LabScene:AdvertisingModel

Labscene is an incredibley structured platform. As such it allows for highly optimized targeted advertising to ensure that adverstisements are being presented to the most appropriate audience. There are many features of the website that can be used to aid in this process of optimization. This page should consist of specific goals and design concpets of the advertising model of Labscene.

=Goals=

Create a separate advertiser user group

 * Used by users who want to advertise on Labscene
 * The Advertisement namespace is only accessible to the Advertiser User Group (actaully may not be required)
 * Users of this group must submit Payment Information
 * Include the option to have access to advertising within the initial user sign up form
 * Create a General Advertisement Template
 * Develop a system for placing advertisements on pages
 * Ads will be required on within all namespaces at specific positions
 * Ads must be placed without the ability for the user to remove them
 * Placement of ads will depend on the options attributed to the ads (such as bid, important semantics, and desired audience)
 * Consider the layout of the pages within the different Namespaces and determine the best place for advertising
 * Ads will be as non-intrusive as possible!!
 * Ads can be liked or removed (like Facebook does)
 * Work out the ad cycling method
 * Attach advertising to data
 * Advertisements should be attached to relevant data not to pages
 * Data transfer has advertisement attachments
 * Bidding can occur for Concepts and Relations

Create a namespace Advertisement

 * Ads can be created within the Advertisement namespace
 * The Advertisement namespace only allows the page creators to edit the page
 * When the page is created a new "advertising campaign" is being created
 * Options for the campaign include the following:
 * Ad Headline
 * Ad Image
 * Ad Description
 * Ad Link (Website)

Create a namespace Campaign

 * Audience Optimization Options:
 * Location (Country/State/Province/City/Zip)
 * Age
 * Gender
 * Discipline (Field of Study)
 * Category
 * Namespaces (Research/Experiments/Studies/Labs...)
 * Investigation Keywords
 * Campaign Options:
 * Existing Campaign
 * Campaign Budget (Lifetime budget/per day budget)
 * Campaign Schedule (Timeline)
 * Objectives
 * Get More Clicks (Charged per click) - Choose cost per click (CPC)

Create a namespace Marketing

 * Includes campaigns and ads

Payment Structure
The way that advertising will be paid for is determined by the options provided through the advertiement creation form; however the actual method of gathering payment information and collecting payments must be designed and implemented. The payment information will be collected during the activation of the Advertiser User Group. This will require a new form for payment information which must be processed securely (NO SEMANTICS HERE!!!.. ie hasCreditcardnumber). Furthermore a simple program must be written that determines when the campaign is to be billed at which point a transaction should be made.

Ad Placement
Example layout for advertisements:

If you click this image it will confuse you:



Ad Auctions
Concepts
 * Ad Rank (AR) - A number used to determine the order by which ads are shown
 * Quality Score (QS) - A rating used as a measure of the quality of an advertisment
 * Max Bid (MB) - The maximum amount the advertiser is willing to pay
 * Price per click (PPC) - Amount that advertiser actually has to pay
 * Number of slots (N) - Number of places available for advertising
 * Number of Advertisements (A) - Total number of advertisements bidding for the slots

Formulas Numbers are used to distinguish between advertisers
 * (MB)*(QS) = (AR)
 * (PPC1)*(QS1)=(MB2)*(QS2)


 * Ads must be cycled based on the following
 * Required impressions
 * Desired impressions
 * Current impressions
 * Bid
 * Quality
 * User Feedback (CTR, Like/Dislike)
 * Relevance -> Uses Semantics
 * Landing Page Quality

Note: This model assumes that the appropriate unit of advertisement is a click; this means that a click is what is being sought and paid for and is the thing that has value. However, more direct value of an advertisement is the transaction that occurs as a result of the advertisement. Therefore if the pay per click model is to be used then attempts should be made to associate clicks with transactions in order to optimize the valuation of advertising opportunities.

=Type: Advertisement= hasHeadline Headline hasDescription Description hasImage Image hasStatus Status hasBudget Budget hasTimeline Timeline hasStatistics Statistics hasRange Range hasFrequency Frequency hasReach Reach hasActions Actions hasImpressions Impressions hasCPM CPM hasClicks Clicks hasCTR CTR hasSpent Spent hasBid Bid hasPrice Price

Advertising Terms Defined for Clarity:
 * CTR - Click through rate: Number of times the ad was clicked divided by the number of times the ad was shown (Clicks/Reach)
 * Clicks = Number of times the advertisment was clicked
 * Reach = Number of people who saw the advertisement
 * Impressions = Number of time the advertisment was seen (includes multiple views per person)
 * Frequency = Average number of times a person saw the ad
 * Actions = Actions taken by a person 24 hours after seeing the advertisement of 28 days after clicking the advertisement (only applies to advertisements that link to pages within Labscene)
 * Spent = Amount spent of campaign so far
 * Budget = Amount will to spend on campaign (per day/total)
 * Bid = Maximum willing to pay per click
 * Price = Average amount being paid per click
 * CPM = Cost per thousand impressions

=References=

Reference material related to advertising concepts for Labscene

Videos explaining related advertising models

 * How Facebook Advertising Works - Good overview of the Facebook Advertising Interface
 * Google PPC Model - Explains the auction process (VERY USEFUL)
 * Google Advertising - Quite general but important mention towards the end of how to determine ad placement
 * Pay per click structure - Lays out the structure the pay per click model --> Labscene will augment this model by using semantics
 * Michael Schwarz Google Talk - An hour long presentation about Generalized Second Price Auctions and how they are utilized for internet advertising (relates to his paper, "Internet Advertising and the Generalized Second Price Auction: Selling Billions of Dollars Worth of Keywords")

Websites explaining related advertising models

 * Facebook Ad Help - A very good overview of the Facebook advertising model
 * Semantic Advertising - Some of the important aspects of semantic advertising

Papers

 * Online Ad Auctions - Good paper authored by Hal Varian