1 / 18

Online Advertisement Campaign Optimization

Online Advertisement Campaign Optimization. Shi Zhong Data Mining and Research Group Yahoo! Inc. Joint work with Weiguo Liu, Shyam Kapur, and Mayank Chaudhary, published in IEEE/INFORMS SOLI Conference. Agenda. Introduction to online advertising Online ad campaign optimization problem

avel
Download Presentation

Online Advertisement Campaign Optimization

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Online AdvertisementCampaign Optimization Shi Zhong Data Mining and Research Group Yahoo! Inc. Joint work with Weiguo Liu, Shyam Kapur, and Mayank Chaudhary, published in IEEE/INFORMS SOLI Conference

  2. Agenda • Introduction to online advertising • Online ad campaign optimization problem • Focus: display advertising (i.e., graphical/banner ads) • Approaches and results • Conclusion

  3. Text Ads Yahoo Sponsored Search

  4. Text Ads Google Content Match

  5. Display Ads on Yahoo LREC, 300x250

  6. Online Advertising • Text ads • Two main categories, a few major players • Sponsored searchE.g., Google search, Yahoo search, Live.com, Ask.com • Content matchE.g., Google adsense, Yahoo YPN • Cost models: CPC • Targeting: search query, page content • Display ads • Fragmented market • Cost models: CPM, CPC, CPA • Targeting: content, demo, geo, behavioral, or none

  7. Online Ad Campaign Optimization Yahoo Display Ads$150k{yahoo top page + LREC, yahoo movie + N, BT=entertainment/movie, …} Google Adwords$250k{dvd rental, online dvd, online movie, …} Ad Agencies DoubleClick$100k{CNN.COM + LREC, IMDB.com + N, …} Netflix, Q4 AdvertisingBudget=$500k,Drive traffic to netflix.com

  8. We focus on … • Display advertising campaigns • Optimize media buys given a campaign budget and/or campaign objectives • Maximize # conversions/clicks for a given budget • Minimize cost for a given number of conversions/clicks • Experiments inside Yahoo • Media buys limited to Yahoo products

  9. A Campaign Example • A campaign contains multiple lines/products • A line specifies a product from the publisher, a quantity, and a price • A product consists of page location, position, and profile

  10. Quantity and Price • Quantity is capped by inventory availability • Price is determined by a bidding process • Except for “guaranteed delivery” – for which advertisers have to pay a premium • Higher bid earns higher priority at ad delivery time, thus has a higher probability getting more impressions

  11. s.t. qi = # imps for line i, in thousands cpm = cost per thousand imps ctr = click through rate rpc = revenue per click Budget = total budget = max fraction of Budget per line = profit marginCapi = available # imps for line i Optimization Formulation - I Maximize profit for a given budget

  12. s.t. Optimization Formulation - II Minimize cost for a desired number of clicks nc = desired # clicks

  13. Test Results • Take a few historical campaigns with Yahoo for some advertiser • Compare simulated results from optimization formulation-II with historical campaigns • Average cost saving (for generating same number of clicks) is 26% sounds simple, but …

  14. Prepare inputs to optimization engine • Collect/generate product lines • Use historical lines of similar advertisers • Use data mining techniques to learn “new” lines that are expected to perform well • Use predictive modeling to discover/explore new lines • Estimate • CTR for each product • Quantity-CPM curve for each product • RPC for a given advertiser/business

  15. Identify High CTR Segments Data examples • Approach: • Extract frequent segments (with min # impressions) with frequent itemset mining algorithm • Calculate CTR for each segment • Check overlap and temporal stability for high CTR segments

  16. Identified Segment Examples • Example high-CTR segments • Page:News + Position:LREC + Age:35-54 CTR=0.31% • Page:Weather + Position:LREC CTR=0.32% • (Baseline average CTR ~ 0.03%) • CTR numbers seen to be stable over time • CPM estimated from most similar historical lines or Yahoo’s internal pricing system

  17. Conclusion • Data mining and optimization work together nicely to enhance campaign effectiveness • An optimized campaign can be very rewarding • Further research • Ad creative optimization • Landing page optimization

  18. Questions?

More Related