Factors that affect consumer trust in product quality: a focus on online reviews and shopping platforms

Online shopping has grown rapidly due to convenience and ease of accessing product information and comparing prices. However, evaluating the quality of experience goods (e.g., clothing, cosmetics) prior to purchase is difficult online, creating uncertainty and making trust a critical factor in purchase decisions. Prior work shows that online reviews and the reputation of shopping platforms can foster trust, as can brand reputation. Yet few studies jointly examine which specific review and platform attributes are most trusted and economically valued by consumers when purchasing experience goods. This study asks: which attributes of online reviews and shopping platforms do consumers trust and value (in monetary terms) when purchasing experience goods online? Trust is defined as a factor that boosts consumer confidence in product quality, reducing perceived risk and helping cope with uncertainty. Reviews, star ratings, and reliable shopping platforms can increase trust for experiential products. The study develops six key attributes, conducts a conjoint choice experiment, and estimates relative importance and marginal willingness to pay using a multinomial logit model, with analyses split by famous versus nonfamous brands.

The related literature highlights that consumers use online review attributes—number of reviews, average star rating, review type (text, image, video), and review length—to inform purchase decisions, with higher star ratings generally perceived as more trustworthy. Image- and video-based reviews can better convey experiential qualities and are often preferred to text-only reviews; longer reviews tend to be more informative. Review reward programs vary widely across countries and platforms: in South Korea (e.g., Naver, 11Street) users often receive points or mileage for reviews; in China (e.g., Taobao) sellers may incentivize positive reviews; Amazon does not pay for reviews but operates Amazon Vine to seed products to reviewers. Shopping platform reputation influences trust formation; platforms can be categorized as personal shopping malls, open markets (e.g., Gmarket), and broader online platforms (e.g., Naver Shopping). Brand reputation also reduces uncertainty and positively influences trust and purchase intention; reviews can especially benefit less-known brands by building trust. Experience goods are those whose quality is known only after use; consumers rely more on online reviews for experience goods than for search goods, spending more time on review information. Prior studies suggest platform reputation and review attributes can reduce uncertainty and increase trust, but an integrated assessment of their relative importance and economic value for experience goods has been lacking.

Design: A choice-based conjoint analysis was used to measure preferences and monetary valuation (MWTP) for six attributes influencing trust in online shopping of experience goods: price, number of reviews, star rating, review type (text; text+picture; text+video), text review length (general: <20 words; premium: >300 words), and shopping platform (personal shopping mall; open market; online platform). Attribute levels (Price: 50,000; 52,000; 54,000 KRW; Number of reviews: 1, 10, 100, 1,000; Star ratings: 1, 3, 5; Review type: text, text+picture, text+video; Review length: general <20 words vs premium >300 words; Platform: personal mall [Hiphoper, Hyber], open market [Gmarket], online platform [Stylec, TheXshop]) were drawn from literature and real marketplaces (Table 2). Fractional factorial orthogonal design (SPSS 25) reduced 648 possible profiles to 25 alternatives. Respondents faced 10 choice tasks total: five for a famous brand (Nike) and five for a nonfamous brand (no brand), each with 5 alternatives per task, choosing one alternative per task. The product context was long-sleeved T-shirts (experience good). Model: Choices were analyzed using a multinomial logit (MNL) model U_nj = Σ_i β_i x_ij + ε_nj with type I extreme value errors, yielding P_nj = exp(V_nj)/Σ_i exp(V_ni). Relative importance (RI) was computed from part-worth ranges across attribute levels and averaged across respondents. Marginal willingness to pay (MWTP) was calculated as MWTP_k = β_k/β_price. Data: An online panel survey in South Korea recruited 528 consumers with online shopping experience (April 28–May 4, 2021), ages 20–59 (approximately even distribution), gender-balanced (50.19% male), with occupations mainly office workers/technicians (46.21%). Inclusion required online/mobile shopping in the last three months.

