GermanyTechnology and telecom2022 – ongoingAwin
Machine learning recommendations of publisher partners for advertisers
Awin's recommendation engine raised the acceptance rate of publisher suggestions ninefold
Problem
Awin runs an affiliate marketing network that links advertisers with publishers. It wanted to point advertisers to the publisher partnerships most likely to succeed.
Approach
Awin built a machine learning recommendation engine on the Databricks platform. It identifies the advertiser and publisher pairings with the highest success rates. The company planned to add churn prediction, demand forecasting and fraud anomaly detection over the following 18 months.
Results
Takeaway
Recommendations only matter if members act on them, so acceptance rate is a direct way to judge whether a matching engine is useful.
Categories
- Industry
- Technology and telecom
- Business function
- Marketing and sales
- Innovation type
- Product and service innovation
- AI technology
- Recommendation and personalization
- Adoption
- Off-the-shelf tool or platform
- Evidence
- Vendor case study
Sources
Source summaries
Databricks reports that Awin used machine learning to recommend publisher partnerships to advertisers. The page says the member acceptance rate for these recommendations rose ninefold. It also lists platform results such as lower compute and hosting costs, which are not part of this case.
Full details are in the original sources.
