Browse Prior Art Database

A Method and System for Evaluating Effectiveness of a Demand Side Platform

IP.com Disclosure Number: IPCOM000239808D
Publication Date: 2014-Dec-03
Document File: 7 page(s) / 300K

Publishing Venue

The IP.com Prior Art Database

Related People

Jian Xu: INVENTOR [+3]

Abstract

A method and system is disclosed for evaluating effectiveness of a Demand Side Platform (DSP). The method and system evaluates the effectiveness of a DSP wherein novel metrics are defined for normalized Return of Investment (ROI) to measure platform performance. The method and system also describes a high-level design of an online A/B test framework as well as a terabyte scale dashboard reporting system for Yahoo Ad Manager Plus* (YAM+). The proposed method and system addresses both the platform level and line level performance measurement problems.

This text was extracted from a Microsoft Word document.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 26% of the total text.

A Method and System for Evaluating Effectiveness of a Demand Side Platform

Abstract

A method and system is disclosed for evaluating effectiveness of a Demand Side Platform (DSP).  The method and system evaluates the effectiveness of a DSP wherein novel metrics are defined for normalized Return of Investment (ROI) to measure platform performance.  The method and system also describes a high-level design of an online A/B test framework as well as a terabyte scale dashboard reporting system for Yahoo Ad Manager Plus* (YAM+).  The proposed method and system addresses both the platform level and line level performance measurement problems.

Description

Disclosed is a method and system for evaluating effectiveness of a Demand Side Platform (DSP).

The method and system proposes a novel metric normalized ROI (nROI) to measure the platform performance, given as,

 

Here, the nROI quantifies the capability to make profit per budget per advertisement (ad) request.  It is also assumed that click is the desired response from user.  The method and system is also applicable to campaigns with action as the desired user response.  Since, the marketplace is so diverse, the method and system introduces the rationale behind this metric in a case by case fashion.

For CPM campaigns, the advertisers pay a fixed amount of money for each impression delivered and there is typically an enhanced Cost per Click (eCPC) goal.  So for such type of campaigns, advertisers take the risk.  Therefore, the nROI is defined from the advertiser’s perspective as:

For Cost per Click (CPC) campaigns, the advertisers pay a fixed amount of money for each click received.  For such type of campaigns, the DSP takes the risk so the nROI is defined from the DSP's perspective as:

For dynamic Cost Per Mille (dCPM) campaigns, the advertisers pay a dynamic amount of money for each impression delivered.  A typical dCPