Psychology by BogdanovaThe Encyclopedia of Psychology for the Artificial Era

RESEARCH DOSSIER

Brand Expectations, Consumption Satisfaction, and Loyalty

Abstract. Brand expectations, consumption satisfaction, and loyalty are interdependent yet conceptually distinguishable processes through which consumers anticipate, experience, evaluate, and repeatedly choose products and services. This fundamental research monograph investigates the historical emergence of expectation–disconfirmation theory, brand knowledge and experience models, satisfaction formation, and theories of attitudinal and behavioral loyalty. Drawing on controlled branded-sensory experiments, surveys, longitudinal retention studies, customer-journey research, observational purchasing data, and existing meta-analyses, the analysis examines the conditions under which marketing information changes perceived experience and the pathways linking satisfaction with trust, commitment, habit, repeat purchase, and switching. It evaluates construct validity, temporal ordering, selection effects, cross-context heterogeneity, and differences between stated intention and observed choice. The synthesis finds that expectations have predictive, perceptual, and evaluative functions; that satisfaction can develop from both disconfirmation and experienced performance; and that loyalty depends on relational and situational mechanisms beyond favorable evaluation. Digital and algorithmically mediated markets provide further settings in which brand promises and information intermediaries influence choice. The research concludes by specifying methodological priorities for causal identification, longitudinal validation, cross-cultural generalization, consumer welfare, and the study of consumption in the Artificial Era.

Keywords: brand expectations; consumer satisfaction; expectation–disconfirmation; customer loyalty; attitudinal loyalty; behavioral loyalty; brand knowledge; brand experience; brand trust; brand affect; repeat purchase; customer retention; switching behavior; consumer decision-making; consumer welfare; Artificial Era

Scientific Overview

Brand expectations, consumption satisfaction, and loyalty form a temporally organized field of consumer psychology. A brand provides learned information that affects what a person anticipates before acquisition, the attributes attended to during use, the interpretation of the experience afterward, and the likelihood of returning to the same offering. Satisfaction is a retrospective evaluation of an encounter or accumulated relationship. Loyalty concerns a pattern of favorable commitment, preference, or patronage that persists across occasions. These processes interact through experience, but each has distinct determinants and observable consequences. The research object of this monograph is the organization of this expectation–experience–evaluation–choice sequence and the evidence for its causal and predictive links. The foundational expectation–confirmation tradition offers one account of these transitions, while research on brand knowledge, affect, relationships, and behavioral retention identifies mechanisms that the basic comparison model leaves unexplained Oliver (1980) Keller (1993) Dick & Basu (1994).

The subject is consequential because a favorable experience does not guarantee a subsequent purchase, and repeated purchase does not by itself establish a favorable psychological bond. Consumers may return because of distribution, contracts, habit, low search costs, convenience, switching penalties, or restricted alternatives. Conversely, individuals can express attachment while facing financial, geographical, or situational barriers to buying a preferred brand. Empirical work therefore separates satisfaction ratings, repurchase intentions, actual transactions, attitudinal preference, and relationship duration. A rigorous account requires outcomes observed over time and methods capable of distinguishing correlation from causal influence Mittal & Kamakura (2001) Bolton (1998) Seiders et al. (2005).

This monograph follows an integrative narrative-review method. It critically examines foundational experiments, conceptual theories, longitudinal studies, field studies, measurement research, meta-analytic evidence, and research on sensory judgments and digital continuance. The synthesis organizes claims by research design and construct rather than treating every cited finding as equally conclusive. Historical theoretical works are evaluated for their continuing explanatory value, whereas contemporary empirical generalizations are assessed against measurement limitations, selection, contextual moderators, and the difference between stated and revealed behavior. The analysis concerns consumer experiences and brand-level expectations across products and services, while detailed numerical price perception and reference-price computation remain a separate neighboring research object Szymanski & Henard (2001) Khamitov et al. (2024).

Research Object, Boundaries, and Scientific Questions

A brand is treated here as an identifiable source of learned associations, expectations, and relational meanings that can influence consumer evaluation. Brand expectations are beliefs or anticipatory standards concerning likely performance, experiential qualities, reliability, treatment, or value. Consumption satisfaction is an evaluative response to a particular consumption episode or to an accumulated sequence of episodes. Loyalty is investigated through both relative psychological preference and behavior repeated under meaningful choice conditions. These definitions demand a specified referent: a respondent may be satisfied with an individual transaction, a service employee, a product category, or a continuing provider relationship. Responses to those different objects are not interchangeable Keller (1993) Oliver (1980) Dick & Basu (1994).

The central research questions are how learned brand associations shape expectations; when expectations alter perception rather than merely serving as comparison standards; how experienced performance and disconfirmation contribute to satisfaction; how affect, trust, commitment, and habit enter the transition from satisfaction to loyalty; and which designs most credibly identify those pathways. Additional questions concern why consumers switch after satisfactory experiences, how brand beliefs are updated following repeated encounters, and how emerging digital intermediaries change the information available before and after consumption. These questions require specifying temporal order, separating the formation of an expectation from its later evaluation, and measuring alternatives that consumers could actually have chosen Boulding et al. (1993) Gustafsson et al. (2005) Lemon & Verhoef (2016).

The domain overlaps with cognitive psychology in prediction, memory, expectancy and selective attention, and with social psychology in trust, commitment, identity, and reputational inference. It also draws on learning, behavioral decision research, and the psychology of culture. Related inquiries into advertising claims and packaging interpretation concern how particular communications are decoded; the present investigation asks what those communications accomplish within the longer postconsumption process. Analyses of the extended self explain some symbolic attachment, but the focal outcome here is how that meaning contributes to satisfaction, continued patronage, or switching. These distinctions maintain a single research identity for a multidimensional phenomenon Arnould & Thompson (2005) Fournier (1998).

Historical Foundations of Consumer Evaluation Research

Modern experimental attention to brand-mediated perception predates the consolidated satisfaction literature. Allison and Uhl studied whether beer brand identification influenced taste judgments, establishing an influential early paradigm for separating information about a product from judgments made in its presence. The historical importance of this design lies in its identification problem: a consumer's report can reflect intrinsic sensory properties, contextual labels, previously learned associations, and the interaction of these causes. The experiment provided an early empirical foundation for treating brand identity as an input into experience instead of a mere label applied afterward Allison & Uhl (1964).

Research in the 1970s and early 1980s developed explicit expectancy-disconfirmation explanations. Anderson examined how disconfirmed expectations influenced perceived product performance; Oliver formulated a cognitive model linking expectations and disconfirmation to satisfaction judgments; Churchill and Surprenant then investigated the relative explanatory roles of expectation and product performance under differing product conditions. Tse and Wilton subsequently compared models of satisfaction formation, testing whether perceived performance contributes directly alongside confirmation. The resulting tradition made comparison standards empirically discussable while also revealing how strongly effects can depend on the kind of product, the manipulation, and the outcome measure Anderson (1973) Oliver (1980) Churchill & Surprenant (1982) Tse & Wilton (1988).

The research program later expanded in three directions. Service-quality investigations considered expectations as differentiated standards of service delivery, including what customers believe a provider will do and what it should do. National satisfaction indices sought cumulative, comparable indicators across organizations and industries. Relationship and loyalty scholarship moved toward distinct constructs of trust, commitment, affect, purchase frequency, and durability of preference. These advances made it possible to investigate a sequence linking experiences to consumer relationships and firm outcomes without equating the stages Grönroos (1984) Boulding et al. (1993) Fornell (1992) Fornell et al. (1996) Dick & Basu (1994).

Brand Knowledge, Schemas, and Anticipatory Prediction

Brand knowledge is a learned structure of associations rather than a single rating. A consumer may associate a name with expected durability, sensory character, social meaning, customer service, environmental conduct, or past satisfaction. Keller's customer-based brand equity framework places brand awareness and the strength, favorability, and uniqueness of associations at the center of differential consumer responses. For the current research object, its significance is that expectations can be carried by memory structures that influence both the selection of information and interpretation of later evidence. The same physical product can therefore be evaluated within different anticipatory frames Keller (1993).

An expectation has at least three identifiable aspects: its content, confidence, and reference. Content specifies what is anticipated, such as taste intensity or delivery reliability. Confidence concerns the degree of certainty attached to that anticipation. Reference indicates whether it describes a particular encounter, a product class, or an enduring brand relationship. Prior use, word of mouth, marketing information, and contextual cues jointly influence these aspects, while subsequent feedback can selectively revise them. Brand promises become psychologically consequential when their implications are concrete enough to be confronted with experience; vague prestige or familiarity may be powerful but more difficult to test through simple performance comparisons Oliver (1980) Boulding et al. (1993).

