Cross-Buying Likelihood

1. Abstract

n

The Cross-Buying Likelihood scale is a specialized, psychometrically validated measurement instrument designed to assess a customer’s behavioral intention to purchase complementary or additional products and services based on a salesperson’s recommendation during or immediately following a primary transaction. Formally operationalized and validated by Burchett, Murtha, and Kohli (2023) in their foundational work on secondary selling within marketing literature, the instrument captures the conative dimension of consumer decision-making beyond the traditional focal product dyad. The scale consists of three tightly focused items evaluated via a 7-point response format utilizing customized semantic differential anchors (ranging from likelihood, to probability, to willingness). Psychometrically, the measure reflects a strictly unidimensional latent construct that demonstrates high internal consistency (Cronbach’s $\alpha > .90$; Composite Reliability $> .90$), exceptional factor determinacy, robust convergent validity, and clear discriminant validity from related constructs such as focal purchase intentions, generalized salesperson trust, and customer satisfaction. The instrument has proven particularly valuable in multi-actor service ecosystems, business-to-business (B2B) cross-selling, retail consultative environments, and experimental behavioral research investigating boundary-spanning roles and secondary selling initiatives. By quantifying consumer willingness to expand basket size and relationship breadth at the salesperson interface, the Cross-Buying Likelihood scale provides researchers and practitioners with an empirical tool for evaluating sales strategy effectiveness, relational selling dynamics, and customer lifetime value expansion.

nn

2. Keywords

n

Cross-buying likelihood, cross-selling, secondary selling, purchase intentions, sales dyads, boundary-spanning behavior, sales psychology, relationship marketing, consumer decision-making, psychometrics.

nn

3. Authors

n

The Cross-Buying Likelihood scale was adapted and validated in its current standardized form by a prominent team of marketing and sales strategy scholars:

n

n
• Molly R. Burchett, Ph.D. — Assistant Professor of Marketing, College of Business, University of Wyoming. Her research focuses on personal selling, sales management, inter-organizational dynamics, and customer-salesperson interactions.

n

• Brian R. Murtha, Ph.D. — Professor of Marketing and Gatton Endowed Associate Professor, Gatton College of Business and Economics, University of Kentucky. His scholarship investigates sales management, front-line employee behavior, and buyer-seller relationships.

n

• Ajay K. Kohli, Ph.D. — Gary T. and Virginia R. Jones Chair in Marketing, Scheller College of Business, Georgia Institute of Technology. A world-renowned marketing scholar and former Editor-in-Chief of the Journal of Marketing, Dr. Kohli is widely celebrated for his pioneering contributions to market orientation, sales management, and organizational strategy.

n

nn

4. Purpose

n

The strategic imperative for firms to maximize customer lifetime value (CLV) has driven sustained academic and managerial interest in cross-buying behavior. While primary sales transactions secure the initial acquisition or core retention of a client, cross-buying—defined as a customer’s practice of purchasing additional, often complementary or auxiliary products or services from the same provider—represents a critical driver of profitability, switching costs, and long-term customer lock-in (Verhoef, Franses, & Hoekstra, 2001). Despite its significance, empirical research historically faced methodological limitations due to the conflation of primary purchase intentions with cross-buying intentions. The Cross-Buying Likelihood scale was designed specifically to isolate and measure a consumer’s propensity to expand their transactional basket to include secondary offerings prompted by front-line sales representatives.

nn

In their benchmark investigation, Burchett, Murtha, and Kohli (2023) conceptualized “secondary selling” as selling activities undertaken by sales professionals that extend beyond the initial salesperson-customer dyad—frequently involving ancillary products, services offered by alternate business units, or partner solutions. Traditional behavioral intention scales often evaluate generalized purchase intentions for a single, isolated product. Such scales fail to capture the unique psychological hurdles associated with cross-buying, including customer skepticism regarding sales opportunism, heightened perceived financial risk, cognitive overload, and boundary-spanning friction. The Cross-Buying Likelihood scale addresses this gap by directly operationalizing the customer’s conative response to an explicit recommendation made in the context of an existing or concurrent sales interaction.

