Suggestions
Andrew Yates
CEO and Co-Founder @ Promoted.ai | (YC W21) Marketplace Search
Andrew Yates is the CEO and Co-Founder of Promoted.ai, a company that specializes in optimizing search and advertising for online marketplaces. He launched Promoted.ai in 2020 alongside co-founder Dan Hill, both of whom previously held engineering leadership roles at Pinterest and Facebook, focusing on ad systems and machine learning.125
Background and Experience
Before founding Promoted.ai, Andrew Yates worked extensively in the tech industry, particularly in advertising technology. He led teams at Facebook, where he was involved in ads ranking and auction systems, and at Pinterest, where he contributed to marketplace engineering. His expertise lies in designing scalable systems that enhance revenue through improved search and advertising strategies.125
Vision for Promoted.ai
Under Yates's leadership, Promoted.ai aims to connect buyers and sellers across various applications by improving marketplace efficiency. The company's technology integrates search results with advertising to boost revenue for clients, which include major players like Wayfair and Upwork. Yates emphasizes that the ultimate goal is not merely to implement ads but to increase average revenue per user (ARPU) sustainably.135
Company Achievements
Promoted.ai is recognized for its innovative approach to marketplace optimization, having been part of Y Combinator's Winter 2021 cohort. The company has successfully increased client revenues by 9% to 11% through its advanced algorithms and engineering solutions.345
Yates continues to advocate for the strategic use of AI in enhancing marketplace dynamics, focusing on creating win-win scenarios for all stakeholders involved.23
Highlights
We're delighted to announce our unified search and native ads launch for Upwork! Thank you #Upwork and the Upwork search and ads team!
Wow, I didn't realize how important data formats like Paimon were for machine learning until we built our own system! Thanks to Xingcan Cui , Pei He, and Jun Sun, we now can update versus overwriting our training examples as labels are updated. This helps us train post-purchase revenue models much more efficiently because each training example is a truncated observation window waiting for a revenue event potentially many days in the future. Thanks, data team!