Overall importance and valuation (famous brand baseline): • Star rating was the most important attribute (RI = 28.75%), followed by number of reviews (17.38%), platform—online platform (12.03%) and open market (8.27%)—price (10.83%), premium review length (10.26%), picture reviews (7.39%), and video reviews (5.10%). Several review and platform attributes were more important than price. • Price coefficient: −0.0002 (p<0.001). • Number of reviews: β = 0.0011 (p<0.001); MWTP ≈ 6.428 KRW per review (≈642.80 per 100; 6,428 per 1,000). • Star rating: β = 0.4446 (p<0.001); MWTP ≈ 2,655.95 KRW per star. A 4-star vs 1-star product implies ≈7,967.85 KRW higher price. • Review types: text+picture β = 0.4570 (p<0.001), MWTP ≈ 2,730.21 KRW; text+video β = 0.3155 (p<0.001), MWTP ≈ 1,885.00 KRW. • Review length: premium (>300 words) β = 0.6345 (p<0.001), MWTP ≈ 3,790.43 KRW. • Platform: open market β = 0.5114 (p<0.001), MWTP ≈ 3,055.26 KRW; online platform β = 0.7442 (p<0.001), MWTP ≈ 4,446.08 KRW. • Consumers are willing to pay more to buy via trusted platforms; e.g., a 50,000 KRW item could command ≈54,446 KRW on an online platform. Famous vs nonfamous brand comparison: • Relative importance shifts slightly for nonfamous brands: star rating (29.18%), number of reviews (18.39%), online platform (11.69%), premium reviews (10.36%), price (9.49%), open market (7.86%), picture (7.09%), video (5.95%). Premium review importance exceeds price for nonfamous brands. • MWTP values are higher across all attributes for nonfamous brands: — Number of reviews: ≈7.763 KRW per review (≈776.30 per 100). — Star rating: ≈3,076.04 KRW per star. — Picture reviews: ≈2,988.61 KRW; video reviews: ≈2,509.49 KRW. — Premium review length: ≈4,368.33 KRW. — Platform: open market ≈3,313.40 KRW; online platform ≈4,931.55 KRW. • Consumers rely more on reviews and platform trust when purchasing nonfamous brands and are willing to pay larger premiums for these trust-building attributes. Sample/context: 528 Korean online shoppers; product context long-sleeved T-shirts (experience good). All key coefficients reported were statistically significant at conventional levels.

The findings directly address the research question by identifying which review and platform attributes most influence trust and by quantifying their economic value. Star ratings and the number of reviews are pivotal in reducing uncertainty for experience goods, with review length and multimedia content further enhancing perceived diagnosticity. Trusted shopping platforms significantly increase willingness to pay, indicating that intermediary reputation mitigates perceived risk beyond seller-level factors. The higher MWTP for all attributes when purchasing nonfamous brands shows that consumers substitute platform and review-based trust for lacking brand equity, paying more for signals that reduce quality uncertainty. Practically, retailers—especially those with weaker brands—can command higher prices or improve conversion by leveraging reputable platforms and incentivizing richer, longer reviews (text plus images/videos). Platforms can calibrate review reward schemes in line with the demonstrated value of premium and multimedia reviews. These insights support pricing and channel strategy decisions, particularly in early product life stages with sparse reviews.

This study contributes by jointly estimating the relative importance and monetary value of six trust-related online shopping attributes for experience goods, showing that star ratings, review volume, review richness (length and media), and platform reputation often outweigh price in importance. Consumers exhibit substantial willingness to pay for trusted platforms and informative reviews, and this willingness is even higher for nonfamous brands, underscoring the role of third-party trust signals in compensating for low brand equity. Implications include: prioritize trusted channels (online platforms, open markets) for experiential products; design review systems and incentives that promote premium-length and multimedia reviews; and set pricing strategies that reflect the value consumers place on trust attributes. Future research should expand to search goods and other categories, test broader and alternative price ranges, include emerging global platforms (e.g., search engines’ shopping services), model preference heterogeneity using mixed logit, and examine potential attribute non-attendance to refine estimates.

• Scope limited to experience goods (clothing), which may constrain generalizability beyond experiential categories. • Conjoint design constraints prevented inclusion of all attribute combinations and required researcher judgment on attribute levels; some potentially relevant attributes were omitted. • Price range and increments (50,000–54,000 KRW) were limited; broader ranges could affect MWTP estimates. • Platform landscape is evolving (e.g., search engines expanding into shopping); results may vary across global platforms and over time. • Potential attribute non-attendance or random responding due to multiple choice tasks, although effort was made to keep attributes within cognitive limits. • Use of multinomial logit does not fully capture preference heterogeneity; future analyses could apply mixed logit or hierarchical Bayes approaches.