Sustainability for Predicting Customer Lifetime Value: A Mediation–Moderation Effect Across SEO Metrics in Europe

Abstract

The aim of this study was to analyse the relationship between sustainability and customer lifetime value (CLV) through the mediation–moderation effect of search engine optimization (SEO) metrics of websites. We obtained a data sample of 296 European sustainable firms from both industrial and technological industries. Based on the theory of source credibility, the firm’s official website, where SEO techniques are applied, is more credible regarding its sustainability activities than other sources such as social media, paid advertising, etc. As a result, we show that sustainability is a precursor of financial performance over time in sustainable firms, represented by CLV. Furthermore, we found that the value of the moderating variable, website traffic, alters the indirect effects produced by the mediating variable called website relevance (domain authority), thereby demonstrating a moderated mediation effect. The contribution of this research to the body of literature is twofold. First, it deepens the understanding of how sustainability predicts marketing outcomes based on both digital and customer metrics over time. Second, we rely on recent literature on prediction-oriented modelling (PLS-SEM) to support that it is not suitable for estimation by reflective measurement models due to the woozle effect.

1. Introduction

Despite the growing importance of Search Engine Optimization (SEO) in the digital marketing landscape, significant gaps persist in the sustainability literature concerning its specific applications and impacts. However, few studies have thus far addressed SEO-related aspects aimed at enhancing sustainability content through user-friendly terminology and strategic keyword selection, with the objective of fostering stakeholder engagement via highly targeted content [1] and consequently improving its ranking on the Search Engine Results Page (SERP) [2]. This relevance also extends to Corporate Social Responsibility (CSR) initiatives, where transparency and tailored content foster a favourable public perception [3], as well as enhance consumers’ perceived authenticity of organisational CSR statements [4,5,6]. Previous research suggests that a positive perception of CSR among stakeholders is not merely advantageous; it not only leads to increased sales but also secures long-term stakeholder affiliation with the firm [7]. Therefore, a holistic SEO strategy that integrates content quality, technical optimisation, and a nuanced understanding of user behaviour is essential for driving superior business performance and sustainable growth in the digital era, while concurrently identifying critical avenues for future research to address existing knowledge gaps.

Understanding how Search Engine Optimization (SEO) tools function is essential for responding effectively to changing customer motivations [8]. Moreover, the synergy between Search Engine Optimization (SEO) and user-centric engagement is paramount for driving successful digital outcomes and shaping business models. To this end, firms must ensure excellence in content provision to achieve high customer conversion rates and substantial visibility for their websites across diverse search engines [9]. In this context, the effectiveness of such content, facilitated by Internet technologies, fosters engagement which, in turn, enhances SEO and thereby improves its ranking on the Search Engine Results Page (SERPs) [2,10]. Consequently, resources must be allocated to the creation of distinctive content to augment user engagement. Furthermore, search engines consistently strive to optimise user experience by presenting results in various formats, including text-only, image, and video. This strategy arises from ongoing efforts to adapt to consumers’ evaluation of search results, prompting a shift from product-focused search engine marketing towards an intent-based approach. Such an approach aligns with the establishment of expertise, authority, and trust, ultimately fostering more meaningful interactions through brand-specific search results in SERPs and influencing purchase decisions [11]. The existing body of literature reveals a lack of clarity regarding the specific nature of the relationship between corporate social responsibility (CSR) and marketing performance, thereby complicating comparative analyses due to divergent emphases on CSR components [7]. Although the significance of perceived CSR authenticity in enhancing brand credibility and influencing consumer purchase behavior is widely acknowledged, scholarly understanding of the evaluative processes consumers employ remains insufficient [4]. While certain studies have demonstrated that altruistic values moderate the perceived trustworthiness and competence of firms in shaping CSR perceptions, the mechanisms underlying this moderating effect are still largely unexamined [5]. Embedding sustainability-oriented messaging within marketing communications confers a substantial competitive edge to firms, effectively catering to the expectations of ethically aware consumers [3,9]. It is firmly established that favorable perceptions of CSR contribute to increased sales and the development of long-term stakeholder relationships [3]. Moreover, strategic digital marketing initiatives can enhance profitability and optimize resource allocation by judiciously distributing investments between social media platforms and search engine optimization (SEO), thereby influencing advertising expenditures and sustainability outcomes [10]. To the best of our knowledge, SEO strategies applied to sustainability-related content have been shown to significantly improve stakeholder interaction by facilitating precise and targeted communication [3,4,5]. Consumers tend to inherently regard top-ranked search engine results as more credible than those appearing lower in the rankings [8,11]. This established association between search visibility and consumer behaviour is strongly supported by Source Credibility Theory [12]. The perception of credibility is derived from the perceived expertise and reliability of the information source [12]. Within the digital marketing ecosystem, the demonstration of expertise, authoritativeness, and trustworthiness (E-A-T) in content is critical, as it directly affects search engine rankings and fosters consumer confidence [13]. Consequently, when consumers perceive a brand’s digital presence—particularly through organic search results—as knowledgeable and trustworthy, they are more likely to engage and convert into buyers [10]. Although the extant literature acknowledges the relevance of SEO techniques for improving search result positioning, it notably lacks detailed recommendations regarding which specific SEO practices should be adopted and in what order to maximize effectiveness [14]. Addressing these gaps is essential, as SEO-driven optimization of sustainability content can enhance stakeholder engagement [1,15], and increased user interaction positively influences SEO performance and page rankings [2]. Nevertheless, online content effectiveness is heavily contingent upon technical factors that facilitate its discoverability via search engines, which are governed by SEO practices. Content clarity is relevant from both the user’s perspective—who needs to comprehend the subject matter—and the search engine’s perspective, which relies on specific signals to match content with search intent [1,15]. It leads firms to transition from product-centric search engine marketing strategies toward approaches that prioritize user intent [11]. To date, no studies have comprehensively addressed these technical dimensions, nor have they proposed robust metrics for evaluating the accuracy of information disseminated outside formal reporting channels [1,15].

This study strives to address the gaps in the literature by analysing companies’ organic traffic and its role as a moderator of search engine optimization (SEO) implementation in the relationship between sustainability activities and long-term customer-derived financial results from a predictive perspective. The novel contribution of this study is twofold. First, the main contribution is to go beyond the relationship between sustainability activities and financial results, demonstrating that SEO implementation also predicts long-term financial results. We demonstrate how third-party metrics (domain authority and domain rating) predict the relevance of the website in search engines, mainly due to the integration of links from other websites (backlinks), as well as their moderation through the metric of organic traffic generated through keywords in the optimization of the website. Second, the findings of our study offer a pathway through Partial Least Squares Structural Equation Modelling (PLS-SEM), which is widely recognised as a technique for causal–predictive research. However, as a method for analysing data within marketing, information system research, and related domains, our literature review on PLS-SEM enabled us to identify factors contributing to the so-called “woozle effect” [16]. Consequently, we take into account recent literature on the woozle effect, which suggests that PLS-SEM cannot be reliably used to estimate reflective measurement models [17,18]. This study is structured as follows: first, we explore the theoretical background of Search Engine Optimization (SEO) and source credibility theory, followed by a discussion of sustainability and marketing outcomes, leading to hypothesis formulation. Second, we outline the materials and methods, present our comprehensive results along with a predictive validity report, and conclude with a discussion of the findings and their implications.

2. Theorical Background

2.1. Search Engine Optimization (SEO) and Source Credibility Theory

Improving website visibility (or online presence) may lead to higher customer turnover and increased customer-to-visitor ratios, which, in turn, enhance organisational performance [19]. Organisations can improve the visibility of their websites and attract organic traffic from prospective clients by implementing SEO techniques such as keyword optimisation [20]. Ultimately, SEO techniques and metrics demonstrate a strong correlation with web traffic, which subsequently boosts conversions and bookings in the airline industry [14]. The success of a website largely depends on improving the positioning of its pages to ensure they appear among the top results on SERPs when users submit queries, thereby contributing to greater visibility and increased web traffic [21].

Prior studies on search engines also consider several factors when evaluating a website before ranking it in the SERPs, including organic keywords, backlinks, website speed, and domain authority (DA), among others [14,22,23]. Organic keywords refer to search terms used by search engines to position websites within their unpaid Search Engine Results Pages. Backlinks are external references from other web pages that serve as endorsements for the designated destination page. Web page load time measures the duration, in seconds, required for the complete transfer and rendering of digital content from a server to a client’s web browser. Domain authority is a key metric for comparing websites with one another. Employing SEO strategies such as keyword optimisation enables businesses to enhance their website’s visibility and attract organic traffic from potential customers [20], thereby improving page positioning to appear among the top results in Search Engine Results Pages (SERPs) in response to user queries, and consequently increasing website traffic and overall web visibility [21]. Domain authority scores have a significant positive impact on web traffic, and optimising technical aspects of a website—such as minifying resources (HTML, CSS, and JavaScript)—directly contributes to faster page load times and an improved user experience [14].

Similarly, the literature lacks comprehensive studies on the comparative effects of organic versus paid search results on customer purchase decisions, as well as on the influence of search result presentation formats on customer attitudes and behavioural intentions [11]. The specific cost components associated with SEO promotion (e.g., copywriting), in contrast to PPC advertising (e.g., Google AdWords), directly affect content effectiveness and the overall profitability of the enterprise [9]. Thus, both Search Engine Optimization (SEO), as organic placement, and Search Engine Marketing (SEM), as paid placement, are essential for enhancing firms’ effectiveness and visibility by strategically positioning them at the top of search results [19]. However, organic traffic, as an outcome of SEO techniques, provides long-lasting benefits for firms through a relatively small investment, in contrast to the nature of paid advertising campaigns [14]. For instance, designing user-friendly websites with fewer pages promotes user engagement and increases visitor numbers by reducing “Pages per Visit” and bounce rates [9]. Customers perceive search results in leading positions as more trustworthy [11], owing to SEO practices that effectively address users’ search intent in a relevant and satisfactory manner, thereby making the quality of the content both evident and comprehensible [15]. According to [24], the source of the communication plays the most critical role in determining the effectiveness and trustworthiness of the information. Individuals assess the reliability of the information source, such as the website. If individuals perceive the information source as credible, they are more likely to seek out that information. A credible source can be defined as a communication channel that provides accurate and trustworthy information [25]. According to source credibility theory, two key indicators are essential for assessing the credibility of communicators: expertise and trustworthiness. Expertise refers to the extent to which consumers believe the information source is sufficiently capable of providing accurate information, while trust denotes the behavioural intention to rely on another person [26]. According to [12], consumers initially evaluate the credibility of an information source, such as online reviews, based on perceived expertise and trustworthiness. Collectively, these studies underscore that consumers’ perceptions of credibility—whether of information sources or the company itself—are either a prerequisite for, or a direct outcome of, their causal inferences regarding brands and their actions. In the context of sustainability, consumers are inherently sceptical of CSR information; therefore, perceived source credibility and favourable attributions regarding a company’s genuine commitment to sustainable practices become paramount for building trust and a positive CSR image. Given the increasing reliance of digital marketing on data and algorithms, it is essential for marketers to focus on cultivating expertise, authority, and trust within search engine environments [13,27].

2.2. Sustainability and Marketing Outcomes Based on Customers

Twenty-first-century sustainable marketing aims to internalise social and environmental costs, create long-term value, and support sustainable development goals. In contrast to traditional marketing, which often promotes overconsumption, sustainability marketing seeks to conserve resources and encourage regenerative practices. It goes beyond the environmental or green marketing of the 1980s–1990s, which focused on developing products with improved socio-environmental performance to cater to the ‘green consumer’ segment [28]. Recent literature emphasises the need to understand consumers’ macro-level impact on the environment (ecological footprint). At the micro level, individual impacts may benefit consumers (e.g., energy bill savings), businesses (e.g., word of mouth [WOM]), and the environment (e.g., reduced energy consumption). At the macro level, Ref. [29] analysed ways to reduce consumer spending and energy consumption to mitigate environmental impacts, suggesting that marketing researchers should conduct further studies on various services and raw materials, renewable energy, longitudinal data, marketing education, and diverse cultural contexts. In turn, at the individual or micro level, Ref. [30] examined a social marketing field study conducted by energy companies on actual customers, finding that a gamified app influenced consumers’ energy-saving behaviours and word of mouth, resulting in significant monetary savings.

Recent studies also provide valuable insights into how sustainability is affecting outcomes in the field of marketing. Consumers generally place greater confidence in interpersonal communication compared to content created by marketers [31], so social media marketing tools also analyse perceived sustainability from customer’s perspective. For instance, Ref. [32] proposed a mediation model for brand attitude that serves to contrast perceived sustainability dimensions (ESG), incorporating trust as a moderating variable. Consequently, the attitude toward the brand-related post, along with overall engagement on social media platforms, primarily influences consumers’ intentions to share content. Therefore, enhancing the perception of sustainability tends to encourage consumers to buy from non-luxury brands, whereas it does not significantly affect purchasing behaviour toward luxury brands. Consequently, framing messages around consumers’ environmental motivation and environmental knowledge positively influences consumer behaviour and supports sustainable manufacturing practices that value participation (intention to immediately return e-waste) in the circular economy and sustainable supply chain management.

Ref. [33] demonstrated that sustainable activities within the traditional fashion industry have a direct and positive effect on brand image. In turn, brand image directly influences consumer perceptions, as reflected in their association with the brand. Perceived sustainable marketing activities—economic, social, environmental, and cultural—encompass decision-making processes and business practices involving the local community and consumers, such as production and sales, the social environment, and environmentally friendly ethics. As a result, the influence of economic, social, and cultural environmental activities on brand image is substantiated. However, social activities—or the social dimension—do not impact brand image, suggesting that consumers are unaffected when the direct benefit is conferred upon a third party.

