Managing customer satisfaction: digital applications for insurance companies

Customer satisfaction management is increasing in importance within the insurance industry. In particular, to define a customer-oriented strategy, installing digital applications based on technologies, e.g. including artificial intelligence or cloud computing, ranks among the major strategic challenges. Against this background, the aim of this paper is to take an integrated perspective on managing customer satisfaction and the digital transformation. Towards this end, we identify and assess a set of digital applications, as a result of a comprehensive review of 106 academic papers and publications of the industry and supervisory authorities. We illustrate the opportunities to increase customer satisfaction and emphasise their impact on insurers at four major customer touch points: contract conclusion, contract modifications, the event of damage and further contacts. Our results are strategic measures to strengthen the position for sales and marketing, to simplify standard processes and to increase efficiency and interaction with the customer.


Introduction
In the past, customer satisfaction in insurance has not been widely discussed as a key steering figure for insurance management, and academic research on customer satisfaction in the insurance industry is notably low compared with other sectors (e.g. Brutyan et al. 2019;Jahnert and Schmeiser 2021;Pooser and Browne 2018). Today, digitalization affects and transforms customer behaviour, customer expectations and customer requirements (Catlin et al. 2015;Cebulsky et al. 2018;Lyskawa et al. 2019). Private insurance customers experience highly transparent, fast service processes in other industries (e.g. firms such as Amazon setting the benchmark for best practices) and transfer their requirements to the quality of the insurance customer experience (e.g. Catlin et al. 2015;Hall 2017;Müller et al. 2015). Moreover, the insurance market observes the phenomenon of InsurTechs, which is beneficial for customers, as they operate flexibly and simplify innovation processes (e.g. Eling and Lehmann 2018;Grima et al. 2020). Hence, managing customer satisfaction has become more important and more challenging for incumbent insurance companies.
At the same time, digitalization also provides considerable benefits for insurance companies in this regard. By enhancing the customer experience (Eling and Lehmann 2018), bringing policyholders and insurance companies closer together (Kotalakidis et al. 2016) and widening the role of insurance companies to become the insured's risk manager, digitalization might help improve customer satisfaction (e.g. IAIS 2018), as insurance companies evolve from pure risk protectors to risk predictors and preventers (Kelley et al. 2018;Schmidt 2018).
Previous literature reveals that customer satisfaction drives customer loyalty and leads to fewer customer complaints (Fornell et al. 1996;Helgesen 2006;Heskett et al. 1994). Moreover, there is an abundance of literature investigating customer satisfaction in general (for an overview, see, e.g. Arora and Narula 2018). In particular, Oliver (1977Oliver ( , 1980 developed the expectation confirmation theory, which argues that customer satisfaction is mainly driven by customer expectations and perceived performance. Outperforming customer expectations with the product or service leads to higher customer satisfaction, while underperforming decreases customer satisfaction (Oliver 1977(Oliver , 1980. In the context of insurance customer satisfaction, existing empirical evidence indicates that lower expenses and combined ratios in the presence of satisfied customers are due to an increasing number of renewed contracts (instead of new policies), with reduced costs for customer acquisition. Therefore, increasing customer satisfaction is discussed to enhance the profitability of insurance operations (Pooser and Browne 2018). 1 Jahnert and Schmeiser (2021) extend the investigations on the relation between customer satisfaction and profitability in the insurance industry by analysing data at the level of individual customers stemming from a Swiss non-life insurance company. Moreover, due to the widespread utilisation of information technology, the option to conduct comprehensive research from the perspective of customers in relation to the characteristics of insurance products, prior to purchasing them, becomes increasingly relevant (e.g. Mau et al. 2015): customers can easily compare products using comparison portals or conveniently buy products online, which is especially important for property and casualty insurance (e.g. Woo-Yeon et al. 2021). To further underpin the relevance of digitalization, recent empirical works discuss process improvements toward customer centricity (e.g. Kreuzer et al. 2020), customer engagement through value co-creation (e.g. Chen and Chen 2017), the trade-off between customisation and complexity (e.g. Grösch and Steul-Fischer 2017;Leischnig et al. 2018), the decisive function of intermediaries (e.g. Dominique Ferreira 2018) and the reliability and responsiveness of customer support (e.g. Ramamoorthy et al. 2018) as prerequisites for customer satisfaction in the insurance industry. As customer satisfaction builds trustful relationships, it thereby increases policyholders' willingness to disclose personal data, which in turn unlocks various benefits for insurers using digital technologies (Steiner and Maas 2018).
