Website Behavioural Intention (WBI)

1. Abstract

The Website Behavioural Intention (WBI) scale is a psychometric instrument designed to assess downstream consumer behavioral intentions within digital environments. Developed by Glen L. Urban, Venkatesh Shankar, Fareena Sultan, and Yakov Bart (2005) in their foundational Journal of Marketing study, the instrument operationalizes the conative component of consumer decision-making following digital interactions. The scale comprises three central indicators capturing three distinct yet interrelated post-visit digital behaviors: intention to transact (purchase financial commitment), intention to recommend (electronic word-of-mouth and advocacy), and intention to register (disclosure of personal identifiable information and digital relationship formation). Administered primarily via a 5-point, 7-point, or 10-point semantic differential or Likert-type scale, the WBI serves as a critical criterion variable in structural equation models evaluating online trust, e-commerce interface design, and human-computer interaction (HCI).

Psychometrically, the instrument demonstrates robust statistical properties across diverse digital platforms, exhibiting an average variance extracted (AVE) of 0.72 and composite reliability (CR) coefficients exceeding 0.85 across multiple consumer cohorts. Despite its extensive adoption and proven predictive validity regarding actual consumer conversion, the operationalization of WBI as a strictly unidimensional construct has generated theoretical debate. Psychometricians and marketing scholars note that while transaction, recommendation, and registration share common variance driven by underlying website trust, they entail disparate consumer risk thresholds: financial risk, social-reputational risk, and privacy-calculus risk, respectively. This article delivers an exhaustive psychometric deconstruction of the WBI scale, analyzing its theoretical underpinnings, structural equation modeling specifications, validity parameters, cross-category generalizability, and applied digital assessment protocols.

2. Keywords

Website Behavioural Intention, online trust, conative loyalty, e-commerce psychometrics, behavioral intention, structural equation modeling, transactional intention, electronic word-of-mouth, information disclosure, consumer risk calculus.

3. Authors

The Website Behavioural Intention operational framework was formulated and empirically validated by an academic team specializing in quantitative marketing, digital trust architecture, and marketing analytics:

• Yakov Bart, Ph.D. — Professor of Marketing, Joseph G. Riesman Research Professor, D'Amore-McKim School of Business, Northeastern University; Faculty Affiliate at the Network Science Institute. Specialized in digital marketing strategy, mobile advertising, and social media analytics.
• Venkatesh Shankar, Ph.D. — Professor of Marketing, Jerry and Kay Cox '72 Chair in Marketing, Mays Business School, Texas A&M University; Director of Research at the Center for Retailing Studies. Specialized in digital business strategy, customer lifetime value, and marketing modeling.
• Fareena Sultan, Ph.D. — Professor of Marketing and Robert Morrison Fellow, D'Amore-McKim School of Business, Northeastern University. Specialized in digital marketing, mobile media, and global high-technology innovations.
• Glen L. Urban, Ph.D. — David Austin Professor of Management Emeritus, MIT Sloan School of Management, Massachusetts Institute of Technology. Pioneer in digital trust-based marketing, pre-market forecasting models (ASSESSOR), and synthetic customer representations.

4. Purpose

The primary purpose of the Website Behavioural Intention (WBI) instrument is to quantitatively capture and evaluate the conative outcomes of digital consumer experiences. In psychometric research and marketing science, understanding what drives a user to visit a website is insufficient; researchers must determine how digital exposures convert into concrete future behaviors. The WBI was formulated to resolve significant operational limitations in early e-commerce literature, which frequently relied on single-item metrics of purchase intent or conflated passive browsing satisfaction with active behavioral commitment.

From an applied research perspective, the WBI operationalizes consumer intent across the digital conversion funnel. It addresses both transactional business models (e.g., retail, travel booking, financial services) and relationship-oriented web environments (e.g., content portals, social platforms, brand advocacy sites). By capturing downstream intentions beyond immediate monetary exchange, the instrument permits scholars and market researchers to evaluate website effectiveness even in non-transactional contexts or across prolonged consideration cycles where immediate purchase is improbable.

