0 cumulative citations
View corpus contextAlgorithmic personalization drives impulse purchases and raises short-term perceived value, but sustaining customer lifetime value hinges on trust: transparency and perceived control jointly shape whether personalization builds or undermines long-term loyalty, with cultural context altering the transparency–control interaction.
Algorithmic personalization has become a key strategy, impacting how customers make spontaneous purchases and their lifetime value - Evidence from Brazil and Indonesia
Using a cross-sectional survey of 177 Brazilian and Indonesian digital consumers and PLS-SEM, the paper finds algorithmic personalization is associated with increased impulse buying and perceived pre-purchase value (price, emotional, social), with consumer trust mediating effects on customer lifetime value and transparency × perceived control interacting (moderated by cultural context).
Citation observations
Cumulative provider counts captured on specific dates; providers are never combined.
Algoritmik kişiselleştirmenin dürtüsel satın alma ve müşteri yaşam boyu değeri üzerindeki etkisi teorik ve ampirik açıdan yeterince araştırılmamış olmakla birlikte, algoritmik kişiselleştirmenin e-ticaret ortamında tüketici davranışlarını ve tüketicilerin hızla değişen algoritmik uygulamalara yönelik tutumlarını önemli ölçüde şekillendirdiği bilinmektedir. Bu çalışma, ilgili literatüre dayanarak bütünleşik bir kavramsal araştırma modeli geliştirmektedir. PLS-SEM temelli modelde, algılanan değerin çok boyutlu yapısı aracı değişken olarak ele alınırken algoritmik şeffaflık ve algılanan kontrolün tüketici güveni üzerindeki etkisi ile tüketici güveninin müşteri yaşam boyu değeri (CLV) üzerindeki etkisi incelenmektedir. Ayrıca kişilik, güç mesafesi ve dijital okuryazarlık kontrol değişkenleri olarak modele dâhil edilmiştir. Ekim–Kasım 2025 döneminde Brezilya’dan 76 ve Endonezya’dan 101 olmak üzere toplam 177 yetişkin dijital tüketiciden kesitsel veriler toplanmıştır. Araştırma kapsamında test edilen yedi hipotezin tamamı desteklenmiştir. Bulgular, algoritmik kişiselleştirmenin dürtüsel satın alma davranışını olumlu yönde etkilediğini göstermektedir. Bu etkiye kısmen algılanan fiyat değeri (dolaylı etki = 0,187) ile duygusal ve sosyal değer aracılık etmektedir. Dışa dönüklük ve dijital öz yeterlilik de dürtüsel satın alma tepkisini güçlendiren faktörler arasında yer almaktadır. Algoritmik şeffaflık ile tüketici güveni arasındaki ilişki kısmen desteklenmiştir. Tüketici güveni, açıklanan varyansın %42,9’unu oluşturarak ilgili etki mekanizmasının en önemli aracısı olarak ortaya çıkmış ve toplam etkinin yaklaşık %20’sine aracılık etmiştir. Ayrıca tüketici güveninin modeldeki en güçlü yordayıcı değişken olduğu belirlenmiştir (β = 0,612; f² = 0,489). Yüksek düzeyde algılanan kontrol ile yüksek algoritmik şeffaflık arasında teorik açıdan sinerjik bir etkileşim bulunmaktadır ve bu ortak etki, iki unsurun bağımsız etkilerinin toplamından daha güçlüdür. Bu etki; güç mesafesi düşük tüketiciler ile dijital okuryazarlık düzeyi yüksek tüketiciler arasında daha belirgindir. Çoklu grup analizi, algoritmik şeffaflık ile algılanan kontrol arasındaki etkileşimin kültürel bağlamın düzenleyici etkisi altında anlamlı olduğunu; buna karşılık temel yapısal ilişkilerin kültürel bağlama göre önemli ölçüde farklılaşmadığını göstermektedir. Çalışma, algoritmik şeffaflık ve algılanan kontrolün ikincil etik meseleler olarak değerlendirilmesi anlayışını tersine çevirmektedir. Algoritmalar çağında sürdürülebilir müşteri yaşam boyu değeri, tüketicileri somut biçimde güçlendiren yönetişim uygulamaları aracılığıyla gerek işletme gerekse tedarik zinciri düzeyinde gerçekleştirilen bir güven inşa süreci olarak değerlendirilmektedir.