However, to the best of our knowledge, there is a lack of academic and practitioneroriented research focusing on the benefits and opportunities of digital applications when managing customer satisfaction from the perspective of insurers, with a focus on the major customer touch points. By including them, we extend the investigations of Eckert and Osterrieder (2020), who take a more technical perspective on implementing digital technologies (e.g. big data, artificial intelligence, cloud computing, the Internet of Things and distributed ledger technology) and discuss the interdependencies between them. In particular, we contribute to previous work by identifying and analysing a set of digital applications. We distinguish between front-end and back-end functionalities and assess the opportunities of these applications at the major customer touch points of insurers, including contract conclusion, contract modifications, the event of damage and further contacts. Hence, our results provide insights and guidance for managing customer satisfaction in a targeted manner on the corresponding customer touch points (e.g. allow a focus on customer touch points, where customer satisfaction is currently low). For each incorporated digital application, we additionally provide a set of necessary requirements, so as to install these and comprehensively discuss their limitations due to concerns, e.g. related to data utilisation and customer attitudes regarding new technologies or regulatory hurdles.
The paper is structured as follows. The next section describes the theoretical background and addresses the terminology. We present the set of digital applications and focus on their benefits and opportunities at the four major customer touch points in the subsequent two sections. Then, we discuss the limitations and derive the related requirements for insurers. The final section summarises the results.

Theoretical background: methodology and analysis
The paper focuses on a comprehensive assessment of digital applications for insurers when managing customer satisfaction. To this end, we create, review and assess a comprehensive literature data sample, which consists of 106 articles in total and incorporates academic research and the publications of industry experts and supervisory authorities. 2 The sample is based on the review approach of Eckert and Osterrieder (2020) and is a result of key word searches including 'insurance' AND 'artificial intelligence', 'insurance' AND 'big data', 'insurance' AND 'blockchain', insurance' AND 'cloud computing', 'insurance' AND 'digitalization', 'insurance' AND 'digital transformation', 'insurance' AND 'distributed ledger technology' as well as 'insurance' AND 'internet of things'. We thereby implement the searches in the selected journal databases (ABI/INFORM Collection, Business Source Complete, EconLit Full Text), and amend the focus of screening and selecting the resulting articles in relation to the effects on customer satisfaction. Besides, to get a reasonable understanding of the current state of research on customer satisfaction in the insurance industry, we search the Scopus database for scientific journals using the queries 'customer' AND 'satisfaction' in the field of business, management and accounting. Finally, to reduce the risk of omitting literature important to this work, we extend our scope to Google Scholar and Google, to incorporate recent discussions in the industry and additionally review and assess the cited references (see also Eckert and Osterrieder 2020;Gatzert and Osterrieder 2020).
Based on this, we have created a set of digital applications for insurers and explain why and how these applications support managing customer satisfaction, which "evaluates whether the customer was satisfied with the insurance services, insurance transaction, and their relationship with the insurance company" (Nguyen et al. 2018, p. 4). 3 To comprehensively assess the aforementioned set of digital applications for insurers, we consider versatile internal options of utilisation that affect different strategic levels within an insurance company. For this reason, we cluster the set of digital applications as a first step based on a categorisation of their main focus on automating or digitally enhancing back-office functionalities, front-office functionalities, or a combination of both (e.g. Günzel and Holm 2013;Marquez 2010;Osterwalder and Pigneur 2010): • Back-office functionalities are related to enhancing the efficiency of insurers by affecting key resources, key activities, key partners or the cost structure. • Front-office functionalities comprise the value-driven parts of insurers' business models by affecting the value proposition, channels, customer relationships or revenue streams. 3 See Giese and Cote (2000) for a detailed literature review and discussion concerning the definition of customer satisfaction. 2 The literature data sample consists of 83 academic articles that are relevant to the subject of the research questions. These 83 articles (mainly peer-reviewed, as, for instance, relevant working papers are included) represent the foundation from which we draw conclusions for the paper. However, we consider 23 web-references as a supplement to capture recent discussions of the topical strategic issue of managing customer satisfaction in insurance companies. Even if the content of the web-references incorporates non-peer-reviewed documents, such as reports or studies of industry representatives (e.g. consultancy companies), that are not representative of scientific knowledge, considering the articles allows us to draw further implications from current developments in practice. Figure 1 illustrates the integrated perspective on the digital transformation and customer satisfaction in the insurance industry. Moreover, it provides an overview of the analysis along with definitions of the relevant terminology and substantiates the derivation of the relevant research questions. In general, the relevance of the illustrated research questions is underpinned by the importance of managing customer satisfaction in the insurance industry: academic work connects customer satisfaction in the insurance industry to customer loyalty (e.g. Abu-Salim et al. 2017;Lee 2019;Nguyen et al. 2018;Ruefenacht 2018), which is important due to its potential impact on the (current and future) economic results of insurance companies (e.g. Eskildsen and Kristensen 2008;Trautinger 2015). In addition to that, Jahnert and Schmeiser (2021), for instance, emphasise the vast amount of non-insurance literature that highlights customer satisfaction as an important factor for a firm's profitability, and state the necessity of differentiated considerations of the effects for the insurance industry. Extending the focus to academic work from outside the insurance industry, empirical findings state the positive impact of customer satisfaction on organisational performance (Ittner and Larcker 1998), such as stock prices (e.g. Aksoy et al. 2008;Fornell et al. 2006Fornell et al. , 2016, cash flows (e.g. Gruca and Rego 2005), or Tobin's Q (e.g. Anderson et al. 2004). Moreover, customer satisfaction increases the market share of firms and creates barriers to customer defection (e.g . Fornell 1992;Matzler and Hinterhuber 1998). Thus, firms benefit from cost savings as retaining already existing customers is typically less expensive than acquiring and nurturing relationships of new customers (e.g. Mittal and Kamakura, 2001;Reichheld and Sasser 1990).