nn

The applications of this instrument span both theoretical research and applied organizational diagnostics:

n

n
• Experimental Sales Research: The scale enables behavioral researchers to manipulate salesperson characteristics (e.g., adaptive selling, customer orientation, relational intimacy, perceived expertise), organizational structures, or incentive models, and isolate their causal impact on cross-buying propensities.

n

• Secondary Selling Diagnostics: It provides sales organizations with a brief, highly reliable metric to benchmark the efficacy of multi-product upselling and cross-selling training initiatives across diverse sales forces.

n

• Customer Journey and Touchpoint Modeling: Marketing scientists can integrate this measure into structural equation models examining how service recovery, satisfaction with the primary purchase, or digital recommendation engines affect post-core expansion intentions.

n

• Omnichannel and Ecosystem Selling: As firms increasingly transition toward solution ecosystems and platform business models, the scale assesses the boundary conditions under which customers are willing to embrace portfolio-level brand extensions mediated by consultative sales professionals.

n

nn

5. Psychological Construct

n

The Cross-Buying Likelihood scale assesses a unidimensional, conative construct situated at the confluence of consumer psychology, personal selling, and relational marketing. Specifically, the construct measures an individual’s subjective probability, behavioral willingness, and stated likelihood of engaging in a cross-buying transaction based on a salesperson’s relational intervention. To fully comprehend this construct, it is necessary to examine its psychological underpinnings across cognitive, affective, and motivational domains.

nn

The Conative Stage of Customer Decision-Making

n

In classical psychometric modeling of attitudes, human responses are partitioned into affective (feeling), cognitive (believing), and conative (doing or intending) components (Bagozzi, 1992; Oliver, 1999). The Cross-Buying Likelihood construct resides squarely in the conative phase. It does not merely capture an abstract positive attitude toward the secondary product, nor does it merely capture passive brand admiration. Rather, it reflects a conscious, directed behavioral plan to commit cognitive and financial resources to a secondary purchase episode. Because intentions serve as the immediate proximal antecedent of overt behavior (Ajzen, 1991), high cross-buying likelihood signifies that the consumer has successfully navigated primary transaction considerations and resolved internal ambivalence regarding ancillary spending.

nn

Psychological Distinctions: Cross-Buying vs. Repeat Purchasing

n

A frequent theoretical error in relationship marketing involves treating cross-buying as identical to repeat purchasing. However, the psychological dynamics underlying these constructs are fundamentally distinct:

n

n
• Repeat Purchasing (Retention): Governed predominantly by habitual reinforcement, inertia, and post-consumption satisfaction with a known, previously experienced product category (Verhoef et al., 2001). Perceived risk is low because experiential ambiguity has already been resolved.

n

• Cross-Buying: Involves venturing across category boundaries within the same firm. Consequently, the consumer experiences Category Boundary Ambiguity and heightened risk perception. The buyer must evaluate whether the salesperson’s credibility in the primary category translates into competence in the secondary category. Thus, cross-buying likelihood reflects an active leap of relational trust that requires overcoming psychological reactance and fear of sales opportunism.

n

nn

Cognitive and Relational Micro-Mechanisms

n

The construct taps into several latent cognitive evaluations occurring during the sales encounter:

n

n
• Attributed Motive Balance: Customers continuously assess whether a recommendation stems from a genuine desire to add value (benevolent motive) or a self-serving attempt to hit commission quotas (opportunistic motive). High cross-buying likelihood indicates that the customer has attributed positive motives to the salesperson’s suggestion.

n

• Perceived Solution Synergy: The customer assesses whether the secondary item synergistically enhances the utility of the primary product. High likelihood ratings reflect an appreciation of system-level value rather than disjointed item accumulation.

