Identifying the necessary elements in audit quality for customer retention

There is a generally accepted assertion that client satisfaction is a necessary condition for client loyalty (Bloemer et al., 1998), with behavioral drivers playing a fundamental role in this satisfaction (Reheul et al., 2013). In the context of auditing, Morton and Scott (2007) highlighted the critical role of audit service quality in fostering client retention and loyalty. Despite numerous studies addressing this issue (Duff, 2009; Ghebremichael, 2019; Sampet et al., 2019, among others), the findings remain inconsistent. A notable gap in the literature is the frequent conflation of necessity and sufficiency (Dul, 2016). Previous analyses have examined whether various aspects of audit quality positively affect client retention and satisfaction. However, according to Richter et al. (2020), while a determinant might be sufficient to produce the outcome, it may not be necessary. The absence of a determinant could be compensated for by other variables. Conversely, if a determinant is necessary, the effect cannot be achieved without it. This distinction raises several critical questions: Are these dimensions of quality significant? Should auditing firms focus only on those aspects of quality perception that have been empirically validated? And, most importantly, which dimensions are necessary? If any of these necessary dimensions are lacking, client perception and loyalty are likely to fail.

This article seeks to address these questions by exploring how perceptions of audit quality contribute to stronger client loyalty and identifying the necessary conditions for this relationship to persist. We argue that audit quality is a cornerstone of competitive advantage, serving as a key differentiator in client retention (Awwad et al., 2024; Hackenbrack and Hogan, 2005; Krishnan and Schauer, 2000). In line with this, there is a growing recognition of auditing as a service-oriented profession (Knechel et al., 2020). Consequently, the quality of audit services must be evaluated from the perspective of the service recipient, the client, while auditors strive to deliver superior service to corporate executives (Ricci, 2022). This necessitates a broader approach to audit quality analysis, one that transcends economic attributes (Morton and Scott, 2007) and encompasses the entire client-auditor interaction (Aghazadeh et al., 2022). In this regard, the quality of audit services emerges as a pivotal element in the competitive strategy of audit firms (Awwad et al., 2024; Ghebremichael, 2019; Ricci, 2022). In this sense, we advance the traditional conceptualization of quality within the auditing field, as initially proposed by DeAngelo (1981), which embraces the perspective of technical quality (Grönroos, 1984).

Although prior studies have explored this service-oriented perspective (Butcher et al., 2013; Reheul et al., 2013; Hackenbrack and Hogan, 2005; Ismail et al., 2006; Knechel et al., 2020; Morton and Scott, 2007, among others), their findings remain inconclusive due to inconsistent empirical results and methodological limitations. While some studies confirmed the relevance of certain SERVQUAL dimensions—such as reliability, responsiveness, and assurance—others found only partial or weak associations, with dimensions like tangibles and empathy often showing no significant effect. These inconsistencies largely stem from the predominant reliance on sufficiency-based models, which assess whether a factor can produce an outcome but not whether it is indispensable. This diversity of results explains the concerns raised by Morton and Scott (2007) and Butcher et al. (2013) about the universal relevance of SERVQUAL dimensions in understanding customer retention. Our findings explain how this lack of consensus does not diminish the importance of quality management efforts by audit firms to enhance client perceptions and retention. Rather, it underscores the need for a more nuanced understanding of the necessary conditions underlying these relationships. Our study addresses this gap by integrating Necessary Condition Analysis (NCA) with PLS-SEM, enabling us to distinguish between sufficient and necessary conditions. This dual approach contributes not only methodologically but also conceptually and managerially, offering a more precise and actionable framework for understanding and managing audit service quality. Future research should continue to explore the interplay between these factors, providing actionable insights for practitioners and policymakers alike.

The primary objective of this study is twofold: first, to evaluate the importance of audit quality, viewed through a service lens, in determining auditor tenure; and second, to identify the necessary conditions among the various elements of audit quality that influence client retention. To achieve this, we build upon the SERVPERF scale (Cronin and Taylor, 1994) as a robust tool for assessing quality perceptions. The SERVPERF scale is well-established in the broader service quality literature (Bruhn, 2023; Park et al., 2023) and has been specifically applied to audit services (Duff, 2004; Ismail et al., 2006; Morton and Scott, 2007).

This paper is structured as follows. First, we review the literature to examine the various approaches used to measure audit quality and the findings they have yielded. Drawing on this evidence, we propose a model that links client loyalty to perceptions of audit quality, establishing hypotheses from a necessary condition perspective. Next, we outline the research methodology and empirical context. The subsequent section presents the results of our analysis, highlighting the relationship between audit quality and client loyalty and identifying the necessary conditions for each dimension of quality. Based on these findings, we discuss implications for audit firms and offer suggestions for future research.

Audit quality has been a central topic in auditing research, with numerous studies attempting to define and measure it. DeAngelo (1981) provided one of the most widely accepted definitions, stating that audit quality is the probability that an auditor will both discover and report breaches in the client’s accounting system. This definition emphasizes the dual role of auditors in detecting errors and communicating them to stakeholders, framing quality as an objective and measurable attribute from the auditor’s perspective. However, this view has been critiqued for its narrow focus on technical aspects, overlooking the relational and service-oriented dimensions of auditing (Montenegro and Brás, 2018).