Regarding the drivers of customer satisfaction, previous research additionally points to the customers' perceived value of a product or service (e.g. Dodds et al. 1991;Parasuraman 1997). According to existing research (e.g. Anjum et al. 2016;Dodds et al. 1991;Fornell et al. 1996;Johnston 1995;Parasuraman et al. 1988;Zeithaml et al. 1990), the perceived value itself is driven by corporate image, service quality and price (see, e.g. Nguyen et al. 2018 for an overview in the context of the insurance industry).
When managing customer satisfaction, it is of great relevance to focus on the customer touch points during the customer journey, as these generate customer perceptions. These customer touch points are commonly used when measuring customer satisfaction (e.g. AssCompact 2020). For this reason, we cluster the illustrated digital applications in terms of applicability for each of those touch points: contract conclusion, policy modifications, event of damage and further contacts. The latter includes points of contact, which are related to providing additional services, offering information on general features or placing general advertisements (e.g. provided by insurance brokers or agents). By integrating the categorisation into our analysis, we are able to derive a set of strategic measures for insurers, aimed at specific areas of the customer journey (e.g. by focusing customer satisfaction management on the touch points, which are characterised by a low customer satisfaction level).

The set of digital insurance applications: overview and characteristics
As previously explained, the set of digital applications for insurers was created based on the literature data sample. By reviewing and analysing the incorporated articles, we further develop the investigation of Eckert and Osterrieder (2020) and consider  Gatteschi et al. 2018), robo-advisory (e.g. OECD 2017) and chatbots (e.g. Gruhn, 2018;Riikkinen et al. 2018). Table 1 provides a comprehensive overview of the incorporated set of digital applications categorised by their main focus on back-office functionalities, frontoffice functionalities, or both, including a description and the major characteristics.
Note that some of the digital applications can be seen on different hierarchical levels. For instance, image recognition can be also used to improve fraud detection. Practical literature (e.g. Bitkom 2018) often refers to a 'periodic table of AI' that consists of elementary applications of artificial intelligence such as image recognition or predictive inference, while applications in practice, e.g. fraud detection, are frequently a combination of these elementary applications of artificial intelligence. In this paper, we focus on the most important digital applications for insurance companies, which often consist of more than one elementary artificial intelligence/digital application and do not aim to go to the level of the elementary applications. However, the areas of applying image recognition in insurance companies go way beyond fraud detection, as explained in Table 1. For this reason, in this case the elementary application itself is important enough to be considered as a separate digital application. Moreover, fraud detection is not only improved by means of image recognition, which is why we investigate fraud detection also as a separate digital application, leading to dependencies between some digital applications in Table 1.

