How to Measure Customers’ Journey?
by Christin Beer & Florian Stahl
Lower prices and novel products or services are said to be no longer sufficient for firms to sustain their competitive edge. Instead, the creation of superior customer experience plays an increasingly crucial role in setting the business apart (Grewal, Levy, and Kumar 2009). Thus, understanding and managing customer experience has become one of the top priorities of both practitioners and scholars in the last decade. Essential prerequisite for this is a deep knowledge of the customer journey (Berry, Carbone, and Haeckel 2002). Advancements in digital and mobile technologies and the resulting emergence of big data enable the obtainment of granular insights into the behavior of individuals (Chen, Chiang, and Storey 2012). On the other hand, this technological proliferation has evoked a complex environment consisting of a plethora of channels, media, and touch points (Day 2011), making it difficult to seamlessly trace respective movements, especially in cross-channel and offline settings (e.g. Ailawadi and Farris 2017). Although many reports (e.g. Gartner 2018) show that the majority of firms are interested in customer experience and customer journey tracking, other sources also emphasize that most of these companies have not yet managed to transform their aspirations into tangible initiatives (e.g. Forrester 2018). Yet, fractions of the customer journey are generally already assessable. The key to such insights often lies in data sets that are readily at management’s disposal. By leveraging and sharing their firm-specific information as well as their key performance indicators (KPIs) from the domains of digital marketing and customer relationship management across company departments, managers could track some of their customers’ footprints, thereby laying the foundation for actively enhancing customer experience.
But what do the terms ‘customer experience’ and ‘customer journey’ actually mean?
Though ‘customer journey’ has only recently emerged as buzzword, the underlying idea can be traced back to Howard and Sheth (1969) who introduced the concept of purchase stages:
- Pre-purchase i.e. need recognition, search, consideration
- Post-purchase i.e. consumption resp. usage, service requests, and engagement
While a notable fraction of scholars asserts that – due to today’s abundance of channels, media, and touch points – an individual’s behavior may not always be as linear as implied by this structured model (e.g. Baker 2016), it is still accepted as fundamental conceptualization of the customer journey. Also, it is widely acknowledged that the customer journey encompasses a variety of touch points across several channels. Thereby, a channel is “a customer contact point, or a medium through which the firm and the customer interact” (Neslin et al. 2006), whereas touch points are “episode[s] of direct or indirect contact with a brand or firm” (Verhoef, Kannan, and Inman 2015) with channels providing the environment in which these episodes occur. Eventually, the customer experience accrues along these touch points and is often described as multi-dimensional construct. For example, Lemon and Verhoef (2016) define the term as “cognitive, emotional, behavioral, sensorial, and social responses to a firm’s offerings during the customer’s entire purchase journey”. The following figure schematically illustrates the fundamental conceptualization of the customer journey and customer experience.
How can managers derive insights about customers’ experience and customers’ journey from primary data?
Current research illustrates that customer experience can be measured via well-established customer feedback metrics that are traditionally collected by means of after-sales surveys, i.e. primary data. In general, academia seems to agree that, though they do not capture the multidimensionality of customer experience, such KPIs serve as modest approximates of experience-related customer perceptions (e.g. Klaus and Maklan 2013). Hereby, customer experience is most commonly approached based on the evaluation of the Customer Satisfaction Score (CSAT). Other authors also consider the Net Promoter Score (NPS) and the Customer Effort Score (CES) as well as (re)purchase intentions as insightful. Yet, researchers perceive such metrics to be too holistic (Meyer and Schwager 2007) or to solely tackle “the sale or after-sale service situation” (Frow and Payne 2007), claiming that they should be assessed more regularly along the customer journey to examine the overall and touch point specific experience. Thus, marketers consulting such KPIs to estimate their customers’ experience should look at different aggregation levels and also combine metrics to generate the most reliable feedback. Although such primary data ergo do entail information on the customer experience, they are said to be “notoriously unreliable [and] skewed by low participation rates, biased questions, and unreliable responses” (Baker 2016) and thus not always indicative of future customer behavior. Against this background, it is suggested to particularly track secondary data that record genuine customer sentiment and behavior.
Which role do secondary data play in assessing customers’ experience and customers’ journey?
