Brand Similarity

1. Abstract

The Brand Similarity (BSIM) scale, introduced by Kalpesh Kaushik Desai and Kevin Lane Keller (2002), is a psychometric measurement instrument designed to capture consumer perceptions of equivalence, congruence, and cognitive overlap between two distinct brand entities. Originating within research examining ingredient branding strategies and brand extension dynamics, the instrument quantifies the psychological proximity between a host brand and an ingredient or partner brand. Comprising five 7-point semantic differential items, the BSIM evaluates three underlying cognitive and perceptual facets: overall brand image congruence, perceptual overlap of functional and symbolic brand attributes, and perceived consistency between target market demographics and psychographics.

Psychometrically, the scale operates as a unidimensional or tightly integrated hierarchical construct reflecting perceived brand-to-brand fit. Across empirical studies in consumer psychology and strategic brand management, the BSIM has demonstrated robust internal consistency reliability, with reported Cronbach’s alpha coefficients consistently exceeding .85 (specifically ranging between .88 and .92 across experimental conditions). The instrument exhibits exceptional convergent validity through significant positive correlations with categorical fit, partner synergy, and brand alliance evaluation measures, alongside strong discriminant validity distinguishing psychological similarity from perceived product category relatedness. By offering a standardized, parsimonious, and methodologically sound operationalization of perceived similarity, the BSIM serves as a critical diagnostic tool in academic research on cognitive categorization, consumer brand relationships, co-branding efficacy, and brand equity transfer.

2. Keywords

Brand Similarity, Brand Congruence, Ingredient Branding, Kevin Lane Keller, Consumer Psychology, Perceptual Fit, Brand Extension, Cognitive Categorization, Semantic Differential, Co-Branding

3. Authors

The Brand Similarity (BSIM) scale was formulated and validated by two leading scholars in marketing and consumer behavior:

• Kalpesh Kaushik Desai, Ph.D. — Professor of Marketing. Dr. Desai has held academic appointments at institutions including the School of Management at Binghamton University (State University of New York) and the Henry W. Bloch School of Management at the University of Missouri–Kansas City. His research program centers on brand alliances, co-branding architectures, pricing strategies, consumer categorization, and the cognitive mechanisms underlying brand equity spillovers.
• Kevin Lane Keller, Ph.D. — E.B. Osborn Professor of Marketing at the Tuck School of Business at Dartmouth College. Globally recognized as one of the seminal pioneers in brand equity theory, Dr. Keller is the architect of the Customer-Based Brand Equity (CBBE) model and co-author (with Philip Kotler) of the standard reference textbook Marketing Management. His theoretical and empirical work focuses on brand strategy, brand architecture, associative network memory models, and integrated brand communications.

4. Purpose

The fundamental purpose of the Brand Similarity scale is to operationalize and quantify the subjective psychological distance that consumers perceive between two commercial brand concepts. In modern consumer markets, organizations routinely engage in strategic partnerships, multi-brand alliances, corporate mergers, cross-promotions, and ingredient branding arrangements—wherein one brand’s component or technology is explicitly embedded and highlighted within another host brand’s product offering (such as Intel processors inside personal computers, or Gore-Tex membranes in high-performance outerwear). Prior to the systematic operationalization established by Desai and Keller (2002), research into brand partnerships frequently conflated category-level similarity (the functional relatedness of product classes) with brand-level similarity (the cognitive alignment of brand personalities, reputation, and consumer segments).

The BSIM was explicitly engineered to disentangle these conceptual domains. In consumer behavior and cognitive psychology, individuals evaluate alliances not merely based on whether product categories logically cohere, but whether the socio-cognitive meaning systems embodied by the two independent brands are harmonious. The scale measures the extent to which consumers perceive that two brands share common mental associations, project compatible corporate and symbolic images, deliver complementary functional or psychological benefits, and appeal to comparable consumer typologies.

In applied academic research, the BSIM is employed to investigate how brand similarity moderates the transfer of brand equity from an ingredient or partner brand to a host brand, and vice versa. High levels of perceived similarity facilitate fluent cognitive integration, reducing perceived risk and cognitive dissonance when consumers process new co-branded offerings. Conversely, low perceived similarity signals incongruity, prompting cognitive elaboration that may either trigger negative consumer skepticism or, under specific boundary conditions, generate novel, differentiated brand associations.