From a theoretical standpoint, the instrument provides a standardized criterion construct for evaluating latent structural models of consumer behavior. As demonstrated by Bart et al. (2005), online trust functions as an essential mediating mechanism between website characteristics (such as navigation, graphic design, security policies, and privacy seals) and downstream behavioral commitments. The WBI provides the precise metric needed to assess the strength and efficacy of this mediation. Clinical and applied human factors researchers employ the tool to benchmark user interface (UI) redesigns, audit digital accessibility implementations, and evaluate how perceived privacy breaches or platform restructurings impact end-user behavioral loyalty.

5. Psychological Construct

The latent construct of Website Behavioural Intention occupies the conative stage of the classical cognitive-affective-conative hierarchy of attitudes. Within this paradigm, cognitive beliefs (e.g., website usability, perceived information quality) and affective responses (e.g., digital affect, institutional trust, visual appeal) culminate in conative predispositions toward action. The WBI operationalizes these predispositions into three distinct functional dimensions:

1. Transactional Intention (Commercial Commitment)

Transactional intention reflects the consumer's subjective probability of executing a monetary exchange on the website. This dimension involves the highest level of perceived financial risk. To execute a digital transaction, an individual must overcome perceived vulnerabilities regarding credit card fraud, billing inaccuracies, merchant non-fulfillment, and post-purchase disputes. Consequently, transactional intention requires high levels of structural assurance and technical security perception. When consumers report high intention to transact, they signal that the website has neutralized perceived financial vulnerability and delivered sufficient perceived utility to justify economic risk.

2. Recommendation Intention (Social Advocacy & e-WOM)

Recommendation intention captures the user's willingness to endorse the website to third parties (friends, family, professional networks, or broader online audiences via electronic word-of-mouth). Unlike transactional intentions, which involve personal economic capital, recommendation intention puts the user's social capital and interpersonal reputation on the line. Endorsing an untrustworthy or ineffective website risks social embarrassment and diminished interpersonal trust. Thus, high recommendation intentions reflect strong affective attachment, high perceived platform credibility, and satisfaction that exceeds baseline expectations.

3. Registration Intention (Information Disclosure & Identity Commitment)

Registration intention denotes the consumer's willingness to create a persistent profile, surrender personal identifiers (name, email, demographic details, behavioral preferences), and permit longitudinal tracking. Grounded in the privacy calculus framework, this behavioral intention is governed by the perceived balance between expected customization benefits and anticipated privacy intrusion (e.g., spam, unauthorized data sharing, secondary data use). Registration represents a foundational step in relationship marketing, transforming an anonymous, transient session into an ongoing institutional relationship.

The integration of these three facets into a unified construct rests on the premise that all three constitute downstream manifestations of brand and website loyalty. However, because each behavior entails distinct risk profiles (financial, social, and privacy), psychometricians must evaluate whether the construct functions as a strictly reflective unidimensional variable or a multidimensional higher-order latent structure.

6. Theoretical Framework

The WBI scale is anchored in foundational psychological and socio-cognitive models of human decision-making and technology adoption:

The Theory of Reasoned Action (TRA) and Theory of Planned Behavior (TPB)

Formulated by Martin Fishbein and Icek Ajzen, the Theory of Planned Behavior asserts that behavioral intention is the most proximal and reliable psychological determinant of actual behavior. Intention represents a calculated conscious plan requiring motivational investment and cognitive effort. Within the WBI framework, consumer attitudes toward the website, combined with subjective norms and perceived behavioral control (e.g., internet self-efficacy, seamless navigation), translate into specific behavioral intentions. These, in turn, predict actual transactional logs and repeat visits.