The scope of secondary data potentially informing customer experience and customer journey initiatives nowadays reaches beyond customer transaction records, as the rise of clickstream and social media data presents modern marketers with the opportunity of monitoring customer traces in the digital environment in unprecedented detail (e.g. Brynjolfsson, Hu, and Rahman 2013). In general, the analysis of these newly available data streams does not only allow for holistically assessing customer perceptions and intentions resp. experience by leveraging unstructured content from customer review platforms, social networks, forums, and blogs by means of text analytics (e.g. Lemon and Verhoef 2016), thus adding to the approaches based on primary data. More importantly, it enables managers to measure in how far the sentiment expressed by their customers is reflected in their real behavior along the different stages, channels, and touch points of the customer journey (e.g. Srinivasan, Rutz, and Pauwels 2016). Recalling the three-staged conceptualization of the customer journey, the next paragraphs introduce how present-day KPIs can be used to acquire a better understanding of the sequence of touch points encountered by customers along the pre-purchase, purchase, and post-purchase stage. Still, taking into account the currently limited availability of fine-grained data in the offline environment (e.g. Ailawadi and Farris 2017) and data integration issues (Dewsnap and Jobber 2000), seamlessly tracing each individual is not yet possible. Thus, up until now, the focus lies on leveraging clickstreams while partially augmenting these vast sets of information with often aggregated, traditional transaction and advertising expenditure data. Furthermore, though they are traditionally residing in the after-sales domain alongside the CSAT and other customer feedback metrics (Bolton, Lemon, and Verhoef 2008), practically applied KPIs such as average resolution time might also be of value to evaluate a firm’s efficiency in dealing with customer requests across all journey phases. Besides, since clickstream data entail records on the type of gadget used for accessing content on the Internet, i.e. smartphone, tablet, or computer, they allow for detecting device switching resp. online versus mobile channel choice along the customer journey (de Haan et al. 2018). In this context, drawing on KPIs used by firms to monitor their customers’ utilization of mobile apps, e.g. the number of downloads, users, and time spent using the app (Edelman and Singer 2015), may also contribute to an improved understanding of customer behavior. While such device-related measures potentially concern the entire customer journey (Shankar et al. 2016), other data sets allow managers to delve into its individual stages. The following figure provides an overview of the data used to measure customers’ experience and customers’ journey.
Pre-purchase stage. A wide range of academics who aim at gaining knowledge on how customers become aware of, search for, and consider a specific brand resp. product or service resorts to evaluating within-website browsing behavior. For this, granular KPIs, i.e. the individual pages of a website requested by a user during one visit as well as the respective duration of such page views, can be adduced, as Bucklin and Sismeiro (2003) show. Also, the authors accentuate that such clickstreams differ between new and repeat visitors, enabling managers to identify the idiosyncrasies in user behavior associated with new versus recurring touch points. The same KPIs are deployed to classify within-website behavior into different categories, i.e. visits serving the purpose of (1) browsing, (2) searching, (3) knowledge-building, or (4) buying (Moe 2003). Tracing such KPIs enables to better understand (potential) customers’ clickstreams, browsing objectives, and purchasing propensities in the context of a company website. Similar metrics are also leveraged to scrutinize information search on third-party platforms like Amazon (e.g. Kim, Albuquerque, and Bronnenberg 2011). Yet, not every website visit is initiated by direct type-in, which is why the monitoring of within-website behavior does not always capture the user’s entire digital footprint (Park and Fader 2004). Rather, many individuals are directed towards a website e.g. by clicking on a banner ad or an organic or paid search result (e.g. Anderl, Schumann, and Kunz 2016), thereby encountering additional touch points. To better grasp these variances in origin, the source of website visitors constitutes an essential KPI. Relatedly, a noteworthy fraction of scholars studies cross-website browsing behavior, mainly by inspecting website log files (e.g. Chan, Wu, and Xie 2011) and cookies (e.g. Kireyev, Pauwels, and Gupta 2016). In fact, academics like Dinner, Van Heerde, and Neslin (2014) contend that both customer-initiated keyword search and subsequent website visits are often preceded by firm-initiated display ads, as, along with traditional media, the latter often trigger initial need recognition and awareness, thus eliciting touch points attributable to the earliest phase of the decision process. Also, in the light of the social environment’s rising influence on customer decision processes (e.g. Wang, Yu, and Wei 2012) and the progressing disentanglement of advocacy and purchase behavior, touch points created by social media should not only be monitored in the post-purchase stage, but also prior to a transaction. While customer experience tracking focusses on sentiment in user-generated content as suggested previously, KPIs in the focal context are impressions of firm-generated posts as well as resulting reactions, measured by the number of likes, shares, clicks, and comments (e.g. Lee, Hosanagar, and Nair 2018). Moreover, as users often resort to review websites to inform buying decisions (Farhadloo, Patterson, and Rolland 2016), respective impressions should also be gauged. However, considering that the customer journey often encompasses touch points across multiple channels (Ailawadi and Farris 2017; Konuş, Verhoef, and Neslin 2008), looking at the digital environment is not always enough to capture the full scope of journey-related customer behavior. Therefore, the focus should be on detecting cross-channel spillover effects. As offline data are scarce, it is difficult to track customers in such settings. Initial attempts to assess the role of traditional advertising in the pre-purchase stage thus mainly draw on expenditure-based measures (e.g. Dinner, Van Heerde, and Neslin 2014). Accordingly, Joo et al. (2014) observe a positive relationship between ad spend and online search volume. Similarly, Fossen and Schweidel (2017) develop an approach that studies the number of brand mentions on social media, i.e. Twitter, within a time frame of two minutes before and after the airing of a television ad, identifying a positive correlation. Yet, the ability to pursue such spillovers is nascent in both research and business practice.
Purchase stage. While a customer’s online buying behavior can be fully captured (Chen, Chiang, and Storey 2012), offline purchase data is usually only accessible aggregately, as buyers often remain anonymous (e.g. Neslin et al. 2006). Academics such as Chan, Wu, and Xie (2011) indicate that B2B firms can more easily trace their clients across bricks-and-mortar channels, as they have close relations with them (Maechler et al. 2017). Yet, irrespective of the aggregation level, when seeking to understand the purchase stage, the absolute or monetary amount of purchases, transactions, or contracts is a prevalent KPI in both academia and practice (Chan, Wu, and Xie 2011). Additionally, it should be monitored which channel is used for the transaction. As research suggests that the type of purchase may differ with channel choice – for example, customers are found to primarily use the mobile channel for buying habitual (Shankar et al. 2016; Wang, Malthouse, and Krishnamurthi 2015) and cheaper products (Huang, Lu, and Ba 2016) – tracking this figure may contribute to an enhanced apprehension of customer behavior in the purchase stage. In this regard, initial research models additionally aim at shedding light not only on the purchase incident itself, but also on the transition between pre-purchase and purchase stage within and across channels. Exemplarily, transaction data can be merged with common advertising and website KPIs as already introduced for the pre-purchase stage, identifying how far customers’ touch points with a paid search ad drive online and offline purchases (Chan, Wu, and Xie 2011). Similarly, ad spend, online search volume, and sales data can be augmented, modeling how advertising impacts consumer interest, i.e. keyword search behavior, website visits, and, subsequently, conversion (Hu, Du, and Damangir 2014). In this context, the phenomenon of ‘research shopping’ as identified by Verhoef, Neslin, and Vroomen (2007) can often be observed, emphasizing that customers may switch from online to offline channels and vice versa when transitioning from pre-purchase to purchase.
Post-purchase stage. The customer journey does not end with the transaction, but also recognizes mostly customer-initiated post-purchase touch points. As per Westbrook (1987), such episodes of contact entail responses to the product or service consumption, finding expression for example in word-of-mouth and firm-directed complaints. While the latter has been measured by the number of customer complaints uttered via online and offline channels all along (e.g. Bell, Mengüç, and Stefani 2004), the tangibility of word-of-mouth has only recently gained traction due to the surge of customer engagement on social media (Bughin, Doogan, and Vetvik 2010). Here, online customer reviews represent a prominent way of sharing former experiences regarding a company or brand with a virtual community (e.g. Voorhees et al. 2017). Hence, next to analyzing the content of these posts to infer customer sentiment as described above, the number of published reviews can be traced to detect which resp. how many customers have encountered this type of post-transaction touch point. In addition to uploading such one-shot contributions, customers also harness social media networks to stay in contact with a brand after the purchase (Edelman 2010). Hence, KPIs such as impressions of firm-generated posts as well as the number of likes, shares, clicks, and comments (Lee, Hosanagar, and Nair 2018) that were already deemed valuable in the pre-purchase phase should also be considered here. As such measures are relevant in more than one part of the customer journey, placing the observed touch point into the right journey stage may be difficult. Recognizing this issue, Vázquez et al. (2014) showcase an algorithm-based method to infer a customer’s current journey position from social media comments. Eventually, customers in both B2C (Court et al. 2009) and B2B (Lingqvist, Plotkin, and Stanley 2013) might conclude the post-purchase stage by entering the so-called ‘loyalty loop’ (Court et al. 2009) or ‘enjoy-advocate-buy loop’ (Edelman 2010), repurchasing from and engaging with a company without passing through consideration and evaluation in the pre-purchase stage again. Relatedly, KPIs such as the churn rate and the number of repurchases can be adduced to detect in how far customers carry out multiple transactions (Lemon and Verhoef 2016) and thus continue their journey with a company or brand.
All in all, the insights presented in this article imply that both primary and secondary data from the domains of digital marketing and customer relationship management may indeed enable marketers to trace fractions of their customers’ experience and journey across channels, touch points, and stages. The figure below summarizes the potential inherent in such proprietary data sets, adopting a customer experience perspective as well as a customer journey perspective.
However, many marketers have not yet translated their interest into this topic into tangible actions, indicating that they may not be ready to partake in the battle for creating superior customer experience yet. Related efforts will only be crowned with success if knowledge from marketing, sales, and after-sales is not managed in siloes. Otherwise, vital fractions of the customer journey such as transitions between the marketing-dominated pre-purchase and sales-oriented purchase phase may not be assessible. Ultimately, if they succeed in employing analytical talent, cross-functional collaboration, and a data-driven culture (e.g. Jain, Aagja, and Bagdare 2017), companies can use their existing data to progress their customer experience and customer journey tracking ambitions. Eventually, such customer insights may pave the way for tailoring marketing efforts to the individual journey stage, thereby providing more valuable interactions and thus a better experience to every customer.