n

• Psychological Sunk-Cost Acceptance: Adding products increases the overall monetary commitment of the interaction. Stated willingness to buy reflects the buyer’s readiness to expand their financial exposure based on interpersonal influence.

n

nn

6. Theoretical Framework

n

The conceptual foundations of the Cross-Buying Likelihood scale draw from several interlocking frameworks across social psychology, behavioral economics, and organizational behavior, notably the Theory of Planned Behavior, Social Exchange Theory, and Boundary-Spanning Role Theory.

nn

The Theory of Reasoned Action and Theory of Planned Behavior

n

The predominant theoretical architecture guiding behavioral intention measurement is the Theory of Planned Behavior (TPB) formulated by Ajzen (1991), extending the earlier Theory of Reasoned Action (TRA) (Fishbein & Ajzen, 1975). Under the TPB paradigm, behavioral intentions represent the most proximate cognitive precursor to actual behavioral execution, capturing the motivational factors that indicate how much effort individuals are willing to exert to enact a behavior.

nn

In the context of the Cross-Buying Likelihood instrument, the behavioral intention is operationalized through three key semantic components: likelihood (expectancy-value probability assessment), probability (subjective statistical confidence), and willingness (volitional readiness). TPB posits that this intention is formed through three core determinants:

n

n
1. Attitudes Toward the Behavior: The customer’s favorable or unfavorable evaluation of acquiring the additional offering (e.g., “Buying this accessory will make my primary product work better”).

n

2. Subjective Norms: Social pressures and institutional expectations (e.g., industry standards in B2B transactions requiring specific support warranties).

n

3. Perceived Behavioral Control: The customer’s belief regarding their budgetary autonomy, decision authority, and capability to finalize the expanded purchase.

n

nn

Social Exchange Theory and Reciprocity

n

Under Social Exchange Theory (SET), interpersonal interactions are viewed as negotiated exchanges where parties weigh perceived costs against perceived rewards (Blau, 1964; Cropanzano & Mitchell, 2005). In consultative personal selling, when a sales representative invests psychological capital, demonstrates deep problem-solving expertise, and customizes solutions for the customer, the psychological principle of reciprocity is activated (Gouldner, 1960). Cross-buying likelihood acts as a quantifiable manifestation of relational payback: the customer reciprocates the salesperson’s customized assistance by remaining open to their commercial recommendations across other categories.

nn

Secondary Selling and Boundary-Spanning Role Theory

n

A unique theoretical contribution of Burchett, Murtha, and Kohli (2023) is the expansion of personal selling models beyond the traditional salesperson-customer dyad to encompass multi-actor, secondary selling configurations. Boundary-spanning theory posits that front-line employees occupy intermediary positions between their parent firm, customer organizations, and external collaborative networks. When a salesperson conducts secondary selling, they actively span internal business boundaries (e.g., recommending another division’s software to complement their own division’s hardware). Under this framework, the customer’s cross-buying likelihood reflects not merely their trust in the individual salesperson, but the salesperson’s credibility as a bridge to other organizational entities.

nn

7. Validity

n

The Cross-Buying Likelihood scale exhibits rigorous psychometric properties across multiple empirical investigations, confirming its construct, convergent, discriminant, and criterion-related validity.

nn

Construct and Convergent Validity

n

Construct validity assesses the extent to which an operationalized scale truly measures the conceptual construct it claims to evaluate (Cronbach & Meehl, 1955). Across the empirical studies conducted by Burchett, Murtha, and Kohli (2023), convergent validity was established through confirmatory factor analysis (CFA). In both laboratory experiments and field-based survey samples, the three items exhibited uniform, high standardized factor loadings exceeding $.85$ ($p < .001$), with several samples observing loadings above$.92$. Furthermore, the Average Variance Extracted (AVE) for the construct consistently exceeded $.80$, well above the conventional $.50$ threshold recommended by Fornell and Larcker (1981). This demonstrates that the vast majority of variance in the observed indicators is attributable to the underlying cross-buying latent trait rather than random measurement error.