The Technology Acceptance Model (TAM)

Originating from Fred Davis (1989), the Technology Acceptance Model posits that behavioral intention to use an information system is governed by its Perceived Usefulness (PU) and Perceived Ease of Use (PEOU). In the context of the WBI, the website is conceptualized simultaneously as an information technology artifact and a commercial storefront. Navigational ease and information quality feed into cognitive valuations of usefulness, which subsequently determine behavioral intentions to interact with, register on, and transact through the system.

The Trust-Commitment Theory and Risk Paradigm

Morgan and Hunt's (1994) Commitment-Trust Theory establishes that relationship commitment and trust are paramount mediating variables in relational exchange. Mayer, Davis, and Schoorman (1995) define trust as the willingness of a party to be vulnerable to the actions of another based on the expectation that the other will perform a particular action important to the trustor, irrespective of the ability to monitor or control them. In digital environments characterized by spatial and temporal separation, asymmetric information, and lack of physical touch, trust serves as an essential cognitive reducer of social and economic complexity. Bart et al. (2005) integrated this framework by showing that online trust directly drives WBI, mitigating perceived risk and providing the psychological bridge between website evaluation and conative intent.

7. Validity

The psychometric validity of the WBI has been evaluated across extensive empirical trials involving thousands of consumers and hundreds of diverse web domains:

Convergent Validity

Convergent validity evaluates whether the operational indicators reflect the underlying latent construct. In the foundational validation study by Bart et al. (2005), which surveyed 6,831 online consumers across 25 distinct website categories (ranging from online banking and retail to medical and portal sites), the three indicators exhibited high standardized factor loadings exceeding the conventional 0.70 threshold (all standardized $lambda ge 0.81$). The Average Variance Extracted (AVE) was established at 0.72, surpassing the 0.50 benchmark proposed by Fornell and Larcker (1981). This confirms that over 70% of the variance captured by the indicators is shared with the latent construct rather than measurement error.

Discriminant Validity

Discriminant validity confirms that WBI remains empirically distinct from its antecedent constructs, such as Online Trust, Perceived Site Quality, Navigational Ease, and Brand Equity. Using the Fornell-Larcker criterion, the square root of the AVE for WBI ($\sqrt{0.72} \approx 0.849$) consistently exceeded the inter-construct correlations between WBI and all other latent variables in structural equation models (where correlations typically ranged between $r = 0.42$ and $r = 0.68$). Subsequent studies applying the modern Heterotrait-Monotrait Ratio (HTMT) criterion report values well below the conservative 0.85 threshold, demonstrating structural divergence between conative intention and cognitive/affective trust.

Predictive and Criterion Validity

The scale demonstrates predictive validity against objective behavioral outcomes. Longitudinal investigations linking survey responses to server log files, clickstream analytics, and customer relationship management (CRM) databases show that WBI significantly predicts actual purchases ($R^2$ ranging from 0.18 to 0.34), repeated site authentication within a 90-day window ($p < 0.001$), and referral link dissemination. However, consistent with the intention-behavior gap documented in behavioral psychology, transactional intention accounts for higher variance in actual purchase behavior than registration intention accounts for long-term platform engagement.

Face and Content Validity Debate

Despite strong statistical convergence, the face validity of treating transaction, recommendation, and registration as an inseparable unidimensional block has drawn scrutiny. Researchers argue that in high-involvement websites (e.g., financial planning, legal counsel, healthcare), users may frequently register to consume information or recommend the platform due to exceptional educational content, while consciously refusing to execute monetary transactions on the site. Conversely, on low-involvement retail platforms, guest checkout allows transactions without registration. Thus, while statistically unified in aggregate cross-sectional samples, content validity requires researchers to verify that all three behaviors are pertinent to their investigated digital environment.