Summary
Main Finding
Algorithmic personalization increases impulse purchases and can raise customer lifetime value (CLV), but its long-term benefit depends crucially on consumer trust — which in turn is shaped by algorithmic transparency and perceived control. Transparency and perceived control interact synergistically (especially across cultural contexts), and trust is the single strongest mediator linking personalization-related perceptions to sustainable CLV (reported beta b = 0.612, effect size f2 = 0.489). Short-term impulse gains are partly mediated by perceived value dimensions (price, emotional, social), with the indirect price effect reported as 0.187.
Key Points
- Sample & context: cross-sectional survey of N = 177 adult digital consumers collected Oct–Nov 2025 (Brazil n = 76; Indonesia n = 101).
- Model & hypotheses: PLS-SEM model with seven hypotheses; model tests effects of algorithmic transparency and perceived control on consumer trust, and trust on CLV; perceived pre-purchase value (price, emotional, social) mediates personalization → impulse buying; controls include personality, power distance, and digital literacy. All seven hypotheses were validated.
- Mediating/value channels:
- Perceived price, emotional, and social value mediate the effect of algorithmic personalization on impulse buying. Reported indirect effect via perceived price = 0.187.
- Consumer trust is the most important mediator in the pathway to CLV (author reports consumer trust related metrics as 42.9% and as contributing nearly 20% of the total pathway).
- Effect sizes:
- Reported direct effect of consumer trust on CLV: b = 0.612, f2 = 0.489 (large effect).
- Moderation & heterogeneity:
- Personality (notably extraversion) and digital self-efficacy amplify impulse responses to personalization.
- Cultural context (tested via multi-group analysis: Brazil vs Indonesia) moderates the interaction between transparency and perceived control: the transparency × control interaction is significant under cultural moderation, while the main structural paths remain largely invariant across the two countries.
- Strategic paradox: tactics that maximize short-term impulse conversions (urgency cues, scarcity messages) can undermine trust and long-term CLV unless paired with transparency and consumer control mechanisms.
Data & Methods
- Design: Cross-sectional online survey (Oct–Nov 2025).
- Sample: 177 adult digital consumers (Brazil 76; Indonesia 101).
- Analysis: Partial Least Squares Structural Equation Modeling (PLS-SEM); multi-group analysis to compare country (cultural) effects.
- Variables:
- Predictors: Algorithmic personalization, algorithmic transparency, perceived control.
- Mediators: Multi-dimensional perceived pre-purchase value (price, emotional, social), consumer trust.
- Outcome: Customer Lifetime Value (CLV).
- Controls: Personality traits (e.g., extraversion), power-distance (cultural), digital literacy/self-efficacy.
- Reported statistics: indirect effect via perceived price = 0.187; consumer trust → CLV b = 0.612, f2 = 0.489; author-reported mediation shares (consumer trust reported around 42.9% and ~20% of total pathway in different descriptive indicators).
Implications for AI Economics
- Investing in transparency and user control has measurable economic returns: improving algorithmic transparency and perceived control helps build trust, which strongly increases CLV (large effect size reported). Firms should view transparency not only as compliance but as an investment in long-term revenue.
- Trade-off management: Optimization metrics that prioritize immediate conversions (impulse buys) may harm CLV if they erode trust. Platform and recommender design should balance short-term revenue tactics with features that sustain trust (explainability, control dashboards, query/refresh profile tools).
- Heterogeneous returns: The economic impact of transparency/control features varies by user traits and cultural context. Segmented deployment (e.g., adjusting visibility/control affordances by market or user digital self-efficacy) can improve ROI.
- Measurement: CLV is a practical endpoint for quantifying the financial impact of algorithmic governance choices. Firms and researchers should incorporate mediators (perceived value, trust) and moderators (personality, power distance, digital literacy) when modeling the economic payoff of personalization investments.
