Customer experience has become a top priority for business (Imhof and Klaus, 2020) and is a strong predictor of customer satisfaction, WOM, and loyalty (De Keyser et al., 2015; Kuppelwieser and Klaus, 2021). Consumers tend to spread their WOM to close others about their own positive experience or someone else’s negative experience (De Angelis et al., 2012) for self-enhancement. In contrast, they are more likely to voice out to an authority (including service provider) after a negative experience (Hirschman, 1970). Some research has suggested that customers’ positive versus negative WOM was driven by different emotions (Alexandrov et al., 2013) due to different expectation confirmations (Nam et al., 2020), but how and when the customer experience valence impacts their WOM requires further investigation.

This paper investigated 1) how consumers respond to positive versus negative service experience, 2) how they spread WOM differently to the receivers (i.e., their family, friends, and social media), and 3) the better timing and reason for service providers to deliver the negative information to their consumers to reduce the inevitable negative WOM.

We recruited 227 U.S. participants and conducted an experiment to examine the likelihood of spreading WOM based on consumer experience valence, WOM receivers, nature of a service failure, and the timing of learning such a failure. We first showed (all) participants the same positive scenario (i.e., getting a free hotel upgrade) and asked them how likely they would share this news with their family, friends, or social media. We then manipulated the negative experience (i.e., ended up not getting the free upgrade) by conducting a 2 (service failure reason: internal (i.e., hotel’s computer system failure) vs. external reason (i.e., leaking due to bad weather) × 2 (delivering timing: telling the customer before the trip started vs. during the hotel check-in) between-subject design.

Paired-samples t-tests showed that, in general, people are more likely to share a positive experience than the negative one with their family, friends, and on social media (all p’s < .001), replicating the results from De Angelis et al. (2012) study. In the context of negative service experience, an ANOVA identified a marginally significant interaction between failure reason and timing only when they share the experience with their friends (F(1, 224) = 2.78; p = .097). Contrast analysis showed that when the negative experience was informed to the customer upon arrival, participants were more likely to share this bad experience with their friends (but not with family or on social media) when they learned that the reason was internal (M = 4.96, SD = .25) than external M = 4.31, SD = .25; F(1, 224) = 2.78; p = .068).

Our study contributes to the service experience literature by examining and comparing how customers respond to positive versus negative experiences in different timing and reasons and different receivers. It indicates the importance of information timing and failure reasons for service providers to reduce consumers’ negative service experience and suggests a way to minimize such inevitable negative WOM.

Customer experience management (CXM) has been conceptualised as a higher-order resource in an organisation with capabilities to improve experiences (Homburg, Jozić and Kuehnl, 2017). Such capabilities are impacted by emergent technologies in service delivery to be augmenting and substituting humans at the organisational frontline (De Keyser et al., 2019). Both machine and human intelligence skills are changing rapidly with progress equally in demands as well as their meaning(Puntoni et al., 2021). The literature in CX(M) needs to focus on an integrated customer-firm perspective that includes varied contexts (for e.g., B2B settings). Hence, this study will contribute to practice and research by understanding the firm capabilities needed to balance human and technology participation based on capability theories applicable to organisations.

For example, three thousand washing machine like robots operate in an Ocado warehouse to pick grocery orders. Fifty item baskets are picked within five minutes accurately without wastage, enabling reliable, faster and cheaper delivery to customers. Humans are still involved in the final stage packing complemented by artificial intelligence, but UK’s leading grocery retailer, Ocado have ambitions to remove human touch from the whole process (Davis, 2021; Ocado Technology, 2021). Such unparalleled delivery of customer experience (CX) is an example of the shifting blend between human and technology participation in managing CX.

However, the balance towards automation may not necessarily be a uniform choice for CXM. For example, B2B settings that thrive on human touch need careful consideration in striking the right balance between human connection and technology-driven efficiency(Walker, 2021). Furthermore, if customers are unwilling to adopt technology, experience led by technology may not always be a suitable choice for CX leaders. This can be derived from Kumar et al.'s (2019) construct on customer engagement and its relationship with technology adoption illustrated by a study concluding that technology acceptance played a key role in adoption of eCommerce platforms(Bilgihan, Kandampully and Zhang, 2016).

Consequently, our understanding of how this blend of human and technology participation influences CXM capabilities needs to be enhanced. The role of customers, employees and technology has been fused and conceptualised as service encounter 2.0 and its proponents recommend further empirical studies in this area(Lariviere et al., 2017). Furthermore, service research continues to call for studies on the role of technology in CX(Mele, Russo-Spena and Kaartemo, 2020; Ostrom et al., 2021). Therefore, the objective of this study is two-fold. First, we explore how are firms balancing the blend between human and technology participation in managing CX. Second, we investigate the impact of this blend on the diversity in skills (both human and AI) on CXM.

To meet the objectives, the study will gain knowledge from industry reports and in-depth interviews with CX and technology leaders at the frontline of this service transformation. We will explore a broad range of settings to derive a conceptual framework to progress our understanding of human and technology participation in CXM. Preliminary findings suggest that the level of technical knowledge (both customers and employees) makes a significant difference to the composition of CXM capabilities.

