Abstract
As firms increasingly uncover their activities to key stakeholders through various media, the perception of these activities is becoming more important for firm performance. Traditionally, access to industry-wide databases provides important metrics on customer perceptions of performance, such as customer satisfaction and brand equity. In addition, numerous studies have highlighted firms' innovation-related actions (e.g., R&D spending and patent counts) as critical metrics linked to their financial performance.
Introduction
Customer-perceived metrics such as customer satisfaction, loyalty, and brand image are important metrics, which have been shown as substantial predictors of a firm’s financial success (e.g., Fornell, Morgeson, and Hult 2016). Accordingly, industry-wide perception metrics like the American Customer Satisfaction Index (ACSI) are available and used by firms and researchers globally. We argue that these customer-perceived measures are likely to become even more important as more firm information is and needs to be made available to various stakeholders across multiple touchpoints. Consequently, firms are also using more resources to communicate
Customer satisfaction and brand equity can be seen as downstream evaluations of a firm’s upstream activities. Furthermore, they tend to be measured cumulatively, which implies that customers consider all previous interactions and make an aggregate evaluation of their perceptions of a firm. This, in turn, limits the diagnostics of the measure (i.e., what caused the evaluation). Consequently, making strategic decisions about how a firm should act becomes challenging, highlighting the need for more upstream information to understand the reasons behind a customer's specific evaluation of a firm. Innovation has been regarded as an important upstream metric due to its strong linkage to financial performance. Studies linking innovation to financial performance typically use objective metrics, such as the number of patents and R&D expenditure, as proxies for innovation activities. Needless to say, these metrics have flaws and may be less suited in a service context as there are fewer objective measures for the service industry, which motivates a need for a measure of perceived firm innovativeness (PFI).
Keiningham et al. justly argue that
Consequently, the creation of the American Innovation Index (Aii), which measures PFI across different service industries, and similar initiatives around the globe are needed and commendable as they create interest in the area. Importantly, Keiningham et al. demonstrate that Aii does capture valuable information. It is notable that the researchers find an impact of Aii since service innovations take on different forms and have different purposes, and may not even have a direct noticeable impact on a market. For instance, service innovation (versus innovation of goods) often focuses on process innovation, is less radical, and frequently involves a recombination of existing innovations (Gallouj and Weinstein 1997).
We agree with Keiningham et al. on the importance of PFI and view this as a key first step. However, to further enhance research on PFI, more research should address the underlying theories and logic behind using PFI as an indicator of service innovation. In this process, the importance of methodological and estimation procedures should also be highlighted. In the next sections, we will build on the approaches taken by Keiningham et al. and identify alternative routes and strategies that future research may take.
Capability of PFI in Capturing Service Innovation
Research on new service development has, since the 1990s, indicated that service firms (as opposed to goods firms) are lagging behind in their capabilities to innovate (Ghanbarpour et al. 2024). Service industries are simply not using sufficient resources, and they still lack sufficiently structured processes to produce innovations. A customer-perceived metric like Aii, which captures service innovation, can be a tool that creates greater focus on this area and inspires firms to change their behavior.
Understanding how customers perceive a firm's innovativeness provides valuable insights into market positioning and customer engagement, and helps firms understand how
Not all innovations are directly related to customers
A significant deficiency of relying solely on customer-perceived innovativeness is that many innovations occur behind the scenes and are not directly observable or understandable to customers. These innovations may involve internal processes, supply chain management, or back-office operations aimed at improving efficiency, reducing costs, or enhancing service delivery capabilities (Gallouj and Weinstein 1997). For example, a service firm may implement a sophisticated data analytics system to optimize its inventory management. While this innovation significantly contributes to the firm’s operational efficiency and cost-effectiveness, customers may remain unaware of these improvements and their indirect benefits to customer service.
Customers might not realize the innovative actions of firms that target them
Customers' ability to recognize and appreciate innovation is limited by their knowledge, experiences, and expectations. Innovative actions that significantly enhance service quality or delivery may go unnoticed if they do not directly change the customer's interaction with the service (Vargo and Lusch 2004). For instance, a bank that develops a new risk assessment algorithm to approve loans more efficiently provides an innovative service that directly benefits customers. However, if the innovation is subtle, as it often is in a service setting, or if its benefits are not immediately apparent, customers may not perceive the bank as more innovative than its competitors. This discrepancy highlights the challenge of measuring firm innovativeness solely based on customer perceptions.
Notably, PFI is probably the only approach capable of capturing the innovation of the goods part of the service. Many service innovations derive from or are enabled by product innovations (Gallouj and Weinstein 1997). However, in the majority of cases, the service firm is not a producer of the goods-based innovation, but rather it delivers a goods-based service. For example, healthcare monitoring services that utilize wearable technology exemplify a complex blend of product and service innovation. Wearable devices with advanced sensors offer continuous health data tracking, integral to personalized healthcare services. In these examples, while the service innovation is the key value contributor, its perception is inherently tied to the product innovation that enabled it. Traditional measures that focus on patents and R&D expenditures may not be able to capture this, making PFI a better measure to capture the true value of the innovation.