In contrast to the perceived customer perspective, in firms’ perspective market research studies, sustainability is considered corporate social responsibility (CSR); in turn, sales metrics are one of these key outcomes. In fact, eco-friendly or fair-trade managers employ various strategies to enhance the benefits of sustainability; thus, investing in strong CSR is an effective approach to boosting sales of new sustainable products [34]. If a brand’s CSR reputation is weak, the success of new sustainable products is likely to be limited, and the product is unlikely to succeed in the market. Conversely, when a brand has a strong CSR reputation, it positively affects new product sales, and further investment in CSR is likely to enhance the performance of all new products. From the perspective of branding and commercial outcomes, inconsistency between the values a brand communicates, and its actual behaviours is perceived as brand disinformation—a phenomenon consumers associate with deceptive advertising and a lack of integrity. In this regard, Ref. [35] examined the advertising campaign of a socially responsible beauty brand, analysing the effect of the company’s new sustainable packaging on sales through the theoretical lens of expectancy disconfirmation theory. Nevertheless, there remains a lack of research on how and why brand disinformation triggers consequential consumer reactions and affects sales. While marketing can enhance external communication of CSR activities, it does not necessarily improve sustainable performance. [36] found a significant and negative relationship between ESG performance and greenwashing based on greenwashing measures derived from a textual analysis. According to [37], a one-point increase in greenwashing results in a 0.56-point decrease in the company’s green brand perception (with a loading factor of 0.78). To reduce greenwashing and enhance stakeholder trust, companies must consider the website content must be trustworthy and accurately reflect the company’s green initiatives. For that, comprehensive non-financial statements should be made publicly available. Moreover, press releases, relevant activities, and environmental audit reports must be accessible on the website as well as mandatory disclosures about its environmental performance.

2.3. Woozle Effect in the Partial Least Squares Structural Equation Modelling (PLS-SEM)

The “Woozle effect,” also known as “evidence by citation,” describes how unfounded claims gain credibility through repeated citation, rather than actual evidence or critical analysis. In the partial least squares structural equation modelling (PLS-SEM) literature, this is particularly problematic regarding the persistent the erroneous belief persists that PLS-SEM is appropriate for evaluating reflective measurement models. Despite extensive scientific evidence, including mathematical proofs and simulation studies, demonstrating biased and misleading results from PLS-SEM for such models, this “zombie Woozle” stubbornly persists. Its propagation is linked to several systemic issues: prioritizing widespread dissemination over scientific accuracy, academic dependencies, misleading graphical representations in software, flawed reasoning (such as inferring suitability from existing use), and a widespread lack of proper referencing. Belief perseverance further obstructs scientific self-correction, as researchers continue to promulgate false claims despite contradictory findings. Therefore, for rigorous marketing research, the choice of SEM approach demonstrably matters for the reliability and validity of your results. The most important initial step involves meticulously aligning the chosen SEM approach with your underlying conceptual or theoretical model, which necessitates a clear understanding of constructs’ nature: whether they are reflectively measured latent variables (common factors) or emergent variables (composites). Given that the theoretical constructs in our study are defined as emergent variables (composites), the use of CB-SEM alongside the Henseler–Ogasawara (H-O) specification, or alternatively PLS-SEM employing Mode B, represents appropriate methodological choices to yield consistent parameter estimates for the proposed model. Ref. [17] indicates that failure to select an estimator that fits the specified statistical model and its assumptions will most likely yield inconsistent and/or biased estimates. Such biases can significantly jeopardize the conclusions drawn from your estimated model, leading to erroneous or questionable findings.

2.4. Hypothesis

The dimensions of sustainability or ESG exhibit a strong association with the financial outcomes of firms. A company’s CSR image represents a form of identity that communicates its commitment to corporate social responsibility initiatives. This image mirrors how stakeholders perceive the organization’s socially responsible actions, whether these are intentional or occur as a byproduct. The majority of consumers neither actively seek CSR-related information nor possess substantial awareness of companies’ CSR efforts. Firms that maintain a robust reputation or image linked to their CSR initiatives tend to elicit favourable consumer reactions, including increased loyalty, intention to purchase, and trust [5]. Nonetheless, CSR practices targeting the environment, workforce, and communities exert a notable influence on financial performance across various dimensions, except in the case of customers, where sustainable development plays a mediating role through green innovation [38].

Customer lifetime value (CLV), as a marketing indicator, represents the total economic exchanges within the customer–firm relationship over a specified timeframe—or potentially without limit—where firms are appraised by discounting future margins or sales across time. Findings from [39] demonstrate that ESG dimensions exert a positive influence on customer lifetime value.

H1: 

Sustainability (ESG) influences Customer Lifetime Value (CLV).

The implementation of Search Engine Optimization (SEO) serves as a mechanism for enhancing website credibility and transparency toward stakeholders. Nonetheless, the application of SEO remains heterogeneous due to the diversity of metrics, techniques, tools, and practices employed. Key determinants of online visibility include elements such as the Alternative Text Attribute, backlinks, robots.txt files, domain authority, and bounce rate. Ref. [14] showed that organic traffic is significantly influenced by SEO indicators—namely domain authority, keyword optimization, and backlink profiles. Moreover, backlinks are incorporated into domain authority due to their interconnected roles in SEO. While SEO implementation is instrumental in improving search engine rankings and driving organic traffic, domain authority emerges as a more robust comparative metric among websites. Websites with elevated domain authority scores tend to achieve higher positions in Search Engine Results Pages (SERPs) relative to those with lower scores. Domain authority thus encapsulates the effectiveness of SEO implementation, leading practitioners—more so than academics—to prioritize its use. Through this predictive metric, organizations seek to enhance their visibility, popularity, and relevance, as reflected in the superior scores attained by high-performing websites [40]. SEO also has a significant effect on the sales performance and business sustainability [41].