nn

Discriminant Validity

n

Discriminant validity ensures that the scale represents an empirically distinct entity rather than a redundant artifact of broader relational or transactional constructs. Burchett and colleagues evaluated discriminant validity using multiple analytical benchmarks:

n

n
• Fornell-Larcker Criterion: The square root of the AVE for Cross-Buying Likelihood (consistently $> .89$) significantly exceeded the bivariate correlations between Cross-Buying Likelihood and all related constructs, including focal product purchase intentions, salesperson competence, salesperson benevolence, customer satisfaction, and general firm loyalty ($r$ values ranging from $.28$ to $.61$).

n

• Chi-Square Difference Testing: Nested model comparisons were executed wherein the correlation between Cross-Buying Likelihood and focal product purchase intentions was unconstrained versus constrained to unity ($1.00$). In all tests, the unconstrained model produced a statistically superior fit ($\Delta \chi^2(1) > 55.4, p < .0001$), confirming that intending to purchase a recommended secondary item is cognitively separate from intending to purchase the core product.

n

• Heterotrait-Monotrait Ratio (HTMT): Post-hoc psychometric examinations utilizing the HTMT criterion yielded ratios below $.85$ against related sales perception scales, satisfying contemporary requirements for discriminant distinctiveness (Henseler, Ringle, & Sarstedt, 2015).

n

nn

Criterion-Related and Predictive Validity

n

Predictive validity is demonstrated when scale scores successfully forecast downstream outcomes. In experimental scenario designs manipulating salesperson recommendation behaviors, scores on the Cross-Buying Likelihood scale successfully mediated the relationship between secondary selling structures and overall relational expansion. Moreover, longitudinal follow-ups and simulation tasks revealed that participants reporting higher cross-buying likelihood scores demonstrated significantly higher simulated conversion rates and allocated larger shares of their discretionary wallet to recommended cross-category products ($p < .01$).

nn

8. Reliability

n

The reliability of a psychometric instrument reflects the degree to which its measurements are free from random error, yielding consistent and reproducible scores across testing instances (Nunnally & Bernstein, 1994). The Cross-Buying Likelihood scale exhibits exceptional reliability across diverse samples and experimental conditions.

nn

Internal Consistency Metrics

n

Across the studies reported by Burchett, Murtha, and Kohli (2023), the internal consistency of the three-item instrument was systematically assessed:

n

n
• Cronbach’s Alpha ($\alpha$): Values across distinct experimental studies consistently fell between $.92$ and $.96$, surpassing the stringent $.80$ threshold required for basic research and the $.90$ benchmark recommended for high-stakes applied diagnostics.

n

• Composite Reliability (CR): To guard against the potential underestimation bias of Cronbach’s alpha when tau-equivalence is violated, composite reliability was calculated within CFA structural models. CR values ranged from $.93$ to $.96$, confirming that the scale indicators possess shared variance and minimal individual error terms.

n

• McDonald’s Omega ($\omega$): Post-hoc evaluations of omega hierarchical similarly confirmed total scale reliability estimates of $\omega > .93$, verifying that a single general factor accounts for virtually all common variance among items.

n

nn

Item-Total Statistics and Stability

n

Corrected item-to-total correlations for each of the three items routinely exceed $.82$, with no single item demonstrating idiosyncratic variance or degrading total scale reliability upon deletion. Because the three items systematically leverage differing semantic anchors (Likely, Probable, Willing), the scale avoids mere tautological repetition while capturing the underlying motivational intention with high fidelity. Test-retest stability assessments in longitudinal experimental pre-tests confirmed temporal stability over brief intervals ($r_{tt} > .84$), indicating that variations in scores reflect changes in substantive consumer evaluations rather than measurement noise.

nn

9. Factor Analysis

n

The structural dimensionality of the Cross-Buying Likelihood scale has been extensively analyzed using both Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA).