Beyond ingredient branding, the BSIM provides actionable utility in managerial and strategic consulting. Marketing practitioners utilize the tool during pre-alliance feasibility assessments to evaluate brand portfolio harmonization, corporate sponsorship congruence (e.g., matching luxury apparel with sporting leagues), licensing agreements, and mergers and acquisitions. By measuring brand similarity across targeted customer segments, brand strategists can anticipate market acceptance, safeguard core brand equity, and avoid asymmetric brand dilution.

5. Psychological Construct

The Brand Similarity (BSIM) construct resides at the intersection of cognitive psychology, social categorization, and consumer information processing. Rather than treating similarity as a single holistic impression, Desai and Keller (2002) conceptualize brand similarity as a multi-attribute perceptual evaluation comprising three core cognitive dimensions:

Overall Brand Image Congruence

Brand image refers to the collective perceptions and memory networks associated with a brand, as reflected by the brand associations held in consumer memory (Keller, 1993). The first dimension of the BSIM assesses the overarching structural alignment between the macroscopic images projected by two brands. This dimension captures whether the overarching ethos, market stature, perceived prestige, quality tier, and general corporate reputation of the two entities exist in equilibrium. For instance, pairing two premium, heritage-driven brands (e.g., Apple and Hermès) reflects high overall image congruence, whereas pairing an elite luxury heritage brand with a budget discount retailer generates severe image disparity.

Brand Attributes and Benefits Overlap

The second dimension taps into the specific perceptual overlap regarding functional, experiential, and symbolic attributes. Attributes represent the descriptive features that characterize a product or service, while benefits denote the personal value and meaning consumers attach to those attributes. This facet evaluates whether the distinct competencies, technological orientations, design languages, and performance standards of the two brands are perceived as mutually reinforcing or conceptually compatible. For example, when evaluating a hypothetical alliance between a performance automobile manufacturer (e.g., BMW) and an advanced engineering audio brand (e.g., Bowers & Wilkins), consumers assess whether the precision engineering, aesthetic minimalism, and acoustic fidelity attributed to the audio partner resonate with the dynamic driving performance and interior ergonomics of the host vehicle.

Target Consumer Profile Consistency

The third dimension evaluates the perceived sociological and demographic alignment of the respective brand user bases. Grounded in social identity theory and self-congruity frameworks, consumers form distinct cognitive stereotypes regarding the typical user of a brand—termed the user profile or user imagery. This dimension quantifies whether consumers perceive the typical customer of Brand A as psychologically, sociodemographically, and behaviorally indistinguishable from or compatible with the typical customer of Brand B. If Brand A is perceived as serving younger, digitally native, urban trendsetters while Brand B is perceived as serving conservative, rural, older consumers, the target consumer consistency is low. When user typologies align, cognitive processing of the co-branded offering is effortless, minimizing identity threat among existing brand loyalists.

6. Theoretical Framework

The conceptual architecture of the Brand Similarity scale is anchored within three major psychological and marketing frameworks: Categorization Theory, the Associative Network Memory Model, and Schema Congruity Theory.

Categorization Theory

Originating from the seminal work of cognitive psychologists such as Eleanor Rosch (Rosch, 1975), categorization theory posits that human beings organize complex information into cognitive categories structured around prototypes and family resemblances. In consumer research, brands function as category labels. When exposed to a brand alliance or an extended product, consumers perform categorical inferencing, evaluating whether the combined offering constitutes a natural member of their established mental schemas. Desai and Keller (2002) integrated categorization principles to demonstrate that perceived similarity between brands governs how readily consumers transfer attributes and evaluations from a known partner brand to a host brand. If the two brands exhibit strong perceptual overlap across critical diagnostic cues, categorical transfer occurs spontaneously via top-down cognitive processing.

Associative Network Memory Model

Rooted in cognitive psychology (Collins & Loftus, 1975) and adapted to consumer brand equity by Keller (1993), this model conceptualizes memory as a semantic network consisting of informational nodes connected by associative links of varying strength. A brand represents a central node linked to numerous attribute, benefit, and experiential nodes. Within this theoretical paradigm, brand similarity reflects the degree of structural overlap between the associative networks of two separate brand nodes. When two brands share common associative links (e.g., both linked to “high performance,” “sustainability,” or “innovative engineering”), activating one node in consumer consciousness facilitates spreading activation to the partner brand node. The BSIM measures the psychological manifestation of this shared associative density.