The literature has explored various factors influencing audit quality, including firm size (Krishnan and Schauer, 2000), the provision of non-audit services (Simon, 1997), audit fees (Butcher et al., 2013), and the characteristics of the audit team (Kilgore et al., 2011). Despite these efforts, a significant portion of the research has focused on internal firm attributes, often neglecting the fact that auditing is fundamentally a service-oriented profession (Seckler et al., 2017). This technical quality perspective, rooted in Grönroos’s (1984) service quality framework, has dominated the discourse, emphasizing measurable outcomes over client perceptions (Ghebremichael, 2019).

However, audit quality is inherently multidimensional and subjective, varying across stakeholders (Wooten, 2003). Montenegro and Brás (2018) argue that client perceptions are critical to understanding the relational and service dimensions of audit quality, which are not fully captured by technical measures alone. This perspective aligns with Knechel et al. (2020) and Reheul et al. (2013), who advocate for a broader service-oriented view of auditing, incorporating elements such as the client-auditor relationship and its impact on audit value and efficiency. While auditing standards aim to standardize quality levels, client-specific needs and expectations often diverge from these standards, highlighting the importance of relational dynamics in service delivery (Aghazadeh and Hoang, 2020; Aghazadeh et al., 2022).

Empirical evidence supports the notion that client perceptions of audit quality significantly influence satisfaction and retention (Montenegro and Brás, 2018). This underscores the need to integrate client perspectives into quality assessments, as clients are the ultimate recipients of audit services. Despite this, the service-oriented view remains underexplored in the literature, with few studies linking client perceptions to auditor retention (Duff, 2004, 2009). This gap highlights the necessity of adopting a more holistic approach to audit quality, one that balances technical and relational dimensions.

Client retention is a critical concern for audit firms, as maintaining long-term relationships with clients is essential for financial stability and reputation. The decision to retain an auditor is influenced by a complex interplay of factors, including perceived service quality, satisfaction, and the alignment of client expectations with audit outcomes (Francis, 2004; Behn et al., 1997). Research has demonstrated that client retention is not solely determined by technical audit quality but also by the relational and service aspects of the audit process (Ciconte et al., 2022).

The marketing literature provides valuable insights into client retention. This literature underscores that customer satisfaction and loyalty are fundamental for establishing long-term relationships with consumers. High-quality service not only increases satisfaction but also strengthens loyalty and promotes client retention over time. This is achieved by consistently improving key service aspects, such as reliability, responsiveness, and personalized attention. This has been studied particularly through service quality models such as SERVQUAL (Parasuraman et al., 1985, 1991) and SERVPERF (Cronin and Taylor, 1992). These models emphasize the importance of client perceptions in shaping satisfaction and loyalty. In the auditing field, Ismail et al. (2006) found that client satisfaction mediates the relationship between service quality and loyalty, suggesting that satisfied clients are more likely to remain with their auditors. Similarly, Pandit (1999) identified responsiveness to client needs and executive involvement as key drivers of retention, while Morton and Scott (2007) observed a weak but significant link between service quality and retention.

The expectancy confirmation theory (Oliver, 1980) further supports the idea that client loyalty is contingent on the alignment between client expectations and perceived service performance. This theory has been applied in auditing research to explain how discrepancies between client expectations and audit outcomes can lead to dissatisfaction and auditor switching (Ismail et al., 2006).

Recent studies have also highlighted the role of trust and communication in fostering client loyalty. For example, Mainardes and Sousa (2022) found that trust in the auditor’s expertise and the quality of client-auditor communication are significant predictors of retention. These findings align with the broader service marketing literature, which emphasizes the importance of relational dynamics in sustaining long-term client relationships (Bloemer et al., 1998; Stauss and Neuhaus, 1997).

Furthermore, research by Aghazadeh et al. (2022) suggests that persuasion and effective communication are crucial for client retention in auditing. The integration of contingency information in judgments of cause and probability, as discussed by Mandel and Lehman (1998), can also influence perceptions of service quality and, consequently, client retention. McEnroe and Martens (2001) address the expectation gap between auditors and investors, which can affect client satisfaction and loyalty. Montenegro and Brás (2018) provide a comprehensive review of audit quality, highlighting the importance of quality measures in client retention.

The integration of audit quality and client retention represents a critical area of research, as it bridges the technical and relational dimensions of auditing. While traditional approaches to audit quality have focused on measurable outcomes, the service-oriented perspective emphasizes the importance of client perceptions in shaping satisfaction and loyalty (Knechel et al., 2020). This dual focus is essential for understanding how audit firms can deliver high-quality services while fostering long-term client relationships.

Empirical evidence suggests that client perceptions of audit quality are strongly correlated with retention decisions (Montenegro and Brás, 2018). Clients who perceive their auditors as responsive, reliable, and empathetic are more likely to remain loyal, even in the face of technical shortcomings (Ismail et al., 2006). This highlights the need for audit firms to adopt a balanced approach, addressing both technical and relational aspects of service delivery.

The concept of the “expectation gap” (Parasuraman et al., 1985) is particularly relevant in this context, as it captures the discrepancy between client expectations and perceived service performance. Reducing this gap requires audit firms to align their service delivery with client needs and preferences, fostering a sense of trust and satisfaction (Knechel et al., 2020). This can be achieved through effective communication, personalized service, and a commitment to continuous improvement.

The SERVPERF model offers a valuable framework for assessing the intersection of audit quality and client retention. By focusing on performance perceptions, this model provides a more accurate measure of service quality, enabling audit firms to identify areas for improvement and enhance client satisfaction (Cronin and Taylor, 1994). Furthermore, the model’s emphasis on relational dimensions, such as empathy and responsiveness, aligns with the broader service marketing literature, underscoring the importance of client-centric approaches in auditing.

The integration of audit quality and client retention requires a holistic approach that balances technical excellence with relational competence. By addressing both dimensions, audit firms can enhance their service delivery, foster client loyalty, and achieve sustainable growth in a competitive market.

Drawing on the literature and the theoretical framework outlined above, we propose the following hypotheses, grounded in the necessary condition logic and the service-oriented perspective of audit quality. To elaborate these hypotheses, we address the relationship between audit quality dimensions, contemplated by the SERVPERF scale, and client loyalty, as well as the necessary conditions for achieving client retention.

The literature establishes client satisfaction as a critical antecedent to loyalty in auditing, with empirical studies linking satisfaction to retention through relational and technical dimensions (Reheul et al., 2013; Morton and Scott, 2007). Auditing scholars highlight that satisfaction in this context arises not merely from technical competence but also from the auditor’s capacity to align service outcomes with client expectations—a dynamic that fosters trust and sustains long-term relationships (Knechel et al., 2020; Montenegro and Brás, 2018). Building on Taylor and Cronin’s (1994) SERVPERF framework, which conceptualizes service quality as a unidimensional construct, this study posits that client satisfaction—synthesized from multifaceted quality attributes—serves as a robust predictor of loyalty. This relationship is particularly salient in competitive markets where differentiation relies on superior client experience (Ghebremichael, 2019; Ricci, 2022).

Elaborating on the analyses by Ismail et al. (2006) and Reheul et al. (2013), this study examines whether individual components of service quality measurement scales independently and significantly influence client loyalty. This approach will allow for a consideration of the sufficient and necessary conditions of each of the dimensions included in the construct. Prior findings reveal inconsistencies: Ismail et al. (2006) identified significant effects of tangibility, reliability, assurance, and empathy on loyalty, though some coefficients were unexpectedly negative—a phenomenon left unexplained. Conversely, Reheul et al. (2013), employing the SERVPERF scale, found significant effects for all dimensions except tangibles.

Reliability, defined as the ability to perform the promised service dependably and accurately (Zeithaml et al., 2002), is a cornerstone of service quality. In the context of auditing, reliability ensures that clients perceive the audit firm as consistent and trustworthy in delivering accurate financial assessments (Ismail et al., 2006). Prior research has shown that clients value reliability as a critical dimension of audit quality, as it directly impacts their confidence in the audit process (Morton and Scott, 2007). Without reliability, clients are likely to perceive the audit service as unreliable, leading to dissatisfaction and potential auditor switching (Duff, 2009). Therefore, we posit that:

Responsiveness refers to the willingness of the audit firm to help clients and provide prompt service (Zeithaml et al., 2002). In the audit context, responsiveness is critical for addressing client concerns, meeting deadlines, and adapting to client-specific needs (Knechel et al., 2020). Studies have highlighted that clients perceive responsiveness as a key indicator of the auditor’s commitment to their needs (Aghazadeh et al., 2022). Failure to respond promptly or adequately to client requests can lead to dissatisfaction and erode trust, ultimately jeopardizing client retention (Montenegro and Brás, 2018). Thus, we argue that responsiveness is a exerts a positive influence on client loyalty, and its absence undermines the client’s perception of service quality.

Assurance encompasses the knowledge, courtesy, and ability of the audit firm to inspire trust and confidence in clients (Zeithaml et al., 2002). In auditing, assurance is particularly important because clients rely on the auditor’s expertise to validate financial statements and provide credible insights (Ghebremichael, 2019). Research has shown that clients perceive assurance as a critical component of audit quality, as it directly influences their trust in the auditor’s competence and professionalism (Reheul et al., 2013). Without assurance, clients are likely to question the credibility of the audit process, leading to dissatisfaction and potential auditor switching (Hackenbrack and Hogan, 2005). Therefore, we propose that:

Empathy refers to the ability of the audit firm to provide individualized attention and care to clients (Zeithaml et al., 2002). In the audit context, empathy is crucial for understanding client-specific needs, building strong relationships, and fostering a sense of partnership (Ricci, 2022). Studies have demonstrated that clients value empathy as a key dimension of service quality, as it enhances their overall satisfaction and loyalty (Ismail et al., 2006). A lack of empathy can lead to a perception of indifference, eroding the client-auditor relationship and increasing the likelihood of auditor switching (Montenegro and Brás, 2018). Thus, we argue that empathy is a sufficient and necessary condition for client loyalty.

Tangibles refer to the physical evidence of the service, such as the appearance of facilities, equipment, and personnel (Zeithaml et al., 2002). In auditing, tangibles may include the professionalism of the audit team, the quality of reports, and the use of advanced tools and technologies (Knechel et al., 2020). While tangibles are often considered less critical than other dimensions of service quality, they play a significant role in shaping client perceptions of the audit firm’s competence and reliability (Duff, 2004). A lack of attention to tangibles can lead to a perception of unprofessionalism, undermining client confidence and loyalty (Morton and Scott, 2007). Therefore, we propose that tangibles are a fundamental condition for client loyalty.

These hypotheses are grounded both in the sufficient and the necessary condition logic. The sufficient condition assumes that an element, if present, guarantees that a particular outcome will occur. However, the outcome can still occur even if the sufficient condition is not present. However, in the necessary condition logic, the existence of a relationship posits that the absence of a necessary determinant prevents the achievement of the desired outcome, even if other determinants are present (Richter et al., 2020). In the context of our model related to audit quality, each dimension (reliability, responsiveness, assurance, empathy, and tangibles) contributes uniquely to the client’s perception of service quality and their decision to remain loyal to the auditor (Ismail et al., 2006; Knechel et al., 2020). If any of these dimensions are lacking, the overall perception of service quality is compromised, leading to dissatisfaction and potential auditor switching (Montenegro and Brás, 2018). These hypothesis are shown in Figure 1.

To test our hypotheses, the largest Spanish companies were identified from a financial information database (SABI) from Bureau van Dijk. We identified all companies for which auditing is compulsory and provided complete financial and audit information for the past nine years. A questionnaire was sent to all these companies (2,935), obtaining 234 complete and usable responses, with margin of error of ±6% at a 95% confidence level.

The sectoral distribution and descriptive information of the surveyed companies are presented in Table 1. Regarding company size, 25 were classified as small (fewer than 50 employees), 75 had between 50 and 200 employees, and 73 employed between 200 and 500 individuals. The remaining companies were categorized as large, with over 500 employees. All questionnaires were completed by university graduates in managerial positions, with 90% of respondents having held their positions within the company for more than one year.

To control for possible constraints on the results obtained, the possible implications of the regulatory regime concerning auditor retention/rotation were studied. The Spanish regulations on the mandatory auditor/audit firm rotation are included in the Law on Auditing of Accounts (Art. 40 LAC). The rule establishes that the regulation in Spain of the audit contract applies the provisions of Article 17 of Regulation (EU) No. 537/2014, of April 16, according to which the minimum duration of the initial period of engagement of statutory auditors in public interest entities may not be less than three years, and the total period of engagement, including extensions, may not exceed the maximum duration of ten years established in Article 17 of the aforementioned Regulation. However, once the total maximum contracting period of ten years of an auditor or audit firm has expired, this period may be extended additionally up to a maximum of fourteen years under certain conditions. Furthermore, the contract can also be terminated during a shorter period, although such circumstances must be communicated to the Institute of Accounting and Auditing (Instituto de Contabilidad y Auditoría de Cuentas).

In this study, we adapted the SERVPERF model developed by Cronin and Taylor (1994), which encompasses five dimensions of perceived service quality: Tangibles, Reliability, Responsiveness, Assurance, and Empathy. The SERVPERF model, which focuses solely on performance perceptions, has been shown to provide a more robust measure of service quality than SERVQUAL, particularly in diagnostic contexts (Cronin and Taylor, 1994; Carrillat et al., 2007). This makes SERVPERF a suitable tool for examining the relationship between audit quality and client retention. To ensure the scales were contextually appropriate for the specific regulatory environment (Carrillat et al., 2007), we evaluated various iterations of the SERVPERF questionnaire, drawing on prior research that assessed perceived quality in accounting and auditing contexts (Sierra-García et al., 2017; Reheul et al., 2013) (see  Appendix for a detailed questionnaire).

Client loyalty was operationalized as the number of years a firm has retained the same auditing firm. Most participating firms (71%) reported maintaining their relationship with their auditing firm beyond the initial five-year period, while only 14% had switched auditing firms within the last two years.

As previously noted, we adopt two perspectives in our analysis of the audit quality construct: (1) a unidimensional view of quality, as proposed by Taylor and Cronin (1994), and (2) a multidimensional perspective, informed by studies that have examined the impact of service quality dimensions on client loyalty in the auditing industry (Butcher et al., 2013; Ismail et al., 2006; Morton and Scott, 2007). To account for both perspectives, we propose a model that considers quality both as a composite construct (to assess its global effect) and as a set of independent indicators (to align with the predominant approach in the literature).

To identify necessary conditions through Necessary Condition Analysis (NCA), we followed the methodological guidelines established by Richter et al. (2020). These guidelines suggest that NCA can be extended to latent constructs by computing factor scores or composite scores derived from the indicators used to measure these constructs (Tuesta-Tapia et al., 2024). Consequently, based on the results obtained from our Partial Least Squares Structural Equation Modeling (PLS-SEM) analysis using SmartPLS 4.0 (Ringle et al., 2024), we applied NCA using R and the NCA package (Dul, 2021; Tuesta-Tapia et al., 2024).

PLS-SEM is a multivariate analysis technique that enables the investigation of causal-predictive relationships in path models with latent variables, as well as the estimation of these relationships (Khan et al., 2019; Richter et al., 2016; Hair et al., 2012; Ringle et al., 2012). In our PLS-SEM model, the strength of the relationships between latent variables reflects the impact of antecedent constructs on our target construct: client loyalty to audit firms.

Our model, grounded in the SERVPERF framework (Cronin and Taylor, 1992), seeks to identify the relationship between audit quality and client loyalty. This model posits that higher perceived quality increases the likelihood of clients maintaining their relationship with their auditing firm. The interpretation of our PLS-SEM findings follows a sufficiency logic (Mandel and Lehman, 1998; Dul, 2016). Thus, using PLS-SEM, we can identify the essential factors (i.e. Service Quality dimensions) that auditing firms must prioritize to foster client loyalty (Richter et al., 2020).

As Richter et al. (2020) argue, “While, according to sufficiency logic, a determinant (e.g. benefit) may be sufficient to produce the result (e.g. the use of a technology), it may not be necessary. The absence of benefit could be compensated for by other determinants, such as a positive evaluation of the technology’s usefulness.” Following this reasoning, our results must demonstrate the interplay between necessary and sufficient conditions to provide a comprehensive understanding of the factors driving client loyalty in the auditing context.

Our research model associates the effect of quality on auditor retention. In this regard, we adopt a dual perspective, with one assuming the one-dimensionality of the quality construct from the SERVPERF model (Cronin and Taylor, 1992). This was studied as a higher order construct composed of five dimensions, which was calculated using the integrated two-stage approach (Sarstedt et al., 2019). However, the first stage of the previous model is used to assess the individual effects, as proposed in the work of Ismail et al. (2006), Butcher et al. (2013), or Morton and Scott (2007).

First-level constructs were estimated in Mode A (reflective), that is, they were evaluated by applying traditional measures of internal consistency, reliability, and validity (Henseler et al., 2016). As a result, most of our indicators, except for one item linked to tangible elements, which was refined, and dimensions, met the reliability requirement, as most of the values exceeded 0.7. Furthermore, the composite reliability (C.R.) measure indicated that all constructs are reliable, since they were all above 0.7, and all constructs reached convergent validity, since their average variance extracted (AVE) values were above 0.5 (Hair et al., 2019) (Table 2). The heterotrait-monotrait ratio of correlations (HTMT) discriminant validity test showed values that remained below the threshold of 0.9 (Hair et al., 2019) in all cases.

In the other perspective, the estimation of the quality construct as a second order construct. Following the two-step procedure aforementioned, after confirming the reliability and validity of the first-order constructs, the second-order construct is estimated using the scores generated by the model. To ensure convergent validity of the second-order construct, the values obtained for the AVE exceed 0.5 and the C.R. is above 0.7, indicating that the dimensions adequately reflect the overall construct (Hair et al., 2019). For discriminant validity, criteria such as the Fornell-Larcker criterion are applied, where the square root of the AVE of the second-order construct must be greater than its correlations with other constructs, as well as the HTMT matrix, whose values should remain below 0.90 to confirm that the second-order construct is distinct from others in the model (Hair et al., 2019). The results obtained fulfill these thresholds and ensure the robustness and applicability of the results obtained (Table 3).

The results of the PLS-SEM model showed that the dependent constructs presented variance inflation factor values lower than 3.3. This result indicated that there were no multicollinearity issues (Hair et al., 2019). Bootstrapping, using 10,000 subsamples as suggested by Ringle et al. (2018) and Streukens and Leroi-Werelds (2016), showed that the path coefficients hypothesised were statistically significant (Table 4). This procedure was performed for the first-order model, which uses the different individual constructs.

These results lead us to accept H1, but to reject all its associated sub-hypotheses (H2a to H6a). The only thing that these contradictory results show is the reality of the one-dimensionality of audit quality. This means that quality should not be understood as a sum of unconnected elements, but rather as a construct resulting from the joint effect of all these dimensions (Taylor and Cronin, 1994). The suitability of the reflective measurement models was also verified by means of goodness-of-fit indicators (Henseler et al., 2016; Benitez et al., 2020; Henseler and Schuberth, 2020). The approximate fit measure of the model shows that the model has a good fit with a Standardised Root Mean Squared Residual (SRMR) of 0.010 (a level below the threshold of 0.08; Hu and Bentler, 1998, 1999). Furthermore, the bootstrap-based exact fit tests show results below the levels and therefore, the model cannot be rejected from a confirmatory point of view (Henseler et al., 2016). In this case, empirical evidence would be found for the postulated model. It would be possible for the empirical data to come from a world that works, as theorised in the model. The results of the sufficiency hypothesis can be seen in Figure 2.

