Sponsored Search Engineering in Electronic Markets: Is It Worth It?
An Empirical Analysis of Search Engine Advertising: Sponsored Search in Electronic Markets
Sponsored search advertising which has grown increasingly popular in all industries over the past decade gives insight into user interface interaction and activity such as click-rates. The search advertising market is presumed to amass 37% compound annual growth rate, more than $33 billion in 2010.
The purpose of search engine advertising is to shift the marketing campaign to a targeted campaign. For example, specific keywords results in the displaying of ads which contains user-generated content (product reviews, product features, values, likes/dislikes). This allows the advertisements to reach a niche audience. Conceptually, advertisers, who represent individual firms, submit their product and services information with chosen keywords for that advertisement to be associated with in the search engine.
In this research article, a Hierarchical Bayesian modeling framework is proposed as it allows for the most detailed representation of search engine advertising. It combines and organizes joint probabilities based on individual clicks and purchase propensities. This relates to our course material the significance of calculating click-through-rate, value-per-click, and clicks-per-hour. These are individual unit of measures used to track customer behavior and also predict questions like: How much will they purchase? Will they purchase? At what hour?
Moreover, the modeling framework allows us to compare and weigh which words and language is most effective for proper sponsored-search advertising. This is executed through a ranking process by the model where it weighs the effect of advertising using different keywords by taking into account the CPC (cost per click) and ‘Quality Score.’ There seems to be a high amount of heterogeneity for each word’s profitability, which implies that most search engines have certain top keywords. Moreover, retailer-specific and brand-specific keywords can falter the results as certain customers may search specific words whereas other search generic words. However, the model does show that there is a considerable difference when the parameters (consumers, advertisers, browsers, search engines) are varied to compare different possibilities.
Furthermore, this paper lacks the consideration of competition in the sponsored-search process. Our data does not have precise information on market competition. Firm model comparison is one of the most influential factors for high return rates as it signifies the most profitable features. Product-specific characteristics and reviews shows us how different products affect click-through and conversion rates. This allows firms to identify the brands with the highest conversion rates and lowest costs per conversion.