H2: 

Sustainability (ESG) influences website relevance (SEO_R).

H3: 

Website relevance (SEO_R) influences CLV.

H4: 

The direct effect between sustainability and CLV is mediated by website relevance (SEO_R).

Organic traffic provides long-lasting benefits with relatively small investment for sustainability, contrasting with the nature of paid advertising campaigns [18]. Prior research has established that website traffic influences domain authority [40], with traffic serving as a proxy for user engagement and site attractiveness. To initiate aligning of the theoretical model, we observe that the moderating effect applies to the entire indirect effect. This is, the path between ESG and SEO relevance is complemented by the path between SEO relevance and CLV, where the entire indirect effect interacts with SEO traffic. According to [42], there are different types of moderation–mediation effects. Our model specification suggests a single moderation affecting the entire indirect effect, in contrast to other models that propose two separate moderating effects—each corresponding to a direct effect that is sequentially connected through a path.

H5: 

The indirect effect between sustainability and CLV is moderated by website traffic (SEO_T).

Based on the aforementioned arguments, Figure 1 shows model of effects and hypothesis.

3. Materials and Methods

We used a sample of 296 Europe firms. We categorized the samples into countries, industries and activities according to information that is provided from Refinitiv Thomson Reuters Eikon database [43]. The four main countries are United Kingdom with 57 firms, Germany with 54 firms, Sweden with 44 firms and Switzerland with 20 firms. The rest of the countries do not include more than 20 firms. In turn, the four main industries are IT Consulting & Other Services with 51 firms, Application Software with 43 firms, Integrated Telecommunication Services with 24 firms and Electronic Equipment & Instruments with 20 firms. The rest of the industries do not include more than 20 firms. Finally, the four main activities are IT Services & Consulting with 56 firms, Integrated Telecommunications Services with 27 firms, Software with 24 firms and Enterprise Software with 24 firms. The rest of the activities do not include more than 20 firms. This study relies on three different sources to collect data.

First, the Refinitiv Thomson Reuters Eikon database [43] provides the sustainability dimension (three indicators from ESG performance pillar scores across environmental, social, and governance) and financial dimension (two indicators based on different calculated Customer lifetime value). Environmental, social, and governance indicators are each scored on a scale of 0 to 100. The environmental indicator refers to three aspects: resource use, emissions, and innovation. The social indicator refers to four aspects: workforce, human rights, community, and product responsibility. The governance indicator refers to three aspects: management, shareholders, and corporate social responsibility strategies. Additionally, the database offers financial data such as sales, profit margin, and the weighted average cost of capital (WACC), which we utilize to calculate customer lifetime value (CLV) as follows: (a) discounting sales from infinity to the year 2024 using the weighted average cost of capital (WACC) of each firm to this year; (b) substituting the profit margin instead of sales and recalculation of another variable.

Second, we collect data from Moz and Ahrefs, which are SaaS firms that provide an all-in-one SEO toolset. Both firms update their databases with the most recent backlinks and their crawlers process up to billions pages a day. They also report two different metrics, which are Domain Authority by Moz and Domain Rating by Ahrefs. We summarize them as follows: on one hand, Domain Authority (DA) constitutes a ranking metric, formulated by Moz, which prognosticates a website’s propensity to achieve high placement within search engine result pages (SERPs). Domain Authority valuations span from 1 to 100, where elevated scores correlate with an increased probability of achieving higher search rankings. This score can then be used when comparing websites or tracking the “ranking strength” of a website over time. Domain Authority is calculated by evaluating multiple factors, including linking root domains and total number of links that uses a machine learning model to predictively find a “best fit” algorithm that most closely correlates these link data with rankings across thousands of actual search results. Therefore, Moz uses them as standards to scale against. For the order hand, Ahrefs has its own proprietary authority metric, called Domain Rating (DR), used to estimate the site’s “link popularity” compared to other websites. As a rule of thumb, a higher Domain Rating (range from 1 to 100) implies a more authoritative website. Website authority is an SEO concept that refers to the overall “strength” of a particular domain. Strength, in Ahrefs, is the ability (or likelihood) of a given domain to rank high in the SERPs and pass its backlink strength (“link juice”) to other websites. The concept of website authority in SEO across Domain Rating should not be confused with the Domain Authority (DA) as a metric. With links of the websites’ firms provided by Refinitiv Thomson Reuters Eikon database, we collect these two metrics that encompass website relevance SEO results of firms in a long-term manner [18,41,44].

Third, SimilarWeb is a web analytics and market intelligence platform with features such as website traffic analysis from firms. Similarweb estimates the total amount of traffic different websites receive as well as key metrics such as total visits and engagement data. Similarweb also segments the traffic information, so users obtain data by introducing a link of each website. With links of the websites’ firms provided by Refinitiv Thomson Reuters Eikon database, we collect two percentage metrics as follows: organic traffic of firms as a percentage during three last months. Furthermore, this metric is again divided and reported in two parts by the Similarweb platform: branded and non-branded. Therefore, we collect the second one to evaluate subpercentage of users that not employed brand keywords in their organic search. It also collects information about users that no awareness brand of the firm obtains results in the SERP in an organic search during 3 last months of 2024. These two metrics encompass website traffic SEO results of firms in a short-term manner [18].

Finally, data were collected in January 2025. For Moz and Ahref, data are from the end of 2024; for Similarweb, data are from the last quarter of 2024; for Refinitiv Thomson Reuters Eikon database, data are from the latest update of the ESG pillar score performances, as well as sales, profit margin and WACC. Partial least squares (PLS) represents a statistical methodology developed to support causal–predictive analysis within the context of structural equation modeling (SEM) [16]. Unlike traditional SEM, which seeks to estimate model parameters that best reproduce the variance–covariance matrix, PLS focuses on maximizing the predictive accuracy of the model’s causal relationships. This fundamental difference in objectives leads to distinct optimization algorithms: while SEM relies on strict assumptions about data distribution, PLS operates without such constraints, making it particularly suitable for exploratory research. Thus, this approach suggests greater openness when defining the timeframe, even though it is academic research, its nature is exploratory data analysis. In this study, PLS is employed to enhance the predictive power of the proposed model, as well as its predictive validity, enabling a more robust understanding of the causal links between the constructs under investigation.

Table 1. Descriptive statistics of sample.

Indicator	Mean	St	Min	Max	N
Environmental	41.8	21.6	1	95	296
Governance	50.2	23.1	3	95	296
Social	52.2	21.6	2	95	296
ln_CLV profit-margin infinite	13.2	2.2	3	19	296
ln_CLV sales infinite	15.7	1.8	11	21	296
Ahref domain rating (DR)	61.2	16.5	7	95	296
Moz domain authority (DA)	49.3	16.1	18	95	296
porce_organic	41.7	15.9	2	79	296
porce_organic_nobranded	24.2	14.1	0	68	296

shows descriptive statistics of the sample.

4. Results

PLS considers two models (measurement and structural) to perform a statistical analysis focusing on variance. To evaluate the measurement model, investigators can incorporate both reflectively and formatively structured measurement models, a process PLS-SEM conducts unrestricted. However, a recent study [16] indicated that Partial Least Squares Structural Equation Modelling (PLS-SEM) proves inadequate for the evaluation of reflective measurement models. Thus, to assess the measurement model with formative composites, we tested individual reliability through absolute (loadings) and relative (weights) contribution of each indicator to the composite. Both loadings and weights are statistically significant at 95% when confidence intervals (lower and upper) do not include the value zero [42,45] We also tested collinearity, which requires values below the threshold, where VIF < 5 [45]. In turn, we also analysed discriminant validity among all composites (ESG, WEB, SEO, CLV), which requires values below the threshold of heterotrait–monotrait (HTMT) < 0.9 [46]. The obtained values among composites are in brackets: SEO_T-CLV (0,12), SEO_R-CLV (0,61), SEO_R-SEO_T (0,10), SEO_TxSEO_R-CLV (0,13), SEO_TxSEO_R-SEO_T (0,04) and SEO_TxSEO_R-SEO_R (0,09).

Table 2. Measurement model of PLS.

Indicator → Composite	Loads	5.0%	95.0%	Weights	5.0%	95.0%	VIF
Environmental → ESG	0.84	0.77	0.90	0.37	0.23	0.50	1.80
Governance → ESG	0.64	0.54	0.73	0.27	0.16	0.37	1.25
Social → ESG	0.91	0.86	0.95	0.57	0.44	0.70	1.78
ahref_domain rating (DR) → SEO_R	0.90	0.80	0.97	0.26	0.03	0.59	3.21
ln_CLV_profit margin_infinite → CLV	0.93	0.89	0.96	0.35	0.19	0.52	3.45
ln_CLV sales infinite → CLV	0.98	0.96	0.99	0.69	0.52	0.83	3.45
moz_domain authority (DA) → SEO_R	0.99	0.95	0.99	0.78	0.45	0.99	3.21
porce_organic → SEO_T	0.99	0.44	0.99	0.89	0.07	0.99	1.35
porce_organic_nobranded → SEO_T	0.64	0.08	0.99	0.18	0.67	0.99	1.35

shows the measurement model of PLS.

To check the structural model, we performed a nonparametric bootstrap resampling procedure (10,000 samples). We tested the hypotheses for the study (direct effect), using the PLS technique of the structural model for the sample. In PLS tests, path coefficients are statistically significant at 95% when confidence intervals (lower and upper) do not include the value zero [42].

Table 3. Structural model of PLS: path coefficients from hypotheses.

