in practice, as part of your daily business activities and routines”. These instructions are clear without being so specific as to limit the participant’s motivation and creativity, and they include a clear definition of the focal event.

Finally, when using this method, it can be challenging to maintain participant motivation. To address this issue, as well as ensuring that instructions are not too restrictive, researchers should meet participants and provide feedback about diary completion (Hurmerinta and Paavilainen-Mäntymäki, 2013). Participants in diary studies have provided positive feedback on the method, noting in particular that they were more aware of their own thoughts, feelings and behaviors (Hurmerinta and Paavilainen-Mäntymäki, 2013; Kenten, 2010), and this could be used to encourage diary completion. To motivate more frequent reporting, it may also help to use a means of delivery that is familiar to the participant (e.g., an audio device).

Satisfying the methodological requirements

The diary method is useful when collecting data on everyday experiences (Kenten, 2010) and has several strengths. First, it does not impose any theoretical framework on participants or on the data they provide (Hurmerinta and Paavilainen-Mäntymäki, 2013). Because it is also phenomenological in nature (Bolger et al., 2003), this method strongly fulfills the first-hand description requirement. However, researchers must ensure that their instructions do not impose theoretical assumptions or expectations that might lead the participant. Second, the diary is an appropriate means of studying multi-layered and multi-actor processes (Makkonen et al., 2012). By allowing the description of various actors and institutions to emerge, it satisfies the second requirement. Finally, as compared to retrospective data, the diary reduces the risk of memory error (Bolger et al., 2003; Kenten, 2010). It also provides an excellent means of investigating temporal dynamics and processes in real time (Bolger et al., 2003), making it the most powerful method here in meeting the third requirement.