A complementary analysis was carried out to determine whether size had an influential role in the model (Sinason et al., 2001). Specifically, we tested two alternative specifications: in the first, size was modelled as having a direct effect on both the independent and dependent variables; in the second, size was evaluated as a potential moderator of the relationship between them. The results indicate that the moderating effect is not statistically significant (β = 0.073, p = 0.344), while the direct effects yielded mixed results—significant for the effect on loyalty (β = −0.289, p = 0.004) and non-significant for the effect on the satisfaction construct (β = −0.130, p = 0.217).

These findings suggest that client size has a direct negative effect on loyalty, independent of satisfaction levels. This result contrasts with prior literature that links firm size to lower auditor-switching rates (Sinason et al., 2001).

Necessary Condition Analysis (NCA) tests whether specific dimensions of Service Quality (Tangibles, Reliability, Responsiveness, Assurance, and Empathy) are necessary determinants of Loyalty to auditing firms. Logical necessity implies that the desired result, or a certain level thereof (loyalty), can only be achieved if the necessary cause is present or is at a certain level. In other words, without such cause, the dependent variable will not exist (Dul, 2016). Furthermore, the absence of a necessary condition cannot be compensated by other determinants, therefore, in its absence, there is no result (Dul, 2016). In conclusion, NCA provides a valuable additional understanding of the critical role of the Service Quality dimensions (i.e. Tangibles, Reliability, Responsiveness, Assurance and Empathy). NCA determines the essential, or bottleneck, factors for auditing firms to have a solid base of loyal clients with long-term contractual relationships (i.e. Loyalty). The scatter diagram (Figure 3), which represents a necessary, but not sufficient, relationship between the independent variables, expressed as continuous variables, and the dependent variable (measured as years of auditing). The graphical results show notable areas that do not contain data points (see the upper left corner in each of the figures). The spaces where the data do not appear indicate that the variable constrains the possible levels of loyalty. This space is called the ceiling zone and is separated from the area containing data by a line called the ceiling line. The ceiling line can be represented in various ways, depending on the statistical assumptions of the individual variables and their distribution. Dul (2016) presents two techniques: Ceiling Envelopment – Free Disposal Hull (CE-FDH) and Ceiling Regression – Free Disposal Hull (CR-FDH). CE-FDH draws an appropriate step function ceiling line for data which is discrete or categorical; CR-FDH creates a regression line through the corners of the CE-FDH line and is usually appropriate for continuous data (Dul et al., 2018).

To determine the must-have or bottleneck factors, NCA determines two key parameters: ceiling accuracy or c-accuracy and the d-necessity effect size. The ceiling accuracy represents the percentage of observations that are above or below the line. The CE-FDH lines are drawn so that all observations must be on or below the line (100% accuracy). CR-FDH lines are drawn so that most observations lie on or below the line. Even when there is no specific threshold that determines the acceptable level of accuracy, its comparison with a reference value (for example, 95% of the estimated accuracy) can help evaluate the quality of the generated solution (Dul, 2016). The effect size of the necessary condition (d) is the area of the ceiling zone (the empty space), divided by the total area, with possible observations (called the scope). Dul (2016) has proposed some initial levels of effect size such that the necessary condition is considered small, if d < 0.1, medium, if the effect sizes are 0.1 ≥ d < 0.3, and large, if the effect sizes are d > 0.3. To prevent NCA effects from being the result of empty space produced by unrelated variables, or variables biased substantially positively and negatively at the same time (Sorjonen et al., 2017), Dul and colleagues developed a statistical significance permutation test of the NCA effect size (Dul et al., 2018, 2019). The p-value resulting from the significance test is interpreted using traditional thresholds. However, before the development of this significance test, Dul (2016) suggested d > 0.1 as a threshold to consider whether a necessary condition was theoretically or practically significant.

The NCA results (see Table 5) indicate that all dimensions of service quality (except for tangibles) are a meaningful (d ≥ 0.1) and a significant (p < 0.05) necessary condition, and therefore a must-have or bottleneck factor, without whose presence Loyalty to the auditing firm will not exist. These results confirm H2b, H3b, H4b, and H5b, leaving the hypothesis about the need for tangible elements (H6b) unconfirmed.

Information from the bottleneck tables (Table 6) allows a more detailed analysis of these must-have factors to be performed. In fact, it is highlighted how, up to a certain duration of the relationship with the auditing firm (up to two years), none of the service quality dimensions are necessary. However, after two years, if the auditing firm wants to have a solid base of loyal clients, five necessary conditions must be met: 25.4% Tangibles, 52.9% Reliability, 41.0% Responsiveness, 36.5% Assurance, and 40.3% Empathy. Therefore, the auditing firm must guarantee that these service quality dimensions are clearly perceived by its clients. In addition, as can be seen in the Bottleneck tables, it does not make sense for the auditing firm to dedicate more resources to raising the Service Quality Perception in any of the dimensions, except in the Assurance dimension, which becomes the Bottleneck Factor, if the auditing firm wants to achieve Loyalty levels beyond three years (4 or 5 years). The proposed relationships would fit within the third scenario of the classification proposed by Richter et al. (2020), that is, our results show non-significant relationships but with a necessary condition. Because of this situation, it is not advisable for an auditing firm to invest excessively in the dimensions beyond the levels expressed, as this is unlikely to lead to increases in the dependent variable, in this case, the years of contracting.