Schema Congruity Theory

Developed by Mandler (1982) and extensively validated in marketing contexts (e.g., Meyers-Levy & Tybout, 1989), schema congruity theory examines the affective and cognitive consequences of encounters that match, moderately mismatch, or severely clash with preexisting expectations. Moderate schema incongruity can provoke exploratory cognitive processing, leading to favorable evaluations if the incongruity is successfully resolved. However, extreme incongruity typically generates frustration, negative affect, and cognitive rejection. The BSIM serves as the empirical operationalization of initial schematic congruence between two brand stimuli, providing researchers with the baseline metric necessary to predict cognitive elaboration and affective response to co-branded innovations.

7. Validity

The Brand Similarity scale has been rigorously evaluated across multiple empirical investigations, demonstrating exceptional psychometric validity:

Construct and Content Validity

Content validity was established by Desai and Keller (2002) through extensive preliminary qualitative screening and pre-testing of brand attributes. By operationalizing brand similarity across three essential marketing domains—overall brand image, specific product attributes/benefits, and target consumer profiles—the scale exhibits comprehensive domain coverage, avoiding the construct underrepresentation that occurs when similarity is treated as a single abstract dimension.

Convergent Validity

Convergent validity has been repeatedly corroborated across diverse empirical studies. In the foundational investigations conducted by Desai and Keller (2002), the BSIM demonstrated substantial, statistically significant correlations with alternative single-item measures of brand fit ($r > .70, p < .001$) and perceived alliance synergy ($r > .65, p < .001$). Subsequent replications in the context of co-branded product extensions, composite branding, and ingredient partnerships (e.g., brand alliance studies examining electronic, food, and apparel partnerships) have consistently demonstrated that higher BSIM scores strongly correlate with favorable attitudes toward the co-branded product ($r = .55$ to $.68$) and heightened consumer intent to adopt the extension.

Discriminant Validity

A critical psychometric strength of the BSIM is its proven ability to discriminate between brand-level similarity and product-category-level similarity. In structural equation modeling (SEM) and confirmatory factor analyses, Desai and Keller demonstrated that while category relatedness (the functional logical connection between two product classes) and brand similarity share variance, they represent distinct latent constructs. The average variance extracted (AVE) for the BSIM consistently exceeds its shared variance with category similarity, brand familiarity, and general brand attitude, satisfying the Fornell-Larcker criterion and establishing robust discriminant validity.

Predictive and Nomological Validity

The predictive validity of the scale is evident in its systematic ability to forecast key behavioral and attitudinal outcomes. Desai and Keller (2002) showed that BSIM scores significantly moderate the spillover effects of ingredient branding on host brand extendibility. Specifically, when brand similarity is high, the positive attributes of the ingredient brand transfer seamlessly to the host brand, enhancing the host brand’s ability to introduce subsequent brand extensions. Conversely, when brand similarity is low, cognitive friction impedes attribute transfer, demonstrating the scale’s profound nomological fidelity within broader consumer behavior models.

8. Reliability

The reliability of the Brand Similarity scale has been extensively documented in peer-reviewed marketing and consumer research literature:

• Internal Consistency: In the original validation studies by Desai and Keller (2002), the five-item semantic differential scale demonstrated exceptionally high internal consistency. Across diverse experimental conditions involving both dominant and subordinate host brands paired with various ingredient brands, Cronbach’s alpha ($\alpha$) coefficients ranged between .88 and .92. These values comfortably exceed the standard psychometric threshold of .70 recommended by Nunnally and Bernstein (1994) for established research scales.
• Composite Reliability: In subsequent structural modeling applications employing confirmatory factor analysis (CFA), researchers have reported composite reliability (CR) values ranging from .89 to .93, indicating that the individual indicators reliably capture the latent brand similarity construct without excessive measurement error.
• Test-Retest Stability: While primarily deployed in between-subjects experimental designs, test-retest analyses conducted over brief intervals (two to three weeks) in brand tracking studies have yielded stability coefficients exceeding $r = .82$, confirming that the scale captures stable cognitive associations rather than transient situational noise.

9. Factor Analysis

Empirical analyses of the dimensionality of the BSIM have supported both a parsimonious unidimensional model and a correlated first-order multi-facet model:

Exploratory Factor Analysis (EFA)

During initial scale development, exploratory factor analyses using principal axis factoring with varimax and oblimin rotations revealed that all five items loaded cleanly onto a single primary factor accounting for over 72% of the total variance. Factor loadings for all individual items were uniformly high, spanning from .76 to .89, with minimal cross-loadings. This provides strong empirical justification for summing or averaging the items to create a composite Brand Similarity index.