Research should distinguish expectations that actually precede consumption from judgments reconstructed after a respondent has already learned the outcome. Retrospectively reported expectation is vulnerable to consistency pressure and memory reconstruction: people may report that they had always expected what subsequently occurred. Pre-use elicitation, timestamped exposure to a brand cue, and randomized information can make the anticipatory phase observable. A design that samples only postpurchase evaluations can estimate associations between brand image and satisfaction, but cannot by itself demonstrate that a prior expectation caused the reported experience. This distinction becomes especially important when brand familiarity, purchasing history, and consumer preference jointly determine who enters a sample Anderson (1973) Lee et al. (2006) Plassmann et al. (2008).

Types of Expectations and Comparison Standards

Consumer research has not always used the word expectation in one sense. Predictive or will expectations concern the level of performance a customer anticipates receiving. Normative or should expectations concern what an organization ought to deliver. Desired, adequate, ideal, and minimum tolerable performance can supply additional standards, and these may diverge within one individual. Boulding and colleagues developed a dynamic account in which will and should expectations can have different effects on perceived service quality and behavioral intention. Collapsing them into a single score obscures whether a person is forecasting the likely outcome or making a claim about acceptable treatment Boulding et al. (1993) Parasuraman et al. (1985).

The reference point matters for disconfirmation. Performance that exceeds a modest prediction may yield favorable surprise while remaining below a consumer's preferred standard. A high-reputation brand can simultaneously raise expected quality, increase willingness to try the product, and expose itself to harsher evaluation if the experience fails to meet the elevated standard. Thus, an apparently adverse relationship between high expectations and later satisfaction may reflect stricter comparison, selection, regression to the mean, or actual differences in delivery. Establishing the mechanism requires measuring the specific standard, assessing its timing, and controlling or manipulating the experienced outcome Anderson (1973) Churchill & Surprenant (1982) Tse & Wilton (1988).

Service settings add the problem of multiple layers of delivery. Consumers may distinguish functional outcomes, interactional treatment, access, timeliness, and the reliability of future responses. Grönroos's service-quality model and subsequent gap-based models brought these dimensions to the foreground. Their historical role is important even though conceptual service-quality gaps and a consumer's overall satisfaction are not identical measurements. Evaluation of service quality may persist as a relatively stable belief, while satisfaction may arise from a particular encounter or a cumulative relationship. Researchers must define whether the intended endpoint is performance estimation, perceived service quality, satisfaction, or loyalty Grönroos (1984) Parasuraman et al. (1985) Zeithaml et al. (1996).

The Consumption Journey as a Temporal Process

Consumers encounter a brand over a sequence of informational and experiential episodes: recognition, search, comparison, purchase, use, support, reflection, sharing, and possible repurchase. These stages may recur, overlap, or involve different channels and organizational actors. The customer-journey framework directs attention to touchpoints before, during, and after a focal purchase, including interactions that a provider does not fully control. For psychological research, the framework is valuable when a stage has a measurable effect on the formation or revision of expectation rather than serving only as a descriptive map Lemon & Verhoef (2016).

The same customer can report distinct satisfaction levels across a journey. A convenient purchase may be followed by disappointing product use; an initial service failure may be followed by effective recovery; repeated adequate interactions may create stability without enthusiasm. Cumulative satisfaction therefore depends on how episodes are weighted and remembered. Frequency, recency, perceived severity, attribution of responsibility, and the relation to prior expectations provide plausible mechanisms. Their relative influence should be estimated with measurements repeated over time rather than assumed from a single end-of-journey survey Bolton (1998) Gustafsson et al. (2005).

A longitudinal process model also requires clear ordering among brand expectations, perceived quality, affect, satisfaction, trust, intention, and actual behavior. One plausible path runs from pre-encounter brand knowledge to anticipated performance, then to encountered performance and disconfirmation, followed by satisfaction and changes in preference. But purchases also generate new experience and new opportunities for expectation revision; stronger loyalty can alter the amount of evidence an individual encounters. The system consequently contains feedback loops that are lost in a one-direction cross-sectional regression. A scientific synthesis must describe the proposed temporal relation and the observations capable of testing it Oliver (1980) Johnson et al. (2006) Lemon & Verhoef (2016).

Expectation–Disconfirmation and Satisfaction Formation

Expectation–disconfirmation theory describes satisfaction as an evaluation shaped by the relation between anticipated and perceived performance. Positive disconfirmation occurs when the experience exceeds the relevant standard; negative disconfirmation occurs when it falls short; confirmation describes correspondence between the two. Oliver's early cognitive model linked this comparison to satisfaction decisions, and subsequent investigations showed that perceived performance may contribute directly rather than acting only through the expectation difference. The framework therefore identifies a possible process rather than a universal arithmetic rule by which satisfaction is mechanically computed Oliver (1980) Churchill & Surprenant (1982) Tse & Wilton (1988).

A simple discrepancy model treats disconfirmation as perceived performance minus expectation, but the two components are rarely interchangeable units. One consumer may report numerical expectations about delivery time, another a general belief about service quality, and a third a socially learned feeling of what a reputable brand should provide. Difference scores can have low reliability and make it difficult to distinguish high-expectation/high-performance situations from low-expectation/low-performance situations that happen to produce the same numerical gap. More informative analyses measure expectations and experienced performance separately, specify the comparison standard, and estimate their independent and interactive relations to satisfaction Anderson (1973) Tse & Wilton (1988).

The theory also suggests competing paths. A brand promise can raise expected performance, which may improve anticipated value and encourage initial trial. The same increase can make equivalent performance seem less satisfactory after consumption if the expectation serves as a strict benchmark. Alternatively, strong expectations can shape perceived performance upward through assimilation. A design that examines only satisfaction after consumers have self-selected a brand cannot distinguish these opposing pathways. Randomized information and direct assessments of perceived performance help reveal whether an effect is perceptual, comparative, or primarily a consequence of sorting consumers into different experiences Churchill & Surprenant (1982) Boulding et al. (1993) Lee et al. (2006).

Perceived Performance, Quality, and Multiple Evaluation Standards

Experienced product performance is a multidimensional psychological judgment built from observable features and interpreted consequences. Reliability, physical durability, sensory character, service responsiveness, accessibility, and fit with a person's goal may each contribute, depending on category. Perceived quality represents an assessment of attributes or excellence; satisfaction represents an evaluative response to consumption; value assessments incorporate what consumers believe they have given up. These constructs often correlate but are not equivalent. In particular, satisfaction with a particular event can change rapidly while a general brand-quality belief remains comparatively stable Parasuraman et al. (1985) Grönroos (1984) Fornell et al. (1996).

Churchill and Surprenant investigated conditions under which expectation and performance contribute differently to satisfaction. Tse and Wilton extended the models by testing alternatives in which perceived performance has a direct influence. These works remain important because they caution against fitting the same expectation-minus-performance specification to every purchase context. A familiar nondurable product, an unfamiliar durable product, and a continuous service relationship afford different opportunities to inspect performance, compare standards, and update learned expectations. The structure of a satisfaction model therefore belongs to an explicit theory of the consumption episode and the evidence available to the consumer Churchill & Surprenant (1982) Tse & Wilton (1988).

Multiple standards also create apparent contradictions across studies. When a researcher asks whether a customer is satisfied, the respondent may compare the experience with the immediately preceding expectation, with competitors, with what was paid, or with an imagined ideal. Standardized response scales cannot eliminate variation in the reference point. Better research records what the question asks respondents to compare, distinguishes transactional from cumulative objects of judgment, and tests whether conclusions persist across measurement instruments. These steps are prerequisites for interpreting national satisfaction indices, retailing surveys, and cross-brand comparisons as evidence about comparable psychological processes Fornell (1992) Szymanski & Henard (2001).

Assimilation, Contrast, and the Timing of Brand Information

A brand cue can alter experience through mechanisms beyond comparison after the fact. Assimilation describes judgments drawn toward a previously activated expectation; contrast describes judgments moving away from an expectation when inconsistency is salient. Their relative occurrence depends on the strength, plausibility, and timing of the information and on the sensory and evaluative task. The significance is that a consumer's report of perceived performance may itself be affected by brand information, thereby complicating any attempt to regard reported performance as an independent physical input Anderson (1973) Allison & Uhl (1964).