Four major customer touch points: benefits and opportunities of digital applications
In what follows, we discuss and assess the benefits as well as the opportunities of the presented set of digital applications, in terms of managing customer satisfaction from the perspective of insurers. Table 2 states the relevance of the applications within the displayed 'opportunity landscape' and emphasises the potential to enhance, e.g. the customer experience, but also the customer journey, based on the literature. Moreover, the specific benefits and opportunities are clustered for each of the four customer touch points (contract conclusion, contract modifications, the event of damage, further contacts). These provide the basis for the strategic decisions of insurers when evaluating applications in regard to deriving measures to Back-office functionalities Customer segmentation • Subdivides the market into customer groups (e.g. depending on their needs or past behaviour) g,m,w,ai • Provides the basis for granular and more precise segments, e.g. for risk assessments, based on an increasing number of data sources and prioritises customers, e.g. depending on profit potential and ability to serve them m,n Digital monitoring and telematics • Record, track and evaluate individual information regarding policyholders l,m as a result of the increasing integration of new data sources (e.g. due to the internet of things with the adoption of sensors and wearables) c,m,aa,ac • Affect internal processes (e.g. insurers gain further data that can be used to improve risk assessment but also product design) j,l,m,ab • Practical examples in regard to data availability: driving behaviour (measured by variables such as speed, braking, distance driven), b,m,at activities to maintain a healthy lifestyle (e.g. fitness activities), m,ar but also 'smart home' data e Digital ratemaking and underwriting • Describes automated and/or digital underwriting and ratemaking processes: while cloud computing or blockchain enable the capture of large amounts of data, blockchain can also automate the underwriting process, n which is due to artificial intelligence no longer being limited to historical data but extended to real-time information t • Increases accuracy with regard to risk selection, insurance rating classification and thus insurance rating factors m,n,t,v,at • Automates and secures the process of underwriting by evaluating, e.g. unstructured customer data with artificial intelligence s Fraud detection • Allows the precise analysis of fraudulent patterns to more effectively detect (e.g. false claims or manipulated claim size) and prevent them h,i,m-o,r,v,al,as • Practical examples for applying enhanced fraud detection: natural language processing and text analytics or speech recognition (e.g. considers contract applications or phone conversations) k Image recognition • Allows the automation of internal processes by translating images as unstructured data into explanatory information, based on machine learning techniques k,m,u • Practical examples refer to applying image recognition for digital fraud detection (thus interdependent applications), the validation of signatures/handwritings, video surveillance, customising services (e.g. VIP identification) or tailored marketing activities (e.g. based on customer's social media content) am Sentiment analysis • Automatically recognises emotions, e.g. based on analysing the voice during a call, specific text messages (e.g. mails and social media content) or videos of customers k,m • Practical examples: sentiment analysis is suitable for enhancing fraud detection, due to its integration in claims management k,m • Describe dialog systems employing natural language processing to process text input, as well as text output, which ensure a basic service, especially in peak hours (e.g. after a hurricane) a,f,x,ak Customer targeting • Tailors (online) product offerings and aligns these to a customer's personal life stage (e.g. marriage) m,v,ab,as • Practical example: customers receive real-time insurance offerings (e.g. a pop-up message on their smartphone) with certain trigger events (e.g. acquisition of a defined product) as Digital first notification of loss • Incorporates functionalities, such as an automated signal to involved parties regarding a defined trigger event, such as a car accident in the context of property and casualty insurance m • Can potentially report additional information on the course of the trigger event: can be accompanied by additional information for the policyholder (e.g. including the application of chatbots, which are described in more detail above), but also for insurers (e.g. predictions on required reserves) m Omni-channel approach (integrated customer portals) • Focuses on adding further online/mobile channels (e.g. 'direct-to-consumer') to the common offline customer interface s,z,ag • Offers information and products via several channels, adjusted to the requirements of digital customers (focus: convenience), ae with further relevance for advanced assistance services ag Up-and cross-selling recommendations • Provide more granular up-and cross-selling probabilities for specific products for individual customers m,af,ai e.g. by taking into account the customer lifetime value n • Provide support to flag the most attractive customer, based on precise up-and cross-selling probabilities, e.g. as a result of trained machine learning models m,n a Abdul-Kader and Woods (2015)        Up-and crossselling recommendations • Extracted information, e.g. stemming from analysing big data, can be applied to e-commerce to increase the probability of a purchase, but also to ease the process of a buying decision i,n,y  Table 2 shows the benefits and opportunities for customer satisfaction in terms of each digital application at each customer touch point in detail, the presented digital applications address all of the three components of customer satisfaction, namely service quality, corporate image and price, and therefore the perceived value (see Fig. 1). We subsequently provide a summary of the major findings aggregated for each touch point and highlight the main effects of utilising the applications as strategic measures for insurers.

Contract conclusion as a customer touch point: strengthening the position for sales and marketing
The digital applications with potential opportunities at contract conclusion are characterised by increasing customer-centricity (e.g. Matouschek and von Hülsen 2015;EIOPA 2019). In general, they aim to better understand the customers and to offer more tailored products to suit the individual needs of a customer perfectly (e.g. Brenner 2019; Eling and Lehmann 2018;Heo and Grable 2017). By building on refined customer segments of precise segmentation, for instance, customer targeting allows enhanced advertising and tailored product offers, based on comprehensive and customer-related data (Eling and Lehmann 2018;EIOPA 2019;Helfand 2017;Lehrer et al. 2018). In practice, the extracted information can be applied to e-commerce to increase the probability of a purchase, but also to ease the process of a buying decision for a customer (e.g. Eling and Lehmann 2018;Gupta and Tham 2018;Owadally et al. 2019). 4 Due to the high number of available data streams (e.g. conventional information of a policyholder, sensor data, social media, information of emotions) and the related methodologies to further process these, insurers have the opportunity to precisely predict or anticipate recent customer needs before and when concluding a contract (e.g. Brenner 2019; Heo and Grable 2017). 5 For this reason, the literature emphasises the strengthened position with regard to sales and marketing activities (e.g. EIOPA 2019; Catlin et al. 2015;Kotalakidis et al. 2016).
To this end, the literature addresses the importance of the 'seamless' customer experience by integrating insurance products into their own respective partner websites and mobile applications or by engaging in e-commerce and online marketplaces. Insurers can also place their products on online comparison websites (e.g. Gupta and Tham 2018;Matouschek and von Hülsen 2015). In this context, the digital applications provide the basis for offering the required, convenient solutions (e.g. product bundles) via flexible channels (e.g. due to integrated customer portals and the omnichannel approach). The resulting customer experience is thus characterised by a lack of friction in relation to switching between different channels, enabling customers to make an informed decision on a particular insurance policy, independent of time and location (e.g. Müller et al. 2015).
However, the application process can be secured, e.g. due to digital ratemaking and underwriting (Gupta and Tham 2018). In this context, the results in Table 2 point to the fact that the applications can help to enhance actuarial models with a more precise risk assessment (e.g. digital monitoring and telematics). In general, the digital applications support optimising core capabilities, such as underwriting, risk selection or claims prevention (e.g. Catlin et al. 2015Catlin et al. , 2018. The latter, moreover, causes a shift from indemnifying losses to predicting and preventing risks, with regard to the role of an insurer (e.g. Spender et al. 2019), while the latest initiatives of insurers focus on the fields of 'mobility', 'home' and 'health' (e.g. Lang et al. 2019).
Overall, Table 2 shows that 13 out of 14 digital applications (except the digital first notification of loss) provide benefits and opportunities at the contract conclusion as customer touch point. Some digital applications with a focus on back-office functionalities (digital monitoring and telematics, digital ratemaking and underwriting, as well as fraud detection) allow insurance companies to more precisely assess risks and to improve risk selection, leading to a better pricing. The depicted applications can thus be connected to increasing customer satisfaction in the context of insurance in the context of insurance. Further digital applications such as image recognition, digital claims management or smart contracts, however, support to accelerate and automate processes. Consequently, these digital applications can enhance service quality, as well as they support to reduce costs (with the potential to lower premiums), resulting in higher customer satisfaction. Another strand of the digital applications with a focus on back-office functionalities aims to better understand customers (e.g. customer segmentation and sentiment analysis). Moreover, by using gathered information in combination with the digital applications that are assigned to the category of front-office functionalities, such as customer targeting and up-and cross-selling recommendations, insurers are able to increase service quality. Further digital applications with a focus on front-office functionalities do not only allow a more targeted customer approach, but also faster responses and better availability (chatbots or robo-advisory) on the preferred channels (omni-channel approach), also increasing service quality and thus addressing customer satisfaction.
Hence, our results show that digital applications at contract conclusion allow insurance companies to benefit in various ways from refined pricing, faster processes and customer centricity, which positively influence the components of customer satisfaction (service quality, corporate image, as well as price, and consequently the perceived value). Moreover, our results show synergies between certain digital applications (e.g. customer segmentation, customer targeting and up-and cross-selling recommendations), implying that implementing all of these respective digital applications is especially valuable for insurance companies.

Contract modifications as a customer touch point: simplifying and accelerating standard processes
Based on the literature, the opportunities for insurers at the touch point of contract modifications range from improving the customer experience by providing fast services with sufficient capabilities, even at peak times, (e.g. Matouschek and von Hülsen 2015) to improving transparency in relation to customers' policies and the status of specific requests, e.g. by configuring standardised self-service functions (e.g. Matouschek and von Hülsen 2015;Müller et al. 2015). The applications, moreover, allow the simplification of standard processes, such as authenticating a customer employing image recognition (e.g. Shang, 2018). In addition, automation due to digital data availability reduces the manual (time intense) processing times of standard customer requests, which is especially beneficial from the perspective of optimising internal processes (e.g. Matouschek and von Hülsen 2015). Against this background, the results in Table 2 emphasise the relevance of the presented digital applications for the automation of internal processes, which primarily result in economies of time, but also prevent human error (e.g. Sharma 2019).
To summarise, Table 2 illustrates that 6 out of 14 reviewed digital applications provide considerable benefits and opportunities for insurance companies at the touch point of contract modifications. While the applications with a focus on enhancing back-office functionalities (digital ratemaking and underwriting, image recognition, digital claims management, smart contracts) positively affect insurers' key activities by simplifying and accelerating processes, digital applications with a focus on frontoffice functionalities (omni-channel approach with integrated customer portals, chatbots) particularly enable self-services for customers and improve service availability. Consequently, these selected digital applications aim to improve service quality as a component of customer satisfaction. Moreover, they have the potential to reduce costs because of increasing automation. Thus, the presented applications contribute to lower premiums that address price as a component of customer satisfaction.