Our work sought to assess the necessary conditions of the dimensions of client satisfaction. The findings of this study contribute to the growing body of literature that views auditing as a service-oriented profession (Knechel et al., 2020; Ricci, 2022). The results underscore the importance of adopting a holistic approach to audit quality, where the interplay of multiple dimensions collectively influences client loyalty. This aligns with previous research that emphasizes the multidimensional nature of audit quality and the need to consider both technical and relational aspects (Montenegro and Brás, 2018; Ghebremichael, 2019).

The results obtained highlight that while individual dimensions of SERVPERF do not independently predict client loyalty, their combined effect is significant. This suggests that audit firms should focus on delivering a balanced service that integrates reliability, responsiveness, assurance, and empathy, rather than prioritizing one dimension over others.

Consistent with previous work (Ghebremichael, 2019; Butcher et al., 2013; Pandit, 1999), our results show that client loyalty is conditioned by their perception of quality. However, moving beyond prior results (Morton and Scott, 2007), where the relationship between audit service quality and auditor retention was considered weak.

The absence of a quality effect is interesting in the early years of the relationship. According to the results, as the end date of the initial contract period approaches, quality becomes more relevant, but without the levels of the different dimensions being seen as sufficient individually (Butcher et al., 2013).

The NCA results provide a novel perspective by identifying the necessary conditions for client loyalty. The finding that reliability, responsiveness, assurance, and empathy are necessary but not sufficient conditions aligns with the expectancy confirmation theory (Oliver, 1980), which posits that client loyalty depends on the alignment between client expectations and perceived service performance. The absence of any of these dimensions would likely result in client dissatisfaction and potential auditor switching, as clients cannot compensate for the lack of these critical elements (Dul, 2016).

Consistent with prior research (Ismail et al., 2006; Morton and Scott, 2007), tangible elements appear to be less critical for clients, who often assume that audit firms possess the necessary resources to perform their function.

Perhaps the most important conclusion is that loyalty, per se, is not based on one attribute or another; rather, the NCA results indicate that there are certain critical determinants that traditional models have not been able to identify. Different dimensions are now identified as representing a necessary but not sufficient condition for the generation of client loyalty. It is therefore necessary that there be a minimum level of all the attributes to guarantee loyalty, without firms having to focus on one particular attribute. Therefore, the auditing company must ensure that these dimensions of service quality are clearly perceived by its customers. Furthermore, results suggest that it does not make sense for the auditing companies to devote more resources to raising the quality perception of their service in any of the dimensions, except for the assurance dimension, which becomes a bottleneck factor if the auditing company wants to maintain higher levels of loyalty after 4 or 5 years of relationship.

Consequently, auditing firms must guarantee a level that favours the maintenance of the relationship established with the client, without higher levels of the dimensions, ensuring greater loyalty from the client. As seen in previous studies (Ismail et al., 2006), the assurance dimension is a fundamental element for the consolidation of the auditor-client relationship, based on quality.

Interestingly, tangibles were not found to be a necessary condition for client loyalty. This finding is consistent with prior research that suggests clients assume audit firms possess the necessary resources to perform their functions, and thus, tangible elements do not significantly influence retention decisions (Ismail et al., 2006; Morton and Scott, 2007). However, this does not imply that tangibles are irrelevant; rather, they may play a secondary role in shaping client perceptions.

From these results, this study clarifies that while audit quality is a multidimensional construct, not all dimensions are equally critical. However, and despite the fact that they cannot determine client retention individually, some become essential thresholds that must be met to sustain long-term client relationships. This explains why Morton and Scott (2007) and Butcher et al. (2013) questioned the universal relevance of dimensions such as responsiveness and empathy. Our results demonstrate that while these dimensions may not be sufficient on their own, they are indeed necessary for client loyalty and therefore clarifying their role in this relationship.

Moreover, the negative effect of firm size on loyalty, independent of satisfaction, reinforces the idea that structural or organizational factors can shape client retention behaviors in ways not captured by service quality alone. This underscores the importance of considering contextual variables when interpreting the dynamics of auditor-client relationships.

The findings of this study offer clear and actionable guidance for audit firms seeking to strengthen long-term client relationships through strategic and data-driven quality management. First, audit firms should prioritize building strong relationships with clients by ensuring high levels of reliability, responsiveness, assurance, and empathy. These dimensions are critical for fostering trust and satisfaction, which are essential for long-term client loyalty. Thus, it becomes essential for audit firms to implement internal quality control systems that monitor each dimension individually rather than relying solely on overall satisfaction metrics.

Second, the results underscore the importance of adopting a balanced, lifecycle-based approach to service delivery by audit firms, ensuring that all necessary dimensions of audit quality are adequately addressed. Assurance, in particular, emerges as a bottleneck factor in long-term relationships, gaining increasing influence beyond the early years of engagement. Overinvesting in one dimension at the expense of others may not yield additional benefits in terms of client retention. This highlights the need for targeted interventions for the efficient allocation of resources in the provision of audit services.