The beer experiments of Lee, Frederick, and Ariely manipulated whether consumers learned of an ingredient before or after tasting. Preference differed with the timing of disclosure, an especially useful result because it separated the prospect of consumption from the retrospective receipt of information. The study illustrates why apparently identical disclosures can have distinct psychological effects depending on whether they frame experience prospectively or supply an explanation afterward. As in other laboratory settings, extension to routine buying requires consideration of product familiarity, the realism of the context, and subsequent behavior rather than relying on preference reports alone Lee et al. (2006).

Plassmann and colleagues studied price information attached to wine under experimental conditions and examined reported pleasantness alongside neural responses associated with experienced reward. The experiment supports the possibility that marketing information influences the experience of a sensory product, including a neural correlate of reported pleasantness. It does not establish that every increase in price or status changes basic sensory processing, nor does it identify brand loyalty from a single tasting response. Within this dossier, the contribution concerns expectation-induced valuation; the calculation of specific reference prices remains the subject of a separate scientific inquiry Plassmann et al. (2008).

Affect, Attribution, and the Emotional Texture of Satisfaction

Satisfaction is commonly modeled as a judgment, yet its phenomenology and consequences also involve affect. Consumers can be pleased, relieved, disappointed, irritated, or ambivalent following an experience, and these reactions need not map perfectly onto a single evaluative score. An adequate explanation therefore permits affect to arise from the experience, from the comparison with expectations, and from the meaning attached to the provider's actions. Emotional response can influence later memory and expression of intention, while the stability and generality of these effects remain design-dependent questions Oliver (1980) Chaudhuri & Holbrook (2001).

Attribution matters especially when performance varies. A consumer who believes an unexpected delay was caused by an uncontrollable event may update expectations differently from one who interprets the same delay as evidence of organizational indifference. Fairness, perceived responsibility, and the appropriateness of remedy can mediate the response. Although evaluative and affective processes frequently co-occur, the causal path from incident appraisal to trust erosion or renewed loyalty must be tested rather than presumed. Research on service quality and relationship commitment provides suitable conceptual dimensions, but many of the strongest observational associations cannot separate the initial incident from the provider's subsequent response Garbarino & Johnson (1999) Gustafsson et al. (2005).

Brand affect is a related but temporally broader construct: consumers can develop positive feelings toward a brand through repeated encounters and associations extending beyond a recent transaction. Chaudhuri and Holbrook investigated links from brand trust and brand affect to purchase and attitudinal loyalty across brands, emphasizing that these forms of loyalty connect differently with performance outcomes. Brand liking may coexist with frustration about a purchase, and short-term satisfaction may occur without deep affective attachment. Measurement should preserve this distinction rather than place affective commitment inside the definition of satisfaction Chaudhuri & Holbrook (2001).

Blinded and Branded Sensory Evaluation

Blind-versus-branded tests offer a direct strategy for examining expectation effects. A researcher can present identical or carefully matched products with different brand cues, randomize exposure, measure expectation before consumption, and compare subsequent ratings. If the information changes while the product remains constant, resulting differences in evaluation provide evidence that context matters. Allison and Uhl's historical beer study and later experiments with ingredient disclosure and wine marketing exemplify the approach. Each identifies a particular effect under specified task conditions, rather than demonstrating universal consumer susceptibility to branding Allison & Uhl (1964) Lee et al. (2006) Plassmann et al. (2008).

The experimental details determine interpretation. Blinding must actually conceal relevant product identity; order and carryover effects must be counterbalanced; sensory fatigue and experimenter demand should be considered; and a manipulation check should verify that participants perceived the brand information. Product category expertise, familiarity, and genuine hedonic preference can moderate outcomes. Whenever possible, measurements should distinguish anticipated liking, actual choice in a consequential setting, immediate pleasantness, delayed recollection, and willingness to obtain the product again. Those observations test different steps in the consumption sequence Lee et al. (2006) Plassmann et al. (2008).

A sensory manipulation may produce a statistically detectable difference in reported pleasantness while leaving later patronage unchanged. The distinction prevents a category error in which an isolated laboratory rating is interpreted as evidence of loyalty. The most valuable extension couples controlled manipulation with delayed behavior, repeated exposure, and feasible alternatives. Such designs also help test whether an expectancy effect declines with product knowledge or strengthens through reinforcement. Where repeated consumption is impractical in the laboratory, linked field data can evaluate the external validity of the immediate judgment Mittal & Kamakura (2001) Seiders et al. (2005).

Brand Experience, Salience, and Consumer Memory

Brand experience refers to sensory, affective, intellectual, and behavioral responses evoked by brand-related stimuli. Brakus, Schmitt, and Zarantonello developed and tested a measure distinguishing these dimensions and connected them to satisfaction and loyalty. Their work expands analysis beyond performance confirmation by emphasizing the organization of contact with branded objects, environments, and communications. A conspicuous design feature, a distinctive service interaction, or a demanding cognitive task can become part of brand experience even when the consumer is not explicitly evaluating product reliability Brakus et al. (2009).

Memory creates continuity across transactions. Consumers encounter selective samples of events, store experiences with varying salience, and reconstruct an overall impression when later asked for satisfaction or preference. Cumulative evaluations may disproportionately reflect especially consequential episodes, recent interactions, or a remembered trajectory. The empirical challenge is that remembered quality can guide future choice while simultaneously being altered by later choice and rationalization. Longitudinal measurement of episode-level reports, combined with transaction records, helps distinguish change in experience from change in retrospective narrative Bolton (1998) Johnson et al. (2006).

Experience measures require attention to content validity. A scale designed for stimulating retail environments may not represent the relevant dimensions of a utility service or low-involvement commodity. Likewise, repeated exposure may heighten familiarity without producing enthusiasm. Researchers can test whether the established multidimensional structure replicates across sectors, cultures, ages, and interfaces before using a composite score as an explanatory variable. Theoretical claims about brand experience should be anchored to the particular stimuli and behaviors actually measured Brakus et al. (2009) Lemon & Verhoef (2016).

Cumulative Satisfaction and the Updating of Expectations

Transaction-specific satisfaction concerns an evaluative episode; cumulative satisfaction concerns a relationship or extended record of use. The difference matters for prediction. A single severe failure can matter differently for a long-standing relationship than for a new consumer with no history of previous good performance. National satisfaction indices and later retention studies generally investigate broader evaluations, while laboratory disconfirmation research often manipulates one short encounter. The resulting measures answer different questions and should not be treated as substitutes Fornell (1992) Fornell et al. (1996) Bolton (1998).

A dynamic account permits prior satisfaction to change the next expectation, and each new outcome to revise both. Bolton's model of relationship duration with a continuous service provider, for example, treated satisfaction as a time-relevant input to retention. Johnson, Herrmann, and Huber examined changing drivers of loyalty intentions, suggesting that the importance of perceived value, satisfaction, and affect may depend on the relationship's stage. These findings motivate models that accommodate a customer history rather than assume the psychological weight of each encounter is constant Bolton (1998) Johnson et al. (2006).

Cumulative change can be assessed with repeated ratings, experience sampling, periodic customer panels, and observed purchase or contract histories. The analyst must still consider attrition: customers who leave may disappear from the satisfaction sample, falsely improving the average rating of those who remain. It is also necessary to compare cohorts at similar relationship stages and observe changes in product or service quality that occurred simultaneously. Without these checks, a positive satisfaction trend may describe sample replacement or survivorship rather than improvement in consumer experience Gustafsson et al. (2005) Seiders et al. (2005).

Loyalty as Attitude, Behavior, and Temporal Persistence

Consumer loyalty contains evaluative and behavioral dimensions. A favorable attitude can be expressed through preference, resistance to alternatives, recommendation, identification, and willingness to maintain a relationship. Behavioral loyalty is observed through repeated purchase, continued subscription, or a greater share of category purchases allocated to a provider. The two dimensions may align, but the alignment is an empirical issue. Dick and Basu proposed a framework distinguishing genuine, latent, spurious, and absent loyalty according to relative attitude and repeat patronage, drawing attention to situational conditions and social norms that mediate their connection Dick & Basu (1994).