Event of damage as a customer touch point: creating intuitive processes for customers, with a positive impact on efficiency
The main opportunities regarding the event of damage are reflected in the digital (and/or automated) assessment of loss and repairment costs, claims management and claims settlement, which provide the basis for creating fast and intuitive processes for the customer (e.g. Brüggemann et al. 2018;EIOPA 2019;Helfand 2017). 6 For instance, handling claims can be complemented with further applications, such as image recognition or speech recognition for assessing or settling claims or for processing automated payments (e.g. Brüggemann et al. 2018;EIOPA 2019). 7 However, due to the previously mentioned, numerous evaluable streams of customer data, a major opportunity is reflected in enhancing service management (e.g. Catlin et al. 2018): the interplay between the digital applications, e.g. digital monitoring and telematics, supports the shift from insurers' role of indemnifying losses to loss prediction and prevention (e.g. EIOPA 2019). 8 In the context of customer satisfaction, the event of damage, in particular, as a major customer touch point, can be eased by the availability of (automated) additional services (e.g. Behm et al. 2019;EIOPA 2019).
Moreover, the results in Table 2 point to (digital) fraud detection, which is especially relevant, amongst further digital applications with a focus on back-office functionalities, in terms of increasing the efficiency of operations, also as a result of (partial) automation. Digital fraud detection and prevention lead to reduced operational costs, as a result of reduced fraudulent claims, but also due to released internal capacities (e.g. Deloitte 2017; EIOPA 2019; Gruhn 2018).
Overall, 9 out of the 14 digital applications provide benefits and opportunities for the event of damage as a further customer touch point (see Table 2). One strand of the applications with a focus on back-office functionalities supports insurers with reducing fraud (digital monitoring and telematics, fraud detection, image recognition, sentiment analysis) and positively affects price as a component of customer satisfaction. In this regard, the mentioned digital applications have numerous synergies; for instance, image recognition and sentiment analysis may provide valuable information for a fraud detection system. Besides, digital applications with a focus on back-office functionalities aim to automate and accelerate processes (digital claims management, smart contracts), while the applications with a focus on front-office functionalities particularly support customers in claims registration with the aim of increasing process transparency (chatbots, digital first notification of loss, omnichannel approach with integrated customer portals). For this reason, there is a clear interlinkage to increasing service quality as a component of customer satisfaction. Table 2 additionally emphasises the opportunities with regard to further contacts with customers: one major aspect refers to the identification of hidden patterns (in line with the results at the touch point of contract conclusion), which impact the measures to steer customer demand, in order to precisely attract specific target groups, e.g. customers with a high up-or cross-selling probability (e.g. Eling and Lehmann 2018;Heo and Grable 2017). Targeting customers with personalised content allows for the compilation of personalised policy suggestions, e.g. consumption-based offerings on smartphones (Eling and Lehmann 2018;EIOPA 2019;Helfand 2017;Lehrer et al. 2018). With regard to the specific touch point, the digital applications also focus on extending the (relevant) communication with a customer over the event of damage, by generating more occasions to interact with the customer. Consequently, the utilisation of the applications helps increase the density of customer interaction (e.g. EIOPA 2019; Spender et al. 2019). Moreover, the literature points to the relevance of collaborations with external partners (e.g. Google as a tech company), which support access to non-insurance customer data to consequently attract customers (e.g. Venkatesh 2019). For instance, an ecosystem strategy is based on the shift of "acquiring customers for a specific financial product and then cross-selling other financial products over the customer lifetime" to "acquiring users with primary needs and then converting them into a customer with financial needs over the user lifetime" (Gupta and Tham 2018, p. 435). 9 As Table 2 shows, 10 out of 14 digital applications provide benefits and opportunities at the touch point of further contacts. Digital applications with a focus on back-office functionalities, as well as those with a focus on front-office functionalities might, again, positively affect a customer's perception of service quality that is, for instance, due to advanced, consistent and transparent digital advice, fluent data management or the creation of explicitly addressable customer groups with automated updates for an insurance company.

Discussion of limitations and requirements from the perspectives of customers and insurers
The results of the previous section display the potential opportunities for increasing customer satisfaction at each of the major customer touch points. Our results reveal that by implementing digital applications, insurance companies can improve customer satisfaction in various ways. However, one has to take into account that the implementation of these digital applications comes with limitations and requirements from the perspective of customers as well as insurers.