Third, the results suggest that disproportionate investment in certain service dimensions will not significantly enhance loyalty. For instance, overinvesting in non-critical dimensions like tangibles—such as branding, office aesthetics, or digital interfaces—which may enhance image but have limited impact on loyalty unless core expectations are fulfilled. Instead, firms should strategically allocate resources toward developing soft skills, including communication, empathy, and proactive client engagement, which directly influence the dimensions essential for retention.

These critical service dimensions should be translated into measurable internal KPIs to support continuous improvement, risk monitoring, and strategic decision-making. These KPIs not only support internal performance tracking but also serve as early warning signals for potential client dissatisfaction. Integrating them into comprehensive quality assurance systems and linking them to contract renewal risk assessments enables a more data-driven, client-centric approach to service management.

By aligning operational efforts with what clients truly value and adapting service quality management throughout the client lifecycle, audit firms can enhance client satisfaction and retention. This means that achieving client loyalty requires a strategic focus on building trust, responsiveness, and empathy, while ensuring that clients perceive a high level of reliability in the audit process. By addressing these necessary conditions, audit firms can enhance client satisfaction and retention, ultimately gaining a competitive advantage in the market. This approach reflects not only a holistic approach for understanding quality but also a precise, evidence-based management of each of the quality dimensions drivers.

This study is not without limitations. First, the research relies exclusively on the SERVPERF scale, which, although validated in various contexts, may not capture all relevant aspects of perceived audit quality. We are aware that there are other ways to measure client satisfaction (Butcher et al., 2013). Besides, despite the simplicity of our model, by using a parsimonious model, we pursue isolating the fundamental effects of key quality dimensions on client retention.

Second, the study was conducted in the Spanish market with a cross-sectional sample, which may limit the generalization of the results to other regulatory and cultural contexts. Although EU regulations are similar, comparative studies with other countries would be useful to validate the applicability of these findings at an international level, as well as a possible difference in the results between industries.

Finally, this study adopts the client’s perspective, excluding the viewpoint of audit firms and other stakeholders. Future studies could include other stakeholders’ perspectives to identify discrepancies and potential perception gaps influencing client loyalty (Aghazadeh and Hoang, 2020).

However, these limitations open several promising research areas. First, it would be relevant to examine whether other service quality measurement tools yield consistent results with those obtained in this study. Additionally, it would be interesting to apply longitudinal approaches to observe the evolution of quality perception and client loyalty over time.

Another promising avenue involves evaluating the applicability of these findings in markets with significantly different regulations from those in the European Union. Finally, integrating the perspective of auditing firms to contrast client and service provider perceptions could provide a more comprehensive and detailed approach to audit quality.

Our results suggest that while firm size does not significantly influence perceived satisfaction with audit services, it has a direct negative effect on loyalty. This may indicate that larger companies manage auditor relationships with a greater degree of formality and reduced interpersonal engagement, which can weaken attitudinal loyalty. Although prior studies have shown that larger firms are less likely to switch auditors, demonstrating higher behavioural loyalty (Sinason et al., 2001), this does not necessarily translate into stronger relational or emotional commitment. These findings point to the relevance of distinguishing between behavioural and attitudinal loyalty in professional services. As a next step, future research could extend the model by incorporating potential moderating variables or by adopting a two-sided perspective that includes the views of audit firms.

Thus, future studies could integrate auditor perspectives (Aghazadeh and Hoang, 2020) to uncover dyadic expectation gaps, or employ longitudinal designs (Tuesta-Tapia et al., 2024) to trace loyalty trajectories post-renewal. Such efforts would deepen our understanding of how audit quality evolves in dynamic regulatory and competitive landscapes.

The ethics review board was not required as we were not using personal data nor intervention in human beings (https://www.investigacion.us.es/sites/investigacion/files/202405/0 b_Recomendaciones_tramitacion_2.pdf)

1. The employed knowledge, technical equipment, specialized bibliography, etc. employed by the auditor seemed appropriate.

2. The procedure used to carry out the work seemed appropriate.

3. The staff showed professional reliability.

4. You thought it was appropriate to be offered services in addition to the audit services.

5. The auditing firm has shown facilities when developing its work.

6. The service has been performed as agreed.

7. The service has been provided within the agreed time.

8. There is reliability and consistency in the service provided.

9. You found the quality/price ratio of the service provided to be good.

10. There is good will when carrying out the service.

11. You have been kept informed about the service provided.

12. Your queries are answered.

13. When the client (you) has a problem, the audit firm shows a sincere interest in solving it.

14. The auditing firm inspires confidence.

15. The employees, in charge of performing the service, inspire confidence in you.

16. You found the treatment polite.

17. You think that the staff dispatched has sufficient knowledge and competence.

18. You are satisfied with the work carried out by the firm.

19. The work provided by the employee of the auditing firm is personal and continuous.

20. The employee is committed to his work.

21. The employee is aware of the importance of service to the client (you).

22. You feel that the employee acknowledges your needs as a client.