Oliver conceptualized loyalty as developing through cognitive, affective, conative, and action-oriented phases. This stage account remains useful as a theoretical description of increasingly stable preference, but the proposed sequence should not be imposed on every consumer history. Some choices are habitual from the outset, some concern low-involvement products, and others depend strongly on institutional contracts or shared household decisions. A person may be loyal for a limited period and then switch when costs, access, or identity changes. Longitudinal observation is required to establish whether stage-like transitions occur in a population Oliver (1999) Gustafsson et al. (2005).

Repeated purchases under conditions of restricted alternatives are insufficient evidence of robust preference. An operational definition should specify the observation period, purchase opportunities, category need, and available alternatives. For a frequently purchased product, loyalty can be estimated from sequences of choices; for a durable good, intention, recommendation, retention of an ecosystem, and replacement decisions may be more realistic indicators. These measures should be interpreted separately because their opportunity structures differ, and because stated preference or recommendation is not equivalent to a consequential choice Dick & Basu (1994) Mittal & Kamakura (2001).

The Satisfaction–Loyalty Relationship and Its Boundary Conditions

Customer satisfaction predicts aspects of loyalty in many settings, yet the association is neither deterministic nor invariant. A satisfied buyer may have competing options that are equally attractive, while a dissatisfied buyer may stay because alternatives are expensive or inaccessible. Variation in market competition, relationship age, product involvement, habit, and the consequences of switching can change the behavioral expression of satisfaction. The most defensible account is conditional: satisfaction is one input into the probability of future choice, whose effect is mediated and moderated by other variables Bolton (1998) Mittal & Kamakura (2001) Seiders et al. (2005).

Mittal and Kamakura examined the relation between satisfaction, repurchase intent, and observed repurchase behavior, including variation across customer characteristics. Their findings are particularly important for separating what respondents say they will do from later behavior. Seiders and colleagues likewise studied moderating influences in a retailing setting, challenging a universal assumption that satisfied customers necessarily purchase more. Designs like these move research beyond identifying a correlation and toward specifying when, for whom, and through which conditions satisfaction contributes to repeat buying Mittal & Kamakura (2001) Seiders et al. (2005).

A nonlinearity hypothesis holds that the marginal effect of satisfaction may change across its range. Very dissatisfied consumers may leave quickly when exit is easy; moderately satisfied consumers may remain because the offering is acceptable; very satisfied consumers may show a stronger preference under high-stakes comparisons. But apparent threshold patterns can emerge from ceiling effects in rating scales, heterogeneous opportunities to purchase, or a censored sample containing mostly survivors. Researchers should use flexible response models and validate thresholds in independent data rather than presume a universal satisfaction score at which loyalty begins Fornell et al. (1996) Gustafsson et al. (2005).

Brand Trust, Affect, and Relationship Commitment

Trust concerns a consumer's willingness to rely on a provider under conditions of uncertainty, including beliefs about competence, integrity, or benevolence. Brand affect refers to favorable or unfavorable emotional reactions associated with a brand. Commitment concerns motivation to maintain a relationship and can reflect emotional identification, calculated benefits, or the perceived costs of leaving. These constructs can accompany satisfaction but have different temporal and causal implications. A customer may trust a brand that has delivered consistently without feeling strong affection, or may feel attached to a brand while distrusting one product claim Chaudhuri & Holbrook (2001) Garbarino & Johnson (1999).

Chaudhuri and Holbrook examined a chain linking brand trust and brand affect to purchase and attitudinal loyalty and, in turn, aspects of brand performance. Garbarino and Johnson found that satisfaction, trust, and commitment could play different roles across customer relationship segments. The studies support multi-construct explanations: satisfaction may matter particularly for relatively transactional encounters, while trust and commitment may become more prominent in sustained relationships. The evidence does not license an invariant causal hierarchy across all categories because much of the reported structure depends on observational measurement and theory-guided models Chaudhuri & Holbrook (2001) Garbarino & Johnson (1999).

A later meta-analysis by Khamitov and colleagues integrated research on consumer trust across numerous contexts and reported that different antecedents and outcomes have meaningfully different associations with trust. Its comparative findings are relevant to a dossier focused on loyalty because they underscore the distinction between attitudinal outcomes and actual behavior. Meta-analytic associations summarize patterns across studies, however, and their causal interpretation depends on the designs of the constituent evidence. Direct observation of purchases, interventions affecting trust-relevant information, and repeated measures remain necessary to test how changes in trust alter consumer behavior Khamitov et al. (2024).

Brand Relationships, Identity, and Symbolic Meaning

Brand attachment and identity introduce meanings extending beyond the immediate product outcome. Consumers may use a brand to express group affiliation, aspiration, continuity, or a conception of self. Fournier's relationship-theory research documented varieties of consumer–brand relationships through qualitative analysis, while consumer culture theory investigated the cultural patterns through which consumption acquires shared significance. These approaches identify mechanisms that conventional performance models may overlook: a brand can become part of a narrative about personal history or social membership Fournier (1998) Arnould & Thompson (2005).

Symbolic attachment can sustain preference during minor setbacks, but its consequences are not uniform. A brand that becomes closely associated with a consumer's identity may be especially valued for consistency with a social role; the same identification can intensify disappointment when conduct appears inconsistent with that meaning. The relevant hypotheses concern the strength, centrality, and social context of identification as well as the perceived significance of failure. General claims about an automatic loyalty advantage for identity-laden brands exceed what a small qualitative sample can establish Fournier (1998) Arnould & Thompson (2005).

Different methodological traditions answer different questions in this area. Ethnography and interviews illuminate motives, narratives, and social practices. Surveys can test the distribution of proposed constructs; experiments can manipulate brand–identity congruence; transaction panels can test whether stated meaning persists in actual choices. A sound synthesis values these methods according to what they can identify, avoiding the error of using qualitative insight as a population prevalence estimate or a cross-sectional association as evidence of causal attachment Fournier (1998) Dick & Basu (1994).

Habit, Convenience, and Switching Costs

Habit is a learned response tendency organized by recurring contexts and reinforced by repeated action. In consumer behavior, established routines can preserve a brand choice even when the individual has little active emotional commitment. Convenience and local availability can have similar effects without requiring a strong psychological bond. Switching costs may include time, learning a new interface, transferring records, losing accumulated benefits, uncertainty about performance, and contractual penalties. A researcher who identifies all repeated purchases as loyalty cannot determine which of these mechanisms actually preserves patronage Dick & Basu (1994) Gustafsson et al. (2005).

Retention studies help separate forms of persistence by considering triggers and relationship commitment. Gustafsson, Johnson, and Roos investigated satisfaction, commitment dimensions, and triggers of retention in a longitudinal setting, emphasizing that switching is responsive to multiple processes. Bolton modeled relationship duration in services, where a customer's opportunity to leave can be observed across time. Together these designs show why retention is best understood through hazard, history, and context rather than one static satisfaction indicator Gustafsson et al. (2005) Bolton (1998).

A discriminating study can exploit natural changes in access, pricing rules, default settings, or switching friction while preserving appropriate ethical safeguards. If a previously stable purchase pattern weakens immediately when a channel becomes unavailable, convenience may have contributed substantially to that pattern. If a preference survives comparable alternatives and reduced exit barriers, stronger loyalty becomes plausible. Such inferences still require attention to concurrent shocks and selection, but they provide a much firmer bridge between psychological construct and behavior than a loyalty label assigned from purchase frequency alone Dick & Basu (1994) Mittal & Kamakura (2001).

Repurchase Intentions, Recommendations, and Observed Choice

Intentions are forecasts of behavior expressed under a particular information state. They can be useful when a future purchase is rare or costly to observe, but they are influenced by social desirability, scale interpretation, changing circumstances, and the respondent's mistaken predictions about future preferences. Observed transactions reflect opportunities and constraints as well as motives. The relation between the two is therefore a substantive object of inquiry. Mittal and Kamakura's comparison of intent and observed repurchase remains a central illustration of this distinction Mittal & Kamakura (2001).

Recommendation intentions add another complication. Saying that one would recommend a brand can express favorable evaluation, perceived social responsibility, identification, or a desire to appear helpful; it may not indicate actual advocacy or subsequent retention. Research testing whether simple recommendation metrics forecast business growth has raised questions about relying on a single item in place of a richer explanatory model. The scientific requirement is to state which behavior is being predicted and compare candidate metrics against an appropriate criterion, preferably outside the sample used to build the prediction Keiningham et al. (2007).