Besides the benefits of more precise risk assessment and improved risk selection, as shown in the previous section, customers often have data privacy concerns in this regard (e.g. BaFin 2018; Venkatesh 2019). In certain cases, a defensive attitude of customers toward data sharing can be observed (e.g. BaFin 2018; Behm et al. 2019). Reasons for this include ethical issues and fairness concerns due to potential discrimination (e.g. BaFin 2018; EIOPA 2019; SCOR 2018). Moreover, the set of digital applications comes with increased cyber risks and risks posed by the IT. For instance, from a customer's perspective, personal data utilisation might not always be clear (EIOPA 2020) and consequently impair the potentially positive impact on customer satisfaction. Even though there are very strict regulatory requirements, e.g. GDPR compliance with the regulation of using personal data (e.g. BaFin, 2018; EIOPA 2019), and potential legal requirements for data acquisition (e.g. Helfand 2017), black boxes due to the use of potentially unclear algorithms substantiate the necessity of employing trustworthy and explainable artificial intelligence (e.g. Franke 2019; Helfand 2017). Hence, in order to be able to exploit the full potential of increasing customer satisfaction due to digital applications in this regard, insurance companies have to consider the specific needs and concerns of their customers, find the right balance, and are challenged to provide a sufficient degree of transparency to customers.
Moreover, while e.g. chatbots, robo-advisory or an omni-channel approach with integrated customer portals can accelerate responses, improve availability, or enable self-services, the option for human interaction at each of the customer touch points remains important (e.g. Eckert et al. 2021;Müller et al. 2015), especially when considering customers who are less familiar with the use of digital technologies. Human interaction is, for example, relevant for the touch point of contract conclusion and the case of purchasing complex products such as life insurance policies. In this context, a robo-advisory, for instance, might not entirely replace human advice; inaccurate algorithms can cause errors that reduce the quality of the opportunities when managing customer satisfaction in practice (e.g. EIOPA 2020). To manage customer satisfaction, it is important for insurance companies to provide customers with the possibilities afforded by these digital applications but also to continue to be available in person if necessary. Insurers are required to define a balanced mix of utilising digital applications and performing physical processes, especially at interfaces with customers. Improving customer experience or accelerating operations are rather a result of combining the advantages of the digital and analog worlds instead of focusing solely on automating or digitizing them (e.g. Müller et al. 2015).
From the perspective of insurance companies, the employment of digital applications comes with specific risks (see Fig. 1). These risks require consideration, as (a) digitalization considerably intensifies their relevance (e.g. EIOPA 2020) and (b) they directly affect the relationship with customers. 10 In what follows, we provide a list of the overarching concerns and issues when implementing digital applications to improve customer satisfaction from the perspective of insurers: • Insufficient data availability, data access and data quality 11 • Data security and data protection issues 12 • Challenges around data processing as a result of a huge amount of (continuously) changing data 13 and portability of data 14 • Lack of compatible datasets to collaborate or combine insights from different data sources (e.g. relevant in the context of fraud detection) 15 • Lack of accurate and reliable data models 16 and resistance to manipulation 17 • Reputational risks 18 and increasing exposure to cyber risks 19 Furthermore, the implementation of digital applications leads to significant change for insurance companies and entails high investment costs, especially regarding IT (e.g. Eckert and Osterrieder 2020). Apart from a lack of appropriate IT infrastructure and required facilities, the literature points to limitations due to data utilisation, the related customer attitude and regulatory requirements. For instance, as digital/automated claims management is closely related to big data and artificial intelligence, major limitations are posed by regulatory requirements, a lack of policyholders' trust with regard to the technology and data access, the necessity of precise data modelling, issues related to fairness or the ethical dilemmas surrounding black boxes, cyber risks and the related reputational risks for an insurer (e.g. BaFin 2018; EIOPA 2019, 2020). Moreover, in the context of improved customer targeting through big data, insurers need to increasingly access private customer information, which can be perceived negatively by the involved insureds (e.g. Venkatesh 2019). However, using social media data is controversially discussed; insureds are only partially poised to share social media information with insurers (e.g. Spender et al. 2019). General limitations are also a result of insufficient data quality. While data can either be entered manually or are automated, errors may occur due to human but also technical failures (Xu and Hoesch 2018). Apart from the presented limitations, further concerns for the implementation of each digital application also incorporate skilled worker shortage, general project risks, as well as issues of compatibility with the corporate culture (e.g. BaFin 2018; EIOPA 2019).
To utilise the previously explained opportunities at each customer touch point, insurers must consider numerous requirements, based on the literature. For this reason, Table A1 in the Appendix provides a set of major requirements, clustered for each digital application.