Overall, while Aii is a very important step to understand the impact of service innovation, the work is, however, not done. The service innovation characteristics highlighted above underscore the need to develop a theoretically based, multi-dimensional approach to measuring innovativeness. The scope of service innovation is simply too wide to be captured with one measure, such as PFI (Gustafsson, Snyder, and Witell 2020). We need to account for both visible and invisible innovations and recognize the complex interplay between service and goods-based innovations. We would also like to stress that in addition to customer-perceived firm innovativeness metrics like Aii, we can clearly see a need for other industry-wide upstream
Methodology and Estimation
Service research, in general, has valued relevance, and we have a long-standing tradition of appealing to practice. We should
The research by Keiningham et al. is based on secondary data from Aii, ACSI, and abnormal stock returns. Combining these three data sources results in a final sample of 78 publicly traded service firms in the United States over a 5-year period, comprising a sample of firms with rather unique traits, such as being relatively large and well-recognized. To put the sample size in perspective, Ghanbarpour and Gustafsson (2022) use a sample of 69 firms from a small set of industries, and Ghanbarpour et al. (2024) include 467 firms from both service and goods-based industries. Secondary datasets always depend on the data that is available, and there is not much we can do about it. However, the sample size, which is influenced by merging different databases and their limited overlap, does impact the generalizability of findings. Therefore, samples and the implications of limited or unique samples need to be discussed more thoroughly. For instance, we can argue that most research using secondary data may not be relevant for smaller service firms as they are generally not covered in the relevant databases. A limited sample size also causes other problems, as it is simply not possible to include all the necessary variables in the estimation model. This may lead to an underspecified model, as it does not control for industry (e.g., industry competitiveness) and firm-level (e.g., accounting indicators) factors (Malshe, Colicev, and Mittal 2020).
Another issue we would like to highlight is the selection of the most relevant metrics. We are fully aware that the choices are limited to what is available to the researcher. Nevertheless, we would still argue that researchers should always strive to select performance measures that are directly linked to marketing activities and customer interactions. In the case of PFI, customer satisfaction is highly relevant, as one of the outcomes of innovation activities can be a change in customer experience. In terms of linking PFI to financial performance, sales figures, which are directly related to customer transactions, might offer a clear and direct measure of how customer perceptions translate into financial performance—and can serve as a mediating mechanism in the relationship between PFI and abnormal stock returns (Katsikeas et al. 2016).
Keiningham et al. employ a relatively advanced estimation procedure, prompting us to issue a call to action for service researchers in this field. We would like to commend Keiningham et al. for using a longitudinal estimation procedure. Longitudinal estimation methods have many advantages that enable the studies to produce more reliable results, as they provide a stronger basis for inferring causal relationships between variables (compared to cross-sectional studies) and allow for the examination of both within-firm and between-firm variations over time. Accordingly, we emphasize the need to use such procedures more extensively in service research.
Discussion
This commentary underscores the challenges in using PFI as a measure of service innovation, where many innovations are invisible to customers. We believe that Keiningham et al. significantly contribute to the innovation and marketing-finance literature, as we need more studies on the impact of service innovation. We are especially excited to see a longitudinal and available upstream metric such as Aii that can help companies create strategic initiatives that are reflected in the existing downstream metrics (e.g., customer satisfaction and financial performance). We would, however, like to see more such metrics made available to researchers and practitioners.
We would also like to use this commentary to advocate for enhanced research into service innovation. Service innovation is crucial for businesses, yet research in this area remains relatively underdeveloped. Service industries often lack the structure and capabilities needed for innovation. We hope that useful indices like Aii will attract more attention and resources to innovation in this sector. However, we also want to highlight that the scope of service innovation is too wide to be encapsulated by a single metric like Aii. We simply need more relevant measures that can be linked to stakeholder perceptions and financial performance.
Another critical area for action is the effort to increase rigor without losing relevance. There are several key factors worth emphasizing more broadly. First, the significance of the sample cannot be overstated. The inferences we draw are only applicable to the sample from which the data is collected, whether primary or secondary. Larger samples allow for broader generalizations and the ability to control for more variables, reducing potential biases in our findings.
Finally, the choice of estimation procedures, assumptions regarding relationships, and the measures employed are all crucial. These elements can significantly influence the outcomes of research and, consequently, the decisions based on those findings. Incorrect assumptions about relationships (e.g., assuming linearity when the relationship is non-linear) or using measures not directly linked to an outcome can introduce biases. Similarly, failing to control for the correct factors can also skew results. Thus, these aspects demand careful consideration to ensure the reliability and applicability of research findings.
Footnotes
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.