A strong measurement design records the time frame and denominator of purchase opportunities, includes competing alternatives, and distinguishes a consumer's own spending from referrals or public statements. Behavioral outcomes may include time to churn, frequency, share of wallet, brand switching, repeat subscription, or verified referral. Their meanings depend on product category and access to data. Where only intentions are observed, the conclusion should remain about intentions rather than be silently expanded into claims of actual loyalty behavior Seiders et al. (2005) Zeithaml et al. (1996).

Service Failures, Recovery, and Relationship Reassessment

A failure can simultaneously change experienced performance, the meaning of earlier expectations, and beliefs about the provider's future conduct. Recovery efforts—acknowledgment, explanation, compensation, correction, and changes to delivery—may alter each of these pathways. A failure followed by effective recovery may yield satisfaction with the response while the original incident remains negatively evaluated. Research should therefore separate incident satisfaction, recovery satisfaction, trust in the provider, and later retention rather than infer one comprehensive response from a single question Parasuraman et al. (1985) Garbarino & Johnson (1999).

The customer's history and available alternatives matter after a failure. A new purchaser lacks evidence of a consistent relationship and may respond differently from a long-term customer who interprets the episode against years of adequate service. Yet a long history may also create higher expectations of fairness and responsiveness, making a perceived breach more consequential. Dynamic models of retention and commitment can accommodate these alternatives, while cross-sectional accounts that interview only current customers systematically miss those who already exited Bolton (1998) Gustafsson et al. (2005).

The methodological priority is to identify recovery effects without confounding intervention with severity. Providers often offer greater compensation after particularly serious incidents; an observational association between compensation and later dissatisfaction may therefore reflect underlying damage rather than compensation's harmful effect. Randomized, ethically acceptable variations in communication or remedy, matched comparisons, and rich longitudinal event histories can improve inference. Claims that recovery universally increases loyalty beyond the pre-failure baseline should be treated as hypotheses requiring explicit evidence rather than as an automatic marketing law Gustafsson et al. (2005) Zeithaml et al. (1996).

From Psychological Evaluation to Retention and Market Outcomes

A scientific account extends beyond private evaluations to their behavioral consequences while preserving distinct levels of analysis. Individual satisfaction may influence retention, repeat buying, word of mouth, and willingness to pay. Aggregated satisfaction indices may covary with market share and firm outcomes, but firm-level correlations cannot establish that changing a particular customer's rating causes proportional profit growth. Anderson, Fornell, and Lehmann studied these links using Swedish market data; the research is influential precisely because it articulated the organizational importance of an often individually measured construct Anderson et al. (1994).

Homburg, Koschate, and Hoyer investigated satisfaction and willingness to pay experimentally, contributing evidence that the economic consequences of accumulated satisfaction deserve direct measurement. Yet willingness to pay is not the same as actual payment, and experimentally elicited valuations may be sensitive to procedure and incentive structure. Similar caution applies when sales or revenue are interpreted as loyalty: market growth can arise from new acquisition, channel expansion, and price changes without increased attachment among existing purchasers Homburg et al. (2005) Anderson et al. (1994).

The most credible pathway from psychological mechanism to organizational consequence uses linked data across levels. Individual expectations and satisfaction must be measured before behavioral outcomes, actual purchasing or retention must be recorded where feasible, and contextual factors such as competition, availability, and advertising must be modeled. Aggregate indices can inform policy and industry comparison, but their construction, respondent composition, and temporal stability must be explicit. This integrated strategy respects both the psychological mechanisms and the economic systems through which they become visible Fornell et al. (1996) Seiders et al. (2005).

Scientific Designs and the Logic of Causal Identification

The central causal question asks whether changing a brand-related expectation, holding relevant aspects of the product and setting constant, alters later experience, satisfaction, or choice. Different methods observe different segments of this chain. Randomized cue experiments can isolate information effects on immediate judgments; field experiments can estimate effects in real encounters; observational customer panels can characterize trajectories; transaction records reveal retained behavior; interviews can illuminate meanings that surveys fail to anticipate. An integrative monograph should connect each claim with the design that supports it and avoid placing explanatory weight on a method incapable of identifying the proposed cause Allison & Uhl (1964) Lee et al. (2006) Bolton (1998).

Causal diagrams can clarify competing explanations. A consumer's prior familiarity may influence both the brand selected and the expectation reported; product performance may influence satisfaction and be influenced by the quality tier a consumer selected; commitment may influence the decision to stay and consequently the opportunity to report future satisfaction. When these variables are jointly determined, adjusting for a large number of measured predictors can still leave selection bias. More seriously, controlling for a mediator such as perceived performance when estimating the total effect of brand information can remove the very mechanism of interest. Studies should distinguish total, direct, and mediated effects in advance Boulding et al. (1993) Plassmann et al. (2008).

Triangulation gains scientific force when evidence sources have meaningfully different limitations. A branded sensory test can identify a short-term expectancy effect but has restricted ecological scope; a field panel observes realistic choices but often cannot separate expectations from unmeasured consumer characteristics; a qualitative account explains what brand relations mean but does not estimate their prevalence. Agreement across such methods strengthens an explanatory interpretation without supplying mathematical proof of generality. Disagreement should prompt analysis of time scale, context, measurement, and the behavior under consideration Fournier (1998) Mittal & Kamakura (2001) Szymanski & Henard (2001).

Experimental Manipulations and Measurement Integrity

An expectation experiment can randomize brand labels, quality claims, independent recommendations, or prior performance descriptions before consumers encounter a standardized product or service. The design should record baseline preferences, expectation content, prior familiarity, and the participant's awareness of the manipulation. It should also measure performance judgments and satisfaction separately. In an ideally blinded comparison, differences in product input are minimized so that a difference in reported experience is more readily attributable to contextual information. The classic branded-taste paradigm and the timing manipulation in beer evaluation show complementary ways to implement this logic Allison & Uhl (1964) Lee et al. (2006).

Manipulation checks should establish what information participants actually interpreted, while avoiding checks that themselves prime the intended mechanism. Concealment, random assignment, counterbalancing, and a clear pre-registration of primary outcomes protect against alternate interpretations. Where participants evaluate multiple products, investigators should consider anchoring, contrast, fatigue, and interference. In sensory neuroimaging, the relation between a neural signal and subjective experience requires particular care: a signal correlated with pleasantness is not an independent measurement of loyalty, and neural differences do not settle whether effects persist into later consumer decisions Plassmann et al. (2008).

A complete experiment reports recruitment, inclusion rules, exclusions, assigned conditions, outcome timing, uncertainty estimates, and relevant preregistered analyses. When feasible, a behavioral follow-up is especially valuable: participants can be given a meaningful subsequent choice or researchers can observe repurchase after an interval. Such extensions can determine whether changes in immediate evaluation survive outside the task. For replication, the identity of the product, the exact cue, baseline brand familiarity, incentives, and cultural setting are not minor logistical details; they help define which psychological effect the study has actually tested Lee et al. (2006) Brakus et al. (2009).

Measurement of Satisfaction, Trust, and Loyalty

Satisfaction instruments may ask about a transaction, recent experience, overall satisfaction, performance relative to expectations, or perceived value. A single response scale can conceal major differences in construct referent and recall period. A psychometrically informative evaluation specifies the construct, separates facets where theory requires it, and tests reliability, factor structure, predictive validity, and responsiveness to change. The American Customer Satisfaction Index established one influential approach to aggregate measurement, but comparability depends on the assumptions and procedures of the index rather than simply on using its familiar name Fornell et al. (1996).

Brand experience, trust, commitment, satisfaction, and loyalty must not become indistinguishable through redundant questionnaire wording. Brakus and colleagues developed a differentiated brand-experience scale; Chaudhuri and Holbrook separated trust and brand affect from purchase and attitudinal loyalty; Garbarino and Johnson analyzed relationship constructs that have different roles among relationship segments. Together these programs offer a reason to test discriminant validity and incremental prediction rather than relying on a high internal-consistency coefficient for an indiscriminate collection of favorable statements Brakus et al. (2009) Chaudhuri & Holbrook (2001) Garbarino & Johnson (1999).

Survey interpretation also requires identifying selection and response processes. Highly satisfied customers may be more willing to complete a survey, and dissatisfied customers may answer primarily following serious failures. Scale responses can differ by cultural norms, language, age, relationship length, and category involvement. Before comparing group means or pooled paths, researchers should evaluate measurement invariance and the possibility that a question carries different social meaning across groups. A measure can reliably reproduce response patterns while still inadequately representing the intended psychological construct Szymanski & Henard (2001) Khamitov et al. (2024).