- Policy and regulation: Findings support policies that encourage transparency and user control mechanisms; such regulation can align consumer welfare with longer-run firm profitability.
- Research priorities: Larger, longitudinal and experimental studies should quantify causal impacts on turnover/churn and validate cross-cultural generalizability; economic models should incorporate trust dynamics and heterogeneous preference weights when forecasting CLV under different personalization regimes.
Limitations noted by the paper (implicit): modest sample size, cross-sectional design, self-reported measures, and testing in two countries — so causal claims and broad generalization require further work.
Assessment
Paper Typecorrelational
Evidence Strengthlow — The study uses cross-sectional self-reported survey data and structural equation modeling, which can establish associations and test mediation models but cannot plausibly identify causal effects; small sample size and potential common-method and measurement bias further weaken causal claims.
Methods Rigormedium — The authors employ PLS-SEM, test mediation and moderation, include control variables (personality, power distance, digital literacy), and run multi-group analysis across two countries, which is appropriate for theory testing. However, the sample is small and likely non-probability, measures appear self-reported, there is no temporal ordering, and there is little evidence of robustness checks or methods that address endogeneity or common-method bias.
SampleCross-sectional convenience sample of N = 177 adult digital consumers surveyed in October–November 2025: Brazil (n = 76) and Indonesia (n = 101); analyses are based on self-reported measures (impulse buying, perceived price/emotional/social value, perceived control, transparency, trust, perceived CLV) and modeled using PLS-SEM.
Themesadoption governance
IdentificationCross-sectional online survey (N=177) analyzed with PLS-SEM to estimate associations, mediation (perceived value, trust) and moderation (transparency × control, cultural multi-group analysis); no experimental manipulation, instrumental variables, or longitudinal variation to support causal identification.
GeneralizabilitySmall, convenience sample limits external validity and power., Only two middle-income countries (Brazil, Indonesia); results may not generalize to high-income markets or other cultural contexts., Measures appear to be self-reported perceptions (including CLV) rather than observed transactional CLV, limiting behavioral inference., Cross-sectional design prevents inference about dynamics or long-term effects., Platform/sector heterogeneity (which e-commerce platforms or product categories) is not specified, reducing applicability across contexts.
Claims (6)
| Claim | Direction | Outcome | Confidence & Evidence | Details |
|---|---|---|---|---|
| In a PLS-SEM model estimated using cross-sectional data, all seven hypotheses were validated. Other | positive | Hypothesis validation within the structural model |
Reading fidelity
high
Study strength
low
|
n=177
|
| Algorithmically personalized impulse-buying behavior was positively associated with algorithmic personalization, with part of the relationship mediated by perceived price, emotional value, and social value. Consumer Welfare | positive | Algorithmically personalized impulse-buying behavior |
Reading fidelity
medium
Study strength
low
|
n=177
indirect effect = 0.187
|
| Extraversion and digital self-efficacy were identified as factors that promote consumers' impulse responses to algorithmically personalized content. Consumer Welfare | positive | Impulse response to algorithmically personalized content |
Reading fidelity
high
Study strength
low
|
n=177
|
| Consumer trust was a major mediator linking algorithmic transparency to customer lifetime value, accounting for 42.9 percent of the pathway and approximately 20 percent of the total pathway. Firm Revenue | positive | Customer lifetime value |
Reading fidelity
medium
Study strength
low
|
n=177
42.9 percent; approximately 20 percent of the total pathway
|
| Consumer trust was the strongest reported predictor in the model, with a standardized coefficient of 0.612 and an f-squared effect size of 0.489. Firm Revenue | positive | Customer lifetime value |
Reading fidelity
high
Study strength
low
|
n=177
b = 0.612, f2 = 0.489
|
| The interaction between algorithmic transparency and perceived control was significant under cultural-context moderation, while the main structural paths did not differ by cultural context. Consumer Welfare | mixed | Transparency-control interaction and stability of structural paths across cultural contexts |
Reading fidelity
high
Study strength
low
|
n=177
|