Path	Beta	2.5%	97.5%	Fulfilment
ESG → CLV (direct effect)	0.54	0.46	0.62	H1 Accepted
ESG → WEB (direct effect)	0.48	0.38	0.58	H2 Accepted
SEO_R → CLV (direct effect)	0.29	0.19	0.37	H3 Accepted
ESG → SEO_R → CLV (indirect effect)	0.14	0.09	0.19	H4 Accepted
(ESG → SEO_R) x (SEO x SEO_R → CLV) (mediation moderation effect)	-	−0086	−0001	H5 Accepted

shows fulfilment hypotheses from structural model of PLS.

We analysed both the indirect effects and whether these effects are conditioned by another variable that acts as a moderating variable. A moderate mediation effect is when the value of the moderating variable (SEO_T) modifies the indirect effects produced by a mediating variable (SEO_R) in the PLS model, so we tested both mediation effects and moderate mediation effects. The coefficient value of moderate mediation effect is not required, as only confidence intervals (lower and upper) do not include the value zero [42,45].

Prior studies have examined the relevance and advantages of marketing results derived from relationships on sustainability. The most frequently employed terms are associated with the key components or rationales that firms utilize to convey the value of their offerings, in light of their continuous ESG initiatives. Corporate websites constitute a primary instrument for ensuring transparency and facilitating communication between firms and their customers [47]. The visibility of a website serves as a critical indicator for evaluating the success of a firm’s marketing efforts and functions as a key determinant in distinguishing organizational performance [48].

The results of this research provide further support for the integration of technical and content aspects of the web from a predictive perspective between sustainability and marketing outcomes. In this sense, SEO metrics also play an important role in measuring the effectiveness of digital marketing strategies [48,49,50]. Two new composites or constructs from SEO metrics, relevance (of the domain) and traffic volume (organic), due to formative indicators are employed. Among sustainability and customer outcomes, the mediator effect of digitalisation shows the overall quality and credibility of a domain through its indexing and positioning by search engines such as Google, Bing and Yahoo. This composite called SEO_R with two metrics from different sources also showed that firms require a long time to build trust relationships with other firms to optimise SEO off-page. This mediation effect is also be moderated through organic traffic volume (quantity of users and/or visits) obtained through on-page SEO implementation (e.g., keywords, technical aspects, etc.). As widely acknowledged, a relationship between off and on page SEO exists.

At the model level, we computed the q²_predict metric and evaluated the prediction errors of the PLS model in comparison with baseline predictions derived from the average values of the training dataset [51]. A positive q²_predict value indicates that the prediction errors from the initial model are lower than those from the benchmark model, suggesting superior predictive capability of the first model regarding performance outcomes. To assess the predictive validity of our path model, we documented the q²_predict ratio for each composite, categorizing the results into three distinct levels. An effect size is considered small when 0.02 ≤ q² < 0.15, moderate when 0.15 ≤ q² < 0.35, and large when q² ≥ 0.35. Additionally, we presented the R2 values for each composite, which range from 0 to 1, and were also classified into three levels of explanatory power. R2 is interpreted as poor when it falls within the interval of 0.19 to 0.32, moderate between 0.33 and 0.66, and substantial when exceeding 0.67. We further evaluated model fit using the SRMR index, ensuring that its value did not surpass the threshold of 0.8 [45].

Table 4. Structural model of PLS: predictive validity and goodness of fit.

Composite	q2_Predict	R-Square	SRMR
CLV	0.46	0.55	0.04
SEO_R	0.22	0.23

presents the results concerning predictive validity and model fit derived from the structural PLS model.

5. Discussion

In this research, we studied how sustainability predicts CLV across website metrics. We relied on data from 296 European firms from the industrial and technological sectors. The proposal prediction model set up three indicators from the Refinitiv Thomson Reuter Eikon database [43], which the sustainability literature also generally recognises: environmental, social and governance. The outcome variable is CLV, and its mediation–moderation effect is analysed based on a website’s traffic and relevance. Our research provides insight about the effects of sustainability on customers, considering SEO as one digital marketing technique (others include social networks, content marketing, viral marketing, etc.). The predictive quality is satisfied (see Table 4). Analysing the indicators, q²_predict shows a large effect (0.46), and the R2 indicator shows a moderate effect (0.55). The SRMR indicator confirms the acceptance of the structural model fit because its value (0.04) is below the criterion recommended by the SmartPLS literature (0.08).

This study offers evidence for the hypotheses for the proposed model (see Table 3). Of the antecedents of CLV, sustainability composite is the main precursor of CLV, with a direct effect coefficient equal to 0.54 (Hypothesis 1). Website relevance (SEO_R) composite is the second precursor of CLV, with a direct effect coefficient equal to 0.29 (Hypothesis 3). Moreover, the sustainability composite has a direct effect coefficient equal to 0.48 on the website relevance composite (Hypothesis 2), as well as an indirect effect coefficient between the sustainability composite and a CLV equal to 0.14 (Hypothesis 4). Finally, we tested the mediation–moderation effect between the sustainability composite and CLV across websites using SEO techniques with an interval confidence not equal to zero. Thus, a website’s traffic volume (‘quantity’) moderates relevance (‘quality’). In this sense, sustainability influences CLV as directly as the mediation–moderation effect. The total or accumulated indirect effect of mediation–moderation (digital marketing by SEO techniques) is higher than the direct effect.

