|
Activity Number:
|
279
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Tuesday, July 31, 2007 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Statistics and Marketing
|
| Abstract - #308492 |
|
Title:
|
Methodology and Challenges in Search Marketing at Business.com
|
|
Author(s):
|
Nathan Janos*+ and Tom Anderson and Paul Dagum
|
|
Companies:
|
Business.Com and Business.Com and Business.Com
|
|
Address:
|
2120 Colorado Blvd, Santa Monica, CA, 90404,
|
|
Keywords:
|
search marketing ; generalized linear mixed models ; revenue optimization ; profit optimization ; portfolio ; demand elasticity
|
|
Abstract:
|
Annual search marketing spending in the U.S. grew 28% to $7.9B in 2006 and is expected to double by 2010. This has lead to an increase in demand for search terms while posing new challenges in running a profitable campaign. Campaigns must systematically analyze large amounts of click data, adjust bid inferences across many search terms, maximize conversion and best monetize spend. At Business.com we have developed a demand-chain profit model for search marketing. We model demand elasticity using generalized linear mixed models with random cluster effects across groups of correlated search terms and repeated measures across search engine partners. Bid positions are adjusted daily using a portfolio profit and revenue optimization for more than 100,000 search terms. We discuss these techniques, information challenges and our results in managing a successful search marketing campaign.
|