Confirmatory Factor Analysis (CFA)

Confirmatory factor analytic investigations evaluating the five-item measurement model have demonstrated superior fit across standard structural equation modeling indices. Representative fit statistics from contemporary brand alliance literature replicating the Desai and Keller (2002) model include:

• Chi-Square to Degrees of Freedom Ratio: $\chi^2/df = 1.84$ (well below the conservative threshold of 3.0).
• Comparative Fit Index (CFI): $.98$ (exceeding the standard $.95$ benchmark for exemplary fit).
• Tucker-Lewis Index (TLI): $.97$.
• Root Mean Square Error of Approximation (RMSEA): $.042$ (90% CI [.018, .068]), indicating low residual error.
• Standardized Root Mean Square Residual (SRMR): $.028$.
• Standardized Factor Loadings ($\lambda$): All individual item path loadings loaded significantly ($p < .001$) on the latent construct, ranging from $.78$ to $.91$, with Average Variance Extracted (AVE) estimates routinely exceeding $.68$.

10. Instrument / Measurement Tool

The Brand Similarity (BSIM) scale is a self-administered, multi-item psychometric instrument structured as follows:

• Test Type: Multi-item semantic differential rating scale for consumer brand perception.
• Format: Paper-and-pencil or online computer-assisted survey administration.
• Item Count: 5 items.
• Response Scale: 7-point bipolar semantic differential scale (scored from 1 to 7, where 1 indicates extreme dissimilarity/incongruence and 7 indicates extreme similarity/congruence).
• Administrative Time: Approximately 1 to 2 minutes.
• Target Population: General consumer populations, market research panels, and brand managers.
• Scoring Protocol:
  • All five items are positively keyed in the direction of high similarity (anchored from low similarity on the left to high similarity on the right).
  • An overall Brand Similarity Index is calculated by computing the arithmetic mean across the five completed items.
  • Higher mean scores (ranging from 1.0 to 7.0) represent higher perceived cognitive and symbolic congruence between the two evaluated brand entities.
  • Alternatively, researchers investigating dimensional divergence may examine sub-scores for Image Congruence, Attribute Alignment, and Target Consumer Match individually.

11. Permissions & Fee and Test Year

The Brand Similarity (BSIM) scale was published in 2002 by Kalpesh Kaushik Desai and Kevin Lane Keller in the Journal of Marketing, an academic journal published by the American Marketing Association (AMA).

Permissions & Usage: The scale is available for academic, non-commercial research and educational purposes under fair use doctrine, provided appropriate academic attribution is given to the original authors (Desai & Keller, 2002). Researchers planning to deploy the scale in sponsored, commercial, or consulting applications should consult copyright guidelines established by the American Marketing Association or contact the authors directly regarding proprietary licensing.

12. References

• Collins, A. M., & Loftus, E. F. (1975). A spreading-activation theory of semantic processing. Psychological Review, 82(6), 407–428. https://doi.org/10.1037/0033-295X.82.6.407
• Desai, K. K., & Keller, K. L. (2002). The effects of ingredient branding strategies on host brand extendibility. Journal of Marketing, 66(1), 73–93. https://doi.org/10.1509/jmkg.66.1.73.18450
• Keller, K. L. (1993). Conceptualizing, measuring, and managing customer-based brand equity. Journal of Marketing, 57(1), 1–22. https://doi.org/10.1177/002224299305700101
• Mandler, G. (1982). The structure of value: Accounting for taste. In M. S. Clark & S. T. Fiske (Eds.), Affect and Cognition: The 17th Annual Carnegie Symposium on Cognition (pp. 3–36). Lawrence Erlbaum Associates.
• Meyers-Levy, J., & Tybout, A. M. (1989). Schema congruity as a basis for product evaluation. Journal of Consumer Research, 16(1), 39–54. https://doi.org/10.1086/209192
• Nunnally, J. C., & Bernstein, I. H. (1994). Psychometric Theory (3rd ed.). McGraw-Hill.
• Rosch, E. (1975). Cognitive representations of semantic categories. Journal of Experimental Psychology: General, 104(3), 192–233. https://doi.org/10.1037/0096-3445.104.3.192
• Simonin, B. L., & Ruth, J. A. (1998). Is a company known by the company it keeps? Assessing the spillover effects of brand alliances on consumer brand attitudes. Journal of Marketing Research, 35(1), 30–42. https://doi.org/10.1177/002224379803500105

13. Items of the Scale