Conclusion
In this article, we comprehensively analyse a set of digital applications, including their benefits for insurers, when managing customer satisfaction. Based on an extensive analysis of the literature, we first present the digital applications with their characteristics and state their relevance by clustering their beneficial effects in relation to customer satisfaction at each of the four customer touch points, namely, contract conclusion, contract modifications, the event of damage as the moment-of-truth and further contacts. In addition, we focus on the limitations and address specific concerns that accompany the utilisation of digital technologies (e.g. artificial intelligence, cloud computing). Based on this, we address the specific requirements for each digital application.
The main findings substantiate how the set of digital applications strengthens the position of sales and marketing, simplifying and accelerating (standard) processes, along with creating intuitive processes for customers, increasing efficiency and enhancing the density of customer interaction. In the context of customer satisfaction in the insurance industry, the results indicate the positive effects on a customer's perception of service quality along with the perceived value, which have been previously defined as the components of customer satisfaction within the theoretical framework of the paper. Moreover, we emphasise the opportunity to offer more attractive pricing for the customer, which might affect the perception of price as a further driver of customer satisfaction.
However, the implementation of digital applications results in high investment costs for insurance companies. As managing customer satisfaction poses a major strategic challenge, we emphasise the strategic relevance of defining a selected portfolio of digital applications over the course of an insurer's holistic digital strategy. The strategic implications from the depicted 'opportunity landscape' thus depend on each specific insurer, as, for instance, the service leader might prioritise digital applications to foster service quality or the perceived value, while cost leaders might prefer to design their business processes (in general, but also at the customer touch points) more efficiently, with an impact on price.
Currently, the academic research on customer satisfaction management in insurance companies is rather limited, especially with a focus on the opportunities resulting from digital transformation. While our paper provides an overview of opportunities for customer satisfaction management based on digital applications, further research should empirically investigate whether (and to which degree) customer satisfaction of insurance companies that increasingly implement the presented digital applications improves. Moreover, since transparency is an important factor for insurance customers, future research should particularly focus on explainable artificial intelligence to increase the transparency of digital applications and therefore win the trust of customers and fully exploit the potential of digital applications.

Table A1
Major requirements for the implementation of digital applications (see Table 1 with accurate data models e,h • Assurance of sufficient data storage capacities g,i • Insurance products that take into account data from the respected devices/sensors, e.g. pricing i,n Digital ratemaking and underwriting • Collection of comprehensive data regarding the applicant and automation of the respective processes i along with the trained models to apply, e.g. artificial intelligence (descriptive, predictive and prescriptive analytics tools) • Appropriate IT architecture g Fraud detection • Implementation of technologies to interlock existing data with the aim of developing comprehensive customer databases (including past data) j • IT solutions to centrally store policy-related documents of customers with insurance line overarching databases to create a 'global picture' on a customer's (potentially fraudulent) behaviour j • Development of trained models (based on previous data) especially for motor and health insurance as these are prone to fraudulent claims h,j • Establishment of internal structures to share best practices of fraud management, but also to centrally point to conspicuous claims j Image recognition • Availability of sufficient data: data collection to train the underlying models i,k • Accurate (analytical) and trained models to operate the underlying machine learning algorithms i,k Sentiment analysis • Accurate (analytical) models trained on previous data f • Design of business processes, which integrate the results of the application resp. underlying the models h Back-office and front-office functionalities Digital claims management • Installation of a comprehensive digital customer journey for the claims process h • Implementation of big data and related big data analytics approaches, including artificial intelligence h,i,r,t • Suitable infrastructure and facilities to implement related notifications, including the first notification of loss, to manage claims digitally and to settle them d,h Robo-advisory • Installation of accurate and robust algorithms p to allow 'intelligent' interaction with customers c Smart contracts • Development of appropriate product design s • Implementation of blockchain technology/accessibility of blockchain along with the requirements for IT g Customer targeting • Access to private/personal data of a customer t , e.g. the possibility to collect data based on the Internet of Things • Integration of the application for existing marketing and sales activities (e.g. a more precise calculation of a customer lifetime value, with enhanced, digital customer segmentation) h Digital first notification of loss • Installation of the required hardware/sensors to document and report potential losses, e.g. with regard to automobile insurance h • Integration of the notification to business processes, e.g. via smartphone applications h Omnichannel approach (integrated customer portals) • Online/mobile access to the products and related services o • Accurate data management to assure robustness and sufficient capabilities to process input of numerous channels and to install the appropriate analytics tools m,o • Adjustment/transformation of existing capabilities with regard to customer service centers and equipment/ training of salespersons to link the digital technologies to their processes o Up-and cross-selling recommendations • Collection of behavioural information on the customer (across a customer's lifetime or previously) so as to train accurate models based on big data analytics/artificial intelligence (machine learning) h a Abdul-Kader and Woods (2015) Venkatesh (2019)