Advertising and promotion · Data-driven optimization and bandit-algorithm study validated through a two-sided randomized field experiment with a large-scale video-sharing platform; the experiment evaluated the proposed shadow bidding with learning algorithms for new-ad cold start.
Cold Start to Improve Market Thickness on Online Advertising Platforms
Renyu Zhang, Xin Chen, Zhiwei Xu, Zikun Ye, Dennis J. Zhang, Heng Zhang (2022). Cold Start to Improve Market Thickness on Online Advertising Platforms. Management Science. https://doi.org/10.1287/mnsc.2022.4550
Plain-English take
The algorithm increased cold-start success rate by 61.62%, with a 0.717% short-term-revenue compromise. It increased overall advertising-market thickness by 3.13% and long-term advertising revenue by at least 5.35%.
How much weight to give this
4/5 Field experiment with bounded scope
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.
Data-driven optimization and bandit-algorithm study validated through a two-sided randomized field experiment with a large-scale video-sharing platform; the experiment evaluated the proposed shadow bidding with learning algorithms for new-ad cold start.
Ads, advertisers, and users in the large-scale video-sharing-platform field experiment; the accessible publisher abstract does not state numeric sample counts.
See the full paper for comparison conditions and analysis.
Publisher page checked; full-text methods not independently reviewed.
Double-blind? Blinding was not established from the checked publisher page.
Research question
What the authors investigated
What does cold start to improve market thickness on online advertising platforms show about customer decisions or business outcomes?
Design
Data-driven optimization and bandit-algorithm study validated through a two-sided randomized field experiment with a large-scale video-sharing platform; the experiment evaluated the proposed shadow bidding with learning algorithms for new-ad cold start.
Sample and setting
Ads, advertisers, and users in the large-scale video-sharing-platform field experiment; the accessible publisher abstract does not state numeric sample counts.
Reported result
What the study found
The algorithm increased cold-start success rate by 61.62%, with a 0.717% short-term-revenue compromise. It increased overall advertising-market thickness by 3.13% and long-term advertising revenue by at least 5.35%.
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
The empirical validation is tied to one large-scale video-sharing platform and its online advertising system; the publisher page does not report the number of ads, advertisers, or users, so transferability and sample scale cannot be evaluated from the accessible article page alone. Full article methods and correction history were not independently reviewed.
Possible founder application
Use this study as a bounded example of cold start to improve market thickness on online advertising platforms; assess whether its population, intervention and outcome match your own decision.
Evidence access
Publisher article page; full-text access varies and was not established in this check.
Publisher page directly retrieved; candidate metadata and reported result checked against that page. Full methods were not independently reviewed.
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.