Longitudinal Panels, Retention Models, and Repeated Measures

Longitudinal data allow a researcher to observe expectation revision, cumulative satisfaction, and subsequent exit in their temporal order. A panel can collect expectations before a transaction, satisfaction after use, and intention after evaluation, then link those observations to later purchases. Relationship duration models can represent whether and when a consumer leaves, while repeated-measures models can describe changes in a person's satisfaction over time. Bolton's duration analysis and Gustafsson and colleagues' retention research exemplify a longer temporal horizon than a cross-sectional survey can provide Bolton (1998) Gustafsson et al. (2005).

Longitudinal modeling is vulnerable to time-varying confounding. Dissatisfaction may precipitate a decline in engagement, which reduces the opportunity to encounter improved service; a competitor's entry may coincide with dissatisfaction; the organization may preferentially contact customers predicted to leave. Attrition is particularly problematic because a customer who switches may cease to appear in provider-administered measurement. Suitable designs record exit as an outcome, account for incomplete histories, and seek independent information on changes in alternatives, price, access, and service quality Bolton (1998) Mittal & Kamakura (2001).

Cross-lagged associations should be interpreted in relation to stable between-person differences and within-person fluctuations. A person who consistently values convenience may report both more satisfaction and more frequent repurchase without a rise in satisfaction causing a rise in repurchase. Repeated measurements can help distinguish persistent characteristics from episode-specific changes, but identification depends on model assumptions. Whenever possible, analytic plans should report sensitivity to alternative lag structures, model specifications, and the treatment of missing observations Johnson et al. (2006) Gustafsson et al. (2005).

Actual Purchase, Switching, and Market Data

Transaction data give consumer research access to behaviors unavailable in attitudinal surveys. Basket composition, timing, repeat orders, subscription renewal, share of category purchases, and switching after a competitor's entry can operationalize different dimensions of loyalty. These records still require careful interpretation. A transaction may reflect household choice rather than an individual preference, and a missing purchase may reflect absent need or temporary unavailability rather than defection. The analyst must define the risk set and observation window before treating retained purchase as a valid loyalty outcome Mittal & Kamakura (2001) Seiders et al. (2005).

Matching a survey to purchase records is powerful because it allows satisfaction and intention to be evaluated as predictors of later action. Yet matching can exclude customers without identifiable accounts, and a sample restricted to known loyal customers can underrepresent occasional purchasers. Retail studies of satisfaction and purchase quantity demonstrate the importance of heterogeneity in the relation between reported experience and expenditure. Predictive models should be validated on later time periods or separate customer cohorts and compared against simple behavioral baselines rather than assessed only on the data used for estimation Seiders et al. (2005).

Organizational data introduce an ecological level that can diverge from individual psychology. A retailer's revenue can rise because more consumers enter a market, because prices increase, or because the product mix changes. Cross-firm associations between satisfaction indices and revenue thus contain economic information without directly identifying individual motivational transitions. Studies should align the unit of analysis, distinguish acquisition from retention, and avoid ecological inference when moving between firm-level and consumer-level findings Anderson et al. (1994) Fornell (1992) Keiningham et al. (2007).

Review Evidence, Publication Bias, and Replicability

Evidence synthesis must begin by distinguishing review method. A narrative synthesis organizes theories and observations; a systematic review requires a stated question, reproducible search and eligibility procedure, screening and extraction record, and structured risk-of-bias assessment; a meta-analysis requires sufficiently comparable effect estimates and quantitative synthesis. The present work is a critical narrative synthesis because it does not claim exhaustive database screening or pooled estimation. Existing meta-analyses are used as separately conducted evidence summaries, not relabeled as results computed by this monograph Szymanski & Henard (2001) Khamitov et al. (2024).

Szymanski and Henard synthesized empirical research on customer satisfaction and highlighted variations in the evidence for candidate antecedents and consequences. Khamitov and colleagues' later trust meta-analysis aggregated a broad research literature across countries and examined differences among trust pathways. Such syntheses provide insight into recurring associations and moderators, but summary effects are constrained by the quality of included measurements, study designs, and publication selection. An aggregate correlation between trust and attitudinal loyalty cannot alone resolve whether increased trust will cause additional purchases after a real competitor enters the market Szymanski & Henard (2001) Khamitov et al. (2024).

Replication should reproduce theoretically significant features rather than merely repeat an analysis on a new convenience sample. For cue-induced sensory evaluation, that includes expectancy timing, realistic stimulus control, and participant familiarity. For satisfaction–retention associations, it includes observation of competitive alternatives, relationship age, and real behavior. Open instruments, documented exclusion rules, registered primary contrasts, reported uncertainty, and independent replications would improve confidence. Where such evidence is absent, the appropriate conclusion identifies an informative but still context-dependent association Lee et al. (2006) Mittal & Kamakura (2001).

Heterogeneity, Culture, and the Limits of Generalization

Consumers differ in the goals they bring to consumption, their prior experiences, information access, familiarity with brands, sensitivity to risk, and relationships to institutions. Category involvement and life circumstances influence what counts as satisfactory performance. Cultural meanings can turn the same brand sign into evidence of familiarity, aspiration, conformity, or exclusion. Consumer culture theory provides a conceptual reminder that evaluation takes place within social practices and historically developed meanings rather than in an isolated choice laboratory Arnould & Thompson (2005) Fournier (1998).

Cross-cultural evidence must address both explanatory generality and measurement comparability. A satisfaction question that assumes an individual purchase decision may poorly represent jointly managed household consumption. Mean comparisons can be misleading if response scales or expectations have different reference points across settings. The global breadth of trust research in Khamitov and colleagues' meta-analysis is valuable for examining heterogeneity, but coverage across countries does not remove the need to evaluate sampling and construct equivalence within each study. Generality is an empirical result established across meaningful variations in context Khamitov et al. (2024).

Digital access, socioeconomic conditions, and disability-related barriers can also shape the consumer's opportunity to choose. Retention under scarce alternatives may reflect constraints, while satisfaction surveys collected from connected users can omit those who lack access or have already switched. Accordingly, a defensible consumer-welfare analysis distinguishes an experience that consumers value from a relationship they are unable to leave. This distinction is essential when applying loyalty metrics to essential services, platform ecosystems, and products with substantial information or switching costs Dick & Basu (1994) Lemon & Verhoef (2016).

Digital Services and Expectation–Confirmation in Continued Use

Digital services make the satisfaction–continuance relation especially visible because a consumer's experience can be repeated frequently and recorded over time. Interface reliability, search relevance, account security, payment friction, social interactions, and changes to a subscription can all affect the evaluation of a provider. Bhattacherjee adapted expectation-confirmation theory to continued information-system use, linking confirmation, perceived usefulness, and satisfaction to continuance intention. The contribution is an explicit application of postadoption evaluation to technologies whose value and operating conditions evolve during use Bhattacherjee (2001).

A digital provider may change its product while retaining the same brand and account relationship. A once-satisfactory experience can cease to predict continued use when pricing, privacy practices, features, or the user's needs change. Moreover, automatic renewal and account lock-in may preserve subscription status after satisfaction has deteriorated. The observational challenge is to connect periodically elicited evaluations with usage behavior and actual choice opportunities, including whether a consumer understands cancellation options. A strong model therefore separates nominal retention, meaningful engagement, willing continuance, and the constraints governing exit Bhattacherjee (2001) Lemon & Verhoef (2016).

Customer journeys in digital environments also unfold across websites, mobile applications, notifications, contact centers, and independent review networks. A person can form expectations from information that the provider did not author, then attribute later failures to the provider whose brand remains salient. The implication is methodological: causal studies must specify which information surface was manipulated, what the consumer understood, and which later behavior was observed. A theory of digital brand loyalty gains explanatory power through such operational detail rather than through treating all technological interaction as a new psychological mechanism Lemon & Verhoef (2016) Keller (1993).

Algorithmic Mediation and Consumer Expectations in the Artificial Era

In the Artificial Era, recommendation systems, automated agents, conversational interfaces, and personalized ranking mechanisms increasingly mediate exposure to brands and the interpretation of alternatives. These systems can change the probability that a consumer encounters a brand, the information associated with it, and the timing of a subsequent purchase. They can also change the experience of a service through adaptive responses. These conditions create a need to distinguish brand-level expectation from expectation about the intermediary that presents or delivers the offering. The framework of the customer journey supplies an established starting point for locating these effects Lemon & Verhoef (2016).