Improving customer experience and journeys can increase customer satisfaction by 20%, lift revenue by up to 15% while lowering the cost of serving customers by as much as 20% (Pulido, Stone and Strevel, 2014). However, for small businesses and particularly start-ups with very limited resources and expertise, understanding and investing in digital customer experience and journey mapping can be beyond the scope and budget of most small businesses. In this paper we reveal preliminary insights on the use of Screencast Videography (Kawaf, 2019) as a possible approach for assessing the digital customer experience and journey mapping for such businesses with relatively simple to use tools and limited resources. We present a case study on Pic Tree; a creative multimedia agency providing services in the B2B market and show how simple screencasting of user journeys can create meaningful insights that are easy to comprehend and apply with limited resources.
Kawaf (2019, p170) defines Screencast Videography as “a research method that adopts a dynamic visual form of inquiry. It is philosophically underpinned by the ontology of the moving image. The method uses screencasts – videos of screen activities or outputs – as its main mode of data collection. The screencast videos capture dynamic on-screen interactions and experiences as they occur. This helps offer detailed records of online experiences (e.g. online shopping, information search, dating, video gaming, gambling, etc.) that are not usually observable using conventional methods”. We employ screencast videography to capture the journeys users go through while on Pic Tree’s website (www.pic-tree.com). As a starting point, the owner of the company is asked to create her own screencast navigating the website, this footage is not viewed by any other members of the research team at this stage and is left merely for comparison once the data from users who are not familiar with the site are captured. The research team then record their own screencasts in ‘think aloud’ mode with audio captured during the progression of the experience. In total, six screencasts are captured (five are from the research team and one from the company owner).
The videos are imported in Nvivo and coded using Critical Incident Analysis (Edvardsson & Roos, 2001). We further edit the video to produce a meaningful video summarising the progression of the user journey throughout the site highlighting all critical moments where improvements are required.
The findings show a clear discrepancy between how the company owner and the users progress through the site hinting to a need for restructure of the site that speaks to the user directly. In addition, the screencasts show excellent examples of ‘tacit knowledge’ that were unobtainable to the company owner otherwise. For example, the website home page appears differently to different users making it harder to interpret what services are offered depending on the version displayed. Various other examples can be seen in the mapping of those journeys and can be shared upon request.

Background, significance, and research aim
Customer experience has been studied extensively over the years. Few review-based studies helped to understand fragmented customer experience literature, including customer experience formation (Lipkin, 2016), levels of customer experience (Kranzbühler et al., 2018), premises of customer experience (Becker & Jaakkola, 2020), and customer experience TCQ (touchpoints, context, qualities) nomenclature (De Keyser et al., 2020). Although much is known about customer experience in general and technology-enabled customer experience (TECX) in particular, research lacks in understanding theoretical, methodological, and contextual lens of TECX.

Therefore, the current study aims to develop a holistic framework of TECX by integrating Theory Context Methods (TCM) and Antecedents Decisions Outcomes (ADO) frameworks. Few recent examples of TCM and ADO frameworks include ‘life insurance purchase behaviour’ (Bhatia et al., 2021) and ‘performance behaviour’ (Lim et al., 2021). These frameworks are useful help to set up future research agendas with a comprehensive focus using theoretical, methodological, and contextual lens (Paul et al., 2017; Paul & Benito, 2018).

Methods and preliminary findings
After a systematic search of the customer experience literature using search string (“customer experience*” OR consumer experience*” OR “consumption experience*” OR “experience* of customer*” OR “experience* of consumer*”) during Aug 2021 from Web of Science database, which is globally recognized for internationally acclaimed publication outlets (Chabowski et al., 2018; Korom, 2019). The search yielded 4264 documents published in English.

After that, 4264 documents were tagged based on technology-enabled focus that resulted in 771 documents using keywords, including ‘technology’, ‘social media’, ‘digital’, ‘online’, ‘internet’, ‘artificial intelligence’, ‘virtual reality’, ‘augmented reality’, ‘big data’, ‘robot’, and ‘blockchain’ (Shi et al., 2020; Lu et al., 2020; Silva et al., 2021).

Following the AJG and ABDC journal indexing 244 high-quality articles retained for further analysis. Out of 244 articles, only 43 appeared to be with qualitative research design and 154 articles followed quantitative research design that shows quantitative dominance. After careful analysis of the models used in 154 articles, 44 articles were relevant to customer experience. Further analysis for TCM and ADO will be completed with 44 articles.

Discussion and conclusion
The paper brought a holistic, current, and relevant view of fragmented TECX literature to understand elements (theory, contexts, methods, antecedents, and outcomes). The paper presents a powerful framework that can be tested in multiple contexts (within and across countries and industries). Having this understanding enables scholars to identify research gap and contribute within the ongoing academic discussions around TECX.

Moreover, the empirical studies on customer experience and journeys have also been getting attention (Kranzbühler et al., 2019; Jaakkola & Terho, 2021). Therefore, after carefully selection of theories, contexts, methods, constructs from the TECX framework, there could be more effective research designs to extend the literature of total customer experience at every stage of customer journey (Lemon & Verhoef, 2016). This will ultimately enhance theoretical, methodological, and contextual contributions to enrich the intellectual discussion on technology-enabled customer experience.