8. Reliability

The internal consistency and temporal stability of the Website Behavioural Intention scale have been established across numerous independent psychometric investigations:

Internal Consistency

• Cronbach's Alpha ($\alpha$): Across empirical replications, the scale yields Cronbach's alpha values ranging from 0.84 to 0.91, comfortably exceeding the 0.70 threshold for research instruments and the 0.80 benchmark for diagnostic utility.
• Composite Reliability (CR): Structural equation modeling assessments report CR values typically centered around 0.88, indicating excellent internal consistency of the latent construct indicators without suffering from item redundancy.
• Average Variance Extracted (AVE): Reported consistently at or above 0.72, demonstrating that the shared variance among the three operational indicators far outweighs error variance components.

Temporal and Cross-Category Stability

In multi-group invariance tests across the 25 website categories examined by Bart et al. (2005), metric and scalar invariance were confirmed across e-commerce categories (e.g., apparel, consumer electronics, travel booking). Test-retest reliability assessments conducted within 14-day intervals (where no website alterations occurred) yield stability coefficients ranging between $r = 0.76$ and $r = 0.83$ ($p < 0.001$), demonstrating that individual conative states remain relatively stable in the absence of external site interventions or negative service encounters.

9. Factor Analysis

The internal structure of the WBI has been evaluated using both Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA) within structural equation modeling (SEM) software environments such as LISREL, AMOS, Mplus, and lavaan (R):

Confirmatory Factor Analysis (CFA) Parameters

Under maximum likelihood estimation, a single-factor first-order CFA model yields robust fit indices when applied to the three indicators:

• $\chi^2 / \text{df}$ Ratio: Frequently non-estimable as a standalone saturated model ($df = 0$ for a 3-indicator single-factor model without constraints), but when embedded within larger structural measurement models, the overall system demonstrates $\chi^2 / \text{df}$ ratios between 1.85 and 2.45.
• Comparative Fit Index (CFI): Values consistently exceed 0.97, often reaching 0.99.
• Tucker-Lewis Index (TLI): Estimates consistently maintain levels above 0.96.
• Root Mean Square Error of Approximation (RMSEA): Values typically range from 0.032 to 0.052, with 90% confidence intervals well below the 0.08 acceptable threshold.
• Standardized Root Mean Square Residual (SRMR): Observed values routinely fall below 0.035.

Standardized Item Loadings

Within a standard reflective measurement specification, the three indicators demonstrate the following representative standardized loadings ($lambda$):

• Intention to Transact: $lambda = 0.82 – 0.88$ ($p < 0.001$, error variance $\theta_\epsilon \approx 0.23 – 0.33$)
• Intention to Recommend: $lambda = 0.84 – 0.89$ ($p < 0.001$, error variance $\theta_\epsilon \approx 0.21 – 0.29$)
• Intention to Register: $lambda = 0.80 – 0.85$ ($p < 0.001$, error variance $\theta_\epsilon \approx 0.28 – 0.36$)

Alternative Modeling: Unidimensional vs. Multi-Factor

While the single-factor model exhibits statistical parsimony and acceptable mathematical fit, methodologists have modeled WBI as a second-order latent factor composed of three separate first-order constructs (each measured by multi-item sub-indicators: e.g., 2 items for purchase, 2 items for WOM, 2 items for registration). In comparative model assessments, while the second-order model provides greater granularity, the 3-indicator single-factor formulation remains the preferred standard in large-scale structural modeling due to its parsimony and preservation of statistical degrees of freedom.

10. Instrument / Measurement Tool

The operational administration parameters and structural guidelines for the Website Behavioural Intention measurement tool are detailed below:

• Instrument Designation: Website Behavioural Intention (WBI) Scale.
• Primary Construct Measured: Conative digital consumer loyalty, specifically reflecting multi-faceted downstream post-visit intentions.
• Structural Typology: Self-report psychometric rating scale; typically specified as a reflective, single-factor latent variable.
• Item Inventory: Exactly 3 core behavioral items: (1) Intention to execute a commercial transaction/purchase; (2) Intention to recommend the platform to peers/network; (3) Intention to register/provide personal profile information.
• Response Formats:
  • Standard Scaling: 7-point Likert scale (1 = Strongly Disagree, 4 = Neither Agree nor Disagree, 7 = Strongly Agree) or 7-point subjective probability semantic differential (1 = Extremely Unlikely, 7 = Extremely Likely).
  • Original Bart et al. (2005) Calibration: 10-point numerical probability scale (1 = Very Low Probability / Strongly Disagree, 10 = Very High Probability / Strongly Agree).
• Scoring and Computational Procedures:
  • Composite Score: Calculated as the unweighted arithmetic mean of the 3 items (Sum of item ratings / 3).
  • Latent Variable Modeling: Retained as independent observed indicators loaded onto a latent WBI construct within a Structural Equation Modeling (SEM) framework, allowing estimation of indicator-specific measurement errors.
  • High vs. Low Intent Benchmarking: Mean scores $ge 5.5$ on a 7-point scale (or $ge 7.5$ on a 10-point scale) are classified as high behavioral loyalty; scores $le 3.0$ (or $le 4.0$ on a 10-point scale) indicate elevated customer churn risk or severe friction.
• Administration Duration: Approximately 1 to 2 minutes; suitable for post-task web usability evaluations, intercept exit surveys, and longitudinal consumer panels.

11. Permissions & Fee and Test Year

The Website Behavioural Intention scale was published in 2005 by the American Marketing Association (AMA) in the Journal of Marketing (Volume 69, Issue 4, pages 133–152). Academic research use of the scale items for non-profit scientific investigation, educational research, and dissertation inquiries is widely recognized under standard academic fair use provisions, provided that full bibliographic attribution is granted to Bart, Shankar, Sultan, and Urban (2005).

Commercial deployment—such as embedding the scale into proprietary enterprise customer experience management (CXM) platforms, commercial benchmarking engines, or monetization software—may fall under copyright restrictions held by the American Marketing Association or the respective journal publishers (Sage Publications). Researchers planning commercial integrations are advised to review the copyright clearance policies of Sage Publications / American Marketing Association or seek direct permissions.

12. References

The academic and psychometric foundations of the WBI are supported by the following literature:

• Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human Decision Processes, 50(2), 179–211. https://doi.org/10.1016/0749-5978(91)90020-T
• Bart, Y., Shankar, V., Sultan, F., & Urban, G. L. (2005). Are the drivers and role of online trust the same for all web sites and consumers? A large-scale exploratory study. Journal of Marketing, 69(4), 133–152. https://doi.org/10.1509/jmkg.2005.69.4.133
• Culnan, M. J., & Armstrong, P. K. (1999). Information privacy concerns, procedural fairness, and impersonal trust: An empirical investigation. Organization Science, 10(1), 104–115. https://doi.org/10.1287/orsc.10.1.104
• Davis, F. D. (1989). Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Quarterly, 13(3), 319–340. https://doi.org/10.2307/249008
• Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention, and behavior: An introduction to theory and research. Addison-Wesley.
• Fornell, C., & Larcker, D. F. (1981). Evaluating structural equation models with unobservable variables and measurement error. Journal of Marketing Research, 18(1), 39–50. https://doi.org/10.1177/002224378101800104
• Mayer, R. C., Davis, J. H., & Schoorman, F. D. (1995). An integrative model of organizational trust. Academy of Management Review, 20(3), 709–734. https://doi.org/10.5465/amr.1995.9508080332
• Morgan, R. M., & Hunt, S. D. (1994). The commitment-trust theory of relationship marketing. Journal of Marketing, 58(3), 20–38. https://doi.org/10.1177/002224299405800302
• Urban, G. L., Sultan, F., & Qualls, W. J. (2000). Placing trust at the center of web site design. Sloan Management Review, 42(1), 39–48.
• Zeithaml, V. A., Berry, L. L., & Parasuraman, A. (1996). The behavioral consequences of service quality. Journal of Marketing, 60(2), 31–46. https://doi.org/10.1177/002224299606000203

13. Items of the Scale