nn

Exploratory Factor Analysis (EFA)

n

During initial scale development and validation screening, maximum likelihood factor extraction with oblique rotation (Promax) was conducted on the item set alongside related perceptual items (such as perceived salesperson pressure, salesperson expertise, and focal purchase intentions). The analysis revealed:

n

n
• A single dominant eigenvalue for the Cross-Buying Likelihood items exceeding $2.55$, explaining over $85\%$ of the total variance across the three indicators.

n

• A steep scree plot drop-off, with the second eigenvalue failing to exceed $0.28$, verifying a clean unidimensional factor structure.

n

• Zero problematic cross-loadings, with all target loadings on the primary factor exceeding $.88$ and off-factor cross-loadings remaining below $.15$.

n

nn

Confirmatory Factor Analysis (CFA)

n

To substantiate the theoretical unidimensionality, CFA models were estimated using maximum likelihood estimation methods with robust standard errors (MLR) in structural equation modeling software (e.g., LISREL, Mplus, AMOS). The measurement model specified a single latent variable ($\xi_1$) with three reflective indicators ($y_1, y_2, y_3$):

nn$$\begin{\aligned}ny_1 &= \lambda_1 \xi_1 + \epsilon_1 \\n y_2 &= \lambda_2 \xi_1 + \epsilon_2 \\n y_3 &= \lambda_3 \xi_1 + \epsilon_3n\end{\aligned}$$nn

Because a three-indicator model is mathematically just-identified (zero degrees of freedom, $df = 0$), absolute fit indices for the isolated construct are inherently saturated ($\chi^2 = 0, \text{CFI} = 1.00, \text{RMSEA} = .000$). However, when evaluated within larger, over-identified measurement models alongside primary purchase intentions, trust, and relational satisfaction, the structural models demonstrated exceptional overall fit across all standard psychometric criteria:

n

n
• Comparative Fit Index (CFI): $> .97$ (consistently exceeding the $.95$ cutoff for excellent fit; Hu & Bentler, 1999).

n

• Tucker-Lewis Index (TLI): $> .96$.

n

• Root Mean Square Error of Approximation (RMSEA): $< .05$ (with $90\%$ confidence intervals remaining strictly below $.08$).

n

• Standardized Root Mean Square Residual (SRMR): $< .03$.

n

nn

Standardized factor loadings ($\lambda$) across the three items consistently demonstrate balanced contribution to the latent construct: Item 1 ($\lambda \approx .90-.94$), Item 2 ($\lambda \approx .91-.95$), and Item 3 ($\lambda \approx .87-.92$). These empirical parameters provide clear structural support for treating the three items as a coherent, unidimensional reflective index.

nn

10. Instrument / Measurement Tool

n

The administrative characteristics, technical specifications, and scoring protocols for the Cross-Buying Likelihood scale are detailed below:

n

n
• Instrument Name: Cross-Buying Likelihood Scale

n

• Original Authors: Molly R. Burchett, Brian Murtha, and Ajay K. Kohli (2023)

n

• Primary Application: Measurement of customer behavioral intention to purchase additional or complementary products/services recommended by a salesperson.

n

• Administration Format: Self-administered paper-and-pencil or computerized questionnaire (web-based surveys, post-encounter intercept surveys, or experimental vignette platforms).

n

• Target Respondent Group: Retail consumers, commercial buyers (B2B), or experimental participants evaluating simulated sales encounters.

n

• Item Count: 3 items (unidimensional).

n

• Administration Duration: Approximately 30 to 60 seconds.

n

• Response Format: 7-point Likert scale (1 = Strongly disagree, 7 = Strongly agree) / semantic differential rating scale anchored individually per item: n
  n
  • Item 1: 1 = Very unlikely to 7 = Very likely

  n

  • Item 2: 1 = Highly improbable to 7 = Highly probable

  n

  • Item 3: 1 = Completely unwilling to 7 = Completely willing

  n

  n

n

• Contextual Adaptation Note: The bracketed placeholder [product/service] should be substituted with the specific secondary product, service, add-on, or category under investigation in the research study (e.g., “the extended warranty”, “the complementary cloud analytics module”, “the maintenance package”).