A testable model separates three causal inputs: information supplied by a brand, information selected or generated by an intermediary, and performance experienced after choice. Research can manipulate ranking order, recommendation explanation, disclosed sponsorship, and the accuracy of predictions while holding product attributes constant. Outcomes should include prior beliefs, confidence, immediate preference, experienced quality, satisfaction, and eventual switching, not only click-through rate. When a conversational interface provides a confident product recommendation, perceived credibility may influence uptake, but the effect on long-term satisfaction and loyalty remains a question for experimentation and longitudinal observation Keller (1993) Bhattacherjee (2001).

The study of artificial mediation also clarifies the distinction between an artificial system's policy of selecting information and a human consumer's psychological loyalty. A model may optimize a retention target without possessing the customer's values or experiencing satisfaction. The relevant scientific variables are observable outputs, selection rules, exposure histories, human evaluations, and downstream consequences. This approach makes artificial systems investigable as parts of the decision environment and avoids explaining consumer behavior by imputing unmeasured mental states to a technological intermediary Lemon & Verhoef (2016) Dick & Basu (1994).

Consumer Welfare, Informed Choice, and Research Ethics

Consumer welfare is not equivalent to the longest possible retention. A customer who stays because cancellation is obscure, data transfer is difficult, or an essential service is geographically monopolized may report low satisfaction and still appear behaviorally loyal. Conversely, switching after new information becomes available can express effective choice rather than a failure of commitment. From a psychological perspective, evaluating an ecosystem requires attention to agency, informed expectations, the accessibility of alternatives, and the consumer's own goals Dick & Basu (1994) Seiders et al. (2005).

Marketing information that changes anticipated or experienced quality also raises questions of transparency and evidence. The experiments on branded sensory judgments show that context can influence evaluation; they do not justify misleading claims or the exploitation of vulnerable consumers. Ethical study designs should communicate the use of experimental information as permitted by applicable research oversight, limit deception to justified protocols with appropriate debriefing, and avoid exposing participants to harmful products or consequential misrepresentation. In digital research, behavioral tracking further requires lawful data practices, meaningful protection of personal information, and proportionate collection Lee et al. (2006) Plassmann et al. (2008).

Fair interpretation of loyalty metrics matters for organizational governance. A rise in repeat purchases may arise from improved service, an expanded product network, or increased exit friction. Those pathways have different welfare implications, even if the short-term retention indicator is identical. Scientific assessment should therefore report satisfaction and revealed behavior alongside access, switching conditions, consumer complaints, and relevant distributional differences. This multidimensional accountability is especially important where algorithmic systems can alter exposure and choice architecture at scale Fornell et al. (1996) Lemon & Verhoef (2016).

Critical Scientific Synthesis of the Main Theoretical Models

The expectation–disconfirmation family explains how anticipated and experienced performance enter a subsequent judgment. Brand-equity models explain how learned associations generate expectations before a particular encounter. Brand-experience models identify sensory, affective, intellectual, and behavioral dimensions of contact. Relationship models incorporate trust, attachment, and commitment; loyalty models distinguish attitudes and action; dynamic customer models introduce feedback, history, and exit opportunities. Their explanatory ranges overlap but are not coextensive. A comprehensive account draws on each at the level where its constructs are measurable Oliver (1980) Keller (1993) Brakus et al. (2009) Dick & Basu (1994).

The strongest directly interpretable experimental evidence in this field concerns specified information effects on judgments under controlled conditions. Evidence linking satisfaction to actual future purchase or relationship duration is often observational but behaviorally grounded, with more realistic opportunities for choice and greater confounding. Meta-analytic findings indicate recurring associations across a broader literature, but the aggregation does not erase differences in construct definitions or design. Qualitative relationship research elucidates meanings and relational forms that standard metrics can miss while supporting a different type of generalization Lee et al. (2006) Plassmann et al. (2008) Mittal & Kamakura (2001) Szymanski & Henard (2001) Fournier (1998).

A parsimonious synthesis distinguishes five analytic states: prior brand knowledge and anticipation; encountered product or service performance; situated appraisal and satisfaction; evolving trust, affect, commitment, and habit; and later behavior under available alternatives. These states are conceptually separable and reciprocally connected through feedback. This ordering is an analytical map derived from the reviewed research, not a newly validated stage theory. It organizes where competing explanations should be measured and where researchers should expect individual, contextual, and temporal heterogeneity Boulding et al. (1993) Gustafsson et al. (2005) Johnson et al. (2006).

Unresolved Problems and a Testable Research Agenda

The first unresolved problem is how much brand information changes the sensory experience itself, as opposed to the explanation and memory of an experience. A preregistered design could cross pre-use and post-use information timing with blinded and labeled conditions, collect both immediate and delayed evaluations, and measure subsequent consequential choices. Manipulation checks would verify that the cue changed expectations; independent sensory comparisons would constrain competing explanations; replication across categories would test generality. Such research would extend historical branded-taste paradigms without treating brain activity or self-report as interchangeable evidence of perception Allison & Uhl (1964) Lee et al. (2006) Plassmann et al. (2008).

The second problem concerns the transition from satisfaction to durable behavioral loyalty. A longitudinal panel can record episode-level satisfaction, competing offers, relationship age, habit strength, and switching friction, then link these to transactions. Competing models would test whether trust, commitment, and prior habit mediate or moderate behavioral effects, and whether effects change at the beginning and later stages of a relationship. Separate validation samples and sensitivity analyses would test whether the pattern is stable rather than a product of scale thresholds or selective retention Mittal & Kamakura (2001) Gustafsson et al. (2005) Johnson et al. (2006).

The third problem is construct equivalence across markets and cultures. Studies should test whether respondents understand satisfaction, trust, and willingness to recommend comparably, and whether observed purchases represent meaningful choice opportunities. The fourth concerns human behavior in algorithmically mediated markets: randomized ranking and recommendation explanations can be followed by real use, satisfaction, and switching measurements. The fifth concerns welfare: researchers need ways to distinguish freely maintained loyalty from retention secured by frictions and asymmetric information. Each program requires explicit ethical protections and a realistic account of the environment in which behavior becomes observable Khamitov et al. (2024) Lemon & Verhoef (2016).

Scientific Implications for Consumer and Economic Psychology

The accumulated literature supports treating expectation as a process with both predictive and evaluative functions. A brand association can motivate trial, shape attention and reported experience, and supply a benchmark used in satisfaction judgments. The evidence also supports distinguishing momentary evaluation from cumulative relationship assessments, and identifying separate routes through which satisfaction, trust, affect, and habit relate to later behavior. Scientific explanation improves when the temporal sequence is measured instead of compressed into a single survey taken after the purchase Anderson (1973) Oliver (1980) Bolton (1998).

A central practical consequence is that consumer research should report the outcome it actually observed. Preference ratings are evidence about stated evaluation; behavioral purchase records are evidence about choice under a particular opportunity structure; exit and renewal are evidence about relationship persistence; and the corresponding psychological mechanisms require further tests. Satisfaction indices and revenue analyses are valuable for describing organizations, but translation to individual causal claims should be made through linked data and a defensible research design Anderson et al. (1994) Fornell et al. (1996) Seiders et al. (2005).

The interdisciplinary importance of the subject lies in its integration of prediction, memory, affect, identity, social trust, and behavioral adaptation. Consumer and Economic Psychology supplies the primary disciplinary home. Cognitive Psychology contributes mechanisms of expectancy and evaluation, and Social Psychology contributes relationship and reputational processes. In the Artificial Era, interaction with adaptive interfaces and recommendation systems creates new experimental conditions for studying established psychological constructs. The task is to identify how those conditions alter observable human experience and choice, preserving operational precision and an explicit separation between evidence and proposed extension Arnould & Thompson (2005) Lemon & Verhoef (2016).

Conclusion

Brand expectations, consumption satisfaction, and loyalty constitute a connected psychological research field organized across time. Learned brand associations influence anticipation before purchase; actual experiences and comparison standards contribute to satisfaction; subsequent memory, trust, affect, commitment, and habit shape how satisfaction enters later decisions. Historical experiments and modern research establish that contextual information can influence reported consumption experiences, while longitudinal and behavioral studies demonstrate that satisfaction is only one determinant of continued patronage. The most productive scientific account preserves these constructs and the evidence supporting them as distinguishable elements of one evolving process Allison & Uhl (1964) Oliver (1980) Dick & Basu (1994) Mittal & Kamakura (2001).