This study’s contribution to marketing literature is to shed light on the relationship between sustainability and marketing outcomes by revealing the impact of sustainability on CLV across the moderation–mediation effect of SEO metrics. In the sustainability context, source credibility about a company’s commitment to sustainable practices are essential for building trust and a positive CSR image. Source credibility theory focuses on establishing expertise, authoritativeness, and trustworthiness, so both digital and analytic marketing on search engines becomes more data- and algorithm-driven. Web users searching for information about firms on websites are familiar with keywords that support the result on the SERP due to the search intention concept. With the constant updating of search engine algorithms, the credibility of sustainability content of the firms in search engine improves. In this sense, firms strive to offer updated relevant content on the web that helps both gain organic traffic and indexation from other sources. Hence, this research considers sustainability activities as the main precursor and new formative composites through SEO metrics calculated by third-party data sources, underlying the role of SEO implementation in the user engagement context of both digital and analytics marketing. Metrics such as Moz’s Domain Authority and Ahref’s Domain Rating encompass and monitor the relevance of firm websites, while organic traffic and the subset of non-brand organic traffic at short-term data moderate these consolidated metrics. Moderated mediation reveals the impact on clients’ long-term results.

The use of PLS-SEM with SEO metrics is also a contribution to marketing literature due to the woozle effect. Prediction-oriented models do not seek to test a theory. These models are used to advance on the exploratory data analysis for a predictive purpose. In this sense, the data of this study were not collected data from customer–user surveys because specialized firms provide data widely collected by third parties in the digital ecosystem. Thus, this approach suggests greater openness when defining the timeframe; even though it is academic research, its nature is exploratory data analysis. It leads us to discuss two issues: the variables under analysis and the emerging concept of digital sustainability. First, the variables used in the measurement model of PLS, using a formative composite, guarantee unbiased coefficients through predictive estimation in PLS mode B for the exploratory data model. Despite the differences from data collection through surveys which are used to include a set of reflective measures, researchers must take into account the woozle effect. Otherwise, conclusions of their studies will not contribute to PLS literature. Second, digital sustainability represents a developing conceptual framework that explores the role of data and digital technologies in advancing sustainable development objectives. Drawing from established research in information systems, digital sustainability leverages digital resources (such as data and technical infrastructure) and digital artifacts (including software applications) to pursue various sustainability objectives, while also navigating the trade-offs and synergies among competing goals and their respective beneficial and adverse impacts [52]. In this sense, prior literature of digital marketing showed sophisticated analytical frameworks to predict the efficacy of optimization efforts within business models. Predictive simulation models have been created to present the intercorrelation between examined metrics as well as possible optimization strategies [10]. These models offer substantial insights into how machine learning systems influence consumer perceptions, levels of satisfaction, and behavioural intentions [11], thereby informing more effective digital strategies. Therefore, these advanced modelling and simulation approaches are relevant for strategic decision-making and for forecasting the impact of SEO and digital marketing initiatives on overall business performance and sustainability.

From a managerial perspective, community-focused and/or socially focused activities, actions in favour of the environment, fair trade, organic products, and equitable services, may be part of a firm’s actions regarding sustainable development. Therefore, these actions are not generated in pursuit of overconsumption. Consumer acceptance leads to predictions of future sales as well as the firm’s financial sustainability based on customer metrics. Policymakers can manage a firm’s website to create value for customers. Both indicators—organic traffic volume and relevance of the website—are crucial to mediate and moderate the relationship between sustainability and CLV.

Regarding the limitations, the primary limitation of this study is the sample we used: publicly traded companies from the Refinitiv Thomson Reuters Eikon database, which is confined to Europe and includes only the industrial and technology sectors. Likewise, our SEO implementation approach could be complemented by other sectors or industries to generalise the research results. We advocate for comprehensive studies involving a broader range of countries and industries to anticipate a wider applicability of our study. A second limitation pertains to the digital marketing metric, as we considered only two composites (four indicators) based on SEO for websites. In fact, SEO techniques encompass a wider array of measures. Additional studies could include data collection for digital sustainability, that is, metrics within the digital and analytics scope.

Future research could further elaborate on how other digital marketing channels (e.g., social media, content marketing) or strategies (e.g., viral marketing, affiliate marketing) might also play a role in sustainability–CLV linkages, extending the scope beyond SEO. Firms that provide processed customer and internet user data are becoming increasingly specialised, which can facilitate the development of predictive causal models in business. Despite the opportunity this represents for academic research, it requires a well-applied methodology, as we have discussed in this study on the Woozle effect; otherwise, we may reach pseudoscientific conclusions.