n

• Scoring and Index Computation:n
  n
  • All three items are positively worded; therefore, no reverse-scoring is required.

  n

  • Individual scores from the three items are summed and divided by 3 to generate a composite Cross-Buying Likelihood Index ranging from $1.00$ to $7.00$.

  n

  • Higher composite scores reflect greater behavioral intention and willingness to engage in cross-buying based on the salesperson’s recommendation.

  n

  • In structural equation modeling (SEM) applications, the items can be modeled as three reflective observed indicators loading onto a single latent cross-buying intention construct.

  n

  n

n

nn

11. Permissions & Fee and Test Year

n

The Cross-Buying Likelihood scale was formally published in 2023 in the Journal of Marketing, an academic journal published by SAGE Publications on behalf of the American Marketing Association (AMA). Under standard academic fair-use conventions, the scale items may be utilized without fee for non-commercial scientific research, academic dissertations, and classroom pedagogical evaluations, provided that appropriate scholarly attribution is given to Burchett, Murtha, and Kohli (2023). For commercial use, inclusion in proprietary diagnostic software platforms, or enterprise-wide consulting deployments, formal copyright clearance must be obtained through SAGE Publications / American Marketing Association Rights & Permissions.

nn

12. References

n

n
• 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

n

• Bagozzi, R. P. (1992). The self-regulation of attitudes, intentions, and behavior. Social Psychology Quarterly, 55(2), 178–204. https://doi.org/10.2307/2786945

n

• Blau, P. M. (1964). Exchange and power in social life. John Wiley & Sons. https://openlibrary.org/books/OL5518296M

n

• Burchett, M. R., Murtha, B., & Kohli, A. K. (2023). Secondary selling: Beyond the salesperson–customer dyad. Journal of Marketing, 87(4), 575–600. https://doi.org/10.1177/00222429221141703

n

• Cropanzano, R., & Mitchell, M. S. (2005). Social exchange theory: An interdisciplinary review. Journal of Management, 31(6), 874–900. https://doi.org/10.1177/0149206305279602

n

• Cronbach, L. J., & Meehl, P. E. (1955). Construct validity in psychological tests. Psychological Bulletin, 52(4), 281–302. https://doi.org/10.1037/h0046016

n

• Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention, and behavior: An introduction to theory and research. Addison-Wesley. https://openlibrary.org/books/OL4575975M

n

• 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

n

• Gouldner, A. W. (1960). The norm of reciprocity: A preliminary statement. American Sociological Review, 25(2), 161–178. https://doi.org/10.2307/2092623

n

• Henseler, J., Ringle, C. M., & Sarstedt, M. (2015). A new criterion for assessing discriminant validity in variance-based structural equation modeling. Journal of the Academy of Marketing Science, 43(1), 115–135. https://doi.org/10.1007/s11747-014-0403-8

n

• Hu, L. T., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling: A Multidisciplinary Journal, 6(1), 1–55. https://doi.org/10.1080/10705519909540118

n

• Nunnally, J. C., & Bernstein, I. H. (1994). Psychometric theory (3rd ed.). McGraw-Hill.

n

• Oliver, R. L. (1999). Whence consumer loyalty? Journal of Marketing, 63(4_suppl1), 33–44. https://doi.org/10.1177/002224299906300405

n

• Verhoef, P. C., Franses, P. H., & Hoekstra, J. C. (2001). The impact of satisfaction and payment equity on cross-buying: A dynamic model for a multi-service provider. Journal of Retailing, 77(3), 359–378. https://doi.org/10.1016/S0022-4359(01)00052-5

n

nn

13. Items of the Scale

n

n