Reviews and reputation · Observational clickstream study of purchases and reviews, supplemented by supervised machine-learning models trained and validated with Amazon Mechanical Turk product-pair labels to identify complementary and substitutive products.
On the Spillover Effects of Online Product Reviews on Purchases: Evidence from Clickstream Data
Young Kwark, Gene Moo Lee, Paul A. Pavlou, Liangfei Qiu (2021). On the Spillover Effects of Online Product Reviews on Purchases: Evidence from Clickstream Data. Information Systems Research. https://doi.org/10.1287/isre.2021.0998.
Plain-English take
Mean ratings of substitutive products negatively affect focal-product purchase while ratings of complementary products have a positive effect; spillovers are stronger on mobile, across different brands for substitutive products, among less experienced consumers, and when rating variance is lower. For your marketing: Use it as a product, platform or behavior mechanism reference rather than a direct forecast for a SaaS funnel.
How much weight to give this
2.5/5 Evidence needs cautious interpretation
The 1–5 evidence-check score is a practical reading aid, not a formal peer review or a probability that a conclusion is true. It combines design strength (2 points), sample reporting (1), comparison/control clarity (1), and how much of the paper could be checked (1). Access affects how confidently this directory can assess a study, not the study's underlying quality. Blinding is reported separately because it is not relevant to every design.
Observational clickstream study of purchases and reviews, supplemented by supervised machine-learning models trained and validated with Amazon Mechanical Turk product-pair labels to identify complementary and substitutive products.
Clickstream data from a large retailer, examining reviews of covisited products and purchases of a focal product; moderators include device (mobile versus PC), brand relationship, consumer experience, and rating variance.
Based on the publisher-described design and comparison; see the article for complete analysis.
Publisher record was fetched directly; full text was not assumed open.
Double-blind? Blinding was not established from the retrieved publisher record.
Research question
What the authors investigated
How do online reviews of products viewed alongside a focal product spill over into focal-product purchases, and how do complementarity/substitution, device, brand, experience, and rating variance moderate the effect?
Design
Observational clickstream study of purchases and reviews, supplemented by supervised machine-learning models trained and validated with Amazon Mechanical Turk product-pair labels to identify complementary and substitutive products.
Sample and setting
Clickstream data from a large retailer, examining reviews of covisited products and purchases of a focal product; moderators include device (mobile versus PC), brand relationship, consumer experience, and rating variance.
Reported result
What the study found
Mean ratings of substitutive products negatively affect focal-product purchase while ratings of complementary products have a positive effect; spillovers are stronger on mobile, across different brands for substitutive products, among less experienced consumers, and when rating variance is lower.
This is a paraphrase of the source. Check the original paper for definitions, estimates, statistical uncertainty and analysis details.
Scope and transfer
Limitations to keep in view
Not stated in the publisher abstract.
Possible founder application
Use it as a product, platform or behavior mechanism reference rather than a direct forecast for a SaaS funnel.
Evidence access
Publisher abstract/metadata retrieved; consult the linked paper for full methods and tables.
INFORMS states that the study uses clickstream data from a large retailer and reports negative substitutive-product and positive complementary-product review spillovers, with the listed moderation patterns. Direct publisher-page retrieval was recorded on 2026-10-09; profile summarizes the accessible record and does not claim independent review of inaccessible full text.
If only an abstract or metadata was available, the profile is limited to claims supported there. A paper's paywall does not change its evidentiary status, but it limits what we can substantiate from accessible text.