The research record rejects a universal equation between a high satisfaction score and future loyalty. Repeated purchase may express favorable commitment, convenient routine, limited alternatives, or barriers to exit. A credible explanation identifies which pattern occurred and under what opportunities for choice. Likewise, an immediate change in a taste rating under a brand cue demonstrates a situated evaluative effect rather than durable attachment. Foundational conceptual models remain essential for framing these distinctions; experiments, longitudinal data, cross-context replication, and appropriately interpreted evidence syntheses determine which proposed links are empirically supported Lee et al. (2006) Gustafsson et al. (2005) Szymanski & Henard (2001).

The resulting research priorities are clear: prospective measurement of expectations, differentiation of perceived performance from postconsumption appraisal, observation of behavior across actual alternatives, careful assessment of trust and relationship mechanisms, and testing across contexts of culture, product category, and digital mediation. Consumer psychology thereby gains an account of loyalty grounded in how expectations are formed, how experiences are evaluated, and how choices are made repeatedly within changing environments. This synthesis provides a durable foundation for further evidence-based investigation of consumption during the Artificial Era Khamitov et al. (2024) Lemon & Verhoef (2016).

References

  • Allison, R. I., & Uhl, K. P. (1964). Influence of beer brand identification on taste perception. Journal of Marketing Research, 1(3), 36–39. https://doi.org/10.1177/002224376400100305
  • Anderson, R. E. (1973). Consumer dissatisfaction: The effect of disconfirmed expectancy on perceived product performance. Journal of Marketing Research, 10(1), 38–44. https://doi.org/10.1177/002224377301000106
  • Anderson, E. W., Fornell, C., & Lehmann, D. R. (1994). Customer satisfaction, market share, and profitability: Findings from Sweden. Journal of Marketing, 58(3), 53–66. https://doi.org/10.1177/002224299405800304
  • Arnould, E. J., & Thompson, C. J. (2005). Consumer culture theory (CCT): Twenty years of research. Journal of Consumer Research, 31(4), 868–882. https://doi.org/10.1086/426626
  • Bhattacherjee, A. (2001). Understanding information systems continuance: An expectation-confirmation model. MIS Quarterly, 25(3), 351–370. https://doi.org/10.2307/3250921
  • Bolton, R. N. (1998). A dynamic model of the duration of the customer's relationship with a continuous service provider: The role of satisfaction. Marketing Science, 17(1), 45–65. https://doi.org/10.1287/mksc.17.1.45
  • Boulding, W., Kalra, A., Staelin, R., & Zeithaml, V. A. (1993). A dynamic process model of service quality: From expectations to behavioral intentions. Journal of Marketing Research, 30(1), 7–27. https://doi.org/10.1177/002224379303000102
  • Brakus, J. J., Schmitt, B. H., & Zarantonello, L. (2009). Brand experience: What is it? How is it measured? Does it affect loyalty? Journal of Marketing, 73(3), 52–68. https://doi.org/10.1509/jmkg.73.3.052
  • Chaudhuri, A., & Holbrook, M. B. (2001). The chain of effects from brand trust and brand affect to brand performance: The role of brand loyalty. Journal of Marketing, 65(2), 81–93. https://doi.org/10.1509/jmkg.65.2.81.18255
  • Churchill, G. A., Jr., & Surprenant, C. (1982). An investigation into the determinants of customer satisfaction. Journal of Marketing Research, 19(4), 491–504. https://doi.org/10.1177/002224378201900410
  • Dick, A. S., & Basu, K. (1994). Customer loyalty: Toward an integrated conceptual framework. Journal of the Academy of Marketing Science, 22(2), 99–113. https://doi.org/10.1177/0092070394222001
  • Fornell, C. (1992). A national customer satisfaction barometer: The Swedish experience. Journal of Marketing, 56(1), 6–21. https://doi.org/10.1177/002224299205600103
  • Fornell, C., Johnson, M. D., Anderson, E. W., Cha, J., & Bryant, B. E. (1996). The American Customer Satisfaction Index: Nature, purpose, and findings. Journal of Marketing, 60(4), 7–18. https://doi.org/10.1177/002224299606000403
  • Fournier, S. (1998). Consumers and their brands: Developing relationship theory in consumer research. Journal of Consumer Research, 24(4), 343–373. https://doi.org/10.1086/209515
  • Garbarino, E., & Johnson, M. S. (1999). The different roles of satisfaction, trust, and commitment in customer relationships. Journal of Marketing, 63(2), 70–87. https://doi.org/10.1177/002224299906300205
  • Grönroos, C. (1984). A service quality model and its marketing implications. European Journal of Marketing, 18(4), 36–44. https://doi.org/10.1108/EUM0000000004784
  • Gustafsson, A., Johnson, M. D., & Roos, I. (2005). The effects of customer satisfaction, relationship commitment dimensions, and triggers on customer retention. Journal of Marketing, 69(4), 210–218. https://doi.org/10.1509/jmkg.2005.69.4.210
  • Homburg, C., Koschate, N., & Hoyer, W. D. (2005). Do satisfied customers really pay more? A study of the relationship between customer satisfaction and willingness to pay. Journal of Marketing, 69(2), 84–96. https://doi.org/10.1509/jmkg.69.2.84.60760
  • Johnson, M. D., Herrmann, A., & Huber, F. (2006). The evolution of loyalty intentions. Journal of Marketing, 70(2), 122–132. https://doi.org/10.1509/jmkg.70.2.122
  • Keiningham, T. L., Cooil, B., Andreassen, T. W., & Aksoy, L. (2007). A longitudinal examination of Net Promoter and firm revenue growth. Journal of Marketing, 71(3), 39–51. https://doi.org/10.1509/jmkg.71.3.039
  • 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
  • Khamitov, M., Rajavi, K., Huang, D.-W., & Hong, Y. (2024). Consumer trust: Meta-analysis of 50 years of empirical research. Journal of Consumer Research, 51(1), 7–18. https://doi.org/10.1093/jcr/ucad065
  • Lee, L., Frederick, S., & Ariely, D. (2006). Try it, you'll like it: The influence of expectation, consumption, and revelation on preferences for beer. Psychological Science, 17(12), 1054–1058. https://doi.org/10.1111/j.1467-9280.2006.01829.x
  • Lemon, K. N., & Verhoef, P. C. (2016). Understanding customer experience throughout the customer journey. Journal of Marketing, 80(6), 69–96. https://doi.org/10.1509/jm.15.0420
  • Mittal, V., & Kamakura, W. A. (2001). Satisfaction, repurchase intent, and repurchase behavior: Investigating the moderating effect of customer characteristics. Journal of Marketing Research, 38(1), 131–142. https://doi.org/10.1509/jmkr.38.1.131.18832
  • Oliver, R. L. (1980). A cognitive model of the antecedents and consequences of satisfaction decisions. Journal of Marketing Research, 17(4), 460–469. https://doi.org/10.1177/002224378001700405
  • Oliver, R. L. (1999). Whence consumer loyalty? Journal of Marketing, 63(Special Issue), 33–44. https://doi.org/10.1177/00222429990634s105
  • Parasuraman, A., Zeithaml, V. A., & Berry, L. L. (1985). A conceptual model of service quality and its implications for future research. Journal of Marketing, 49(4), 41–50. https://doi.org/10.1177/002224298504900403
  • Plassmann, H., O'Doherty, J., Shiv, B., & Rangel, A. (2008). Marketing actions can modulate neural representations of experienced pleasantness. Proceedings of the National Academy of Sciences, 105(3), 1050–1054. https://doi.org/10.1073/pnas.0706929105
  • Seiders, K., Voss, G. B., Grewal, D., & Godfrey, A. L. (2005). Do satisfied customers buy more? Examining moderating influences in a retailing context. Journal of Marketing, 69(4), 26–43. https://doi.org/10.1509/jmkg.2005.69.4.26
  • Szymanski, D. M., & Henard, D. H. (2001). Customer satisfaction: A meta-analysis of the empirical evidence. Journal of the Academy of Marketing Science, 29(1), 16–35. https://doi.org/10.1177/009207030102900102
  • Tse, D. K., & Wilton, P. C. (1988). Models of consumer satisfaction formation: An extension. Journal of Marketing Research, 25(2), 204–212. https://doi.org/10.1177/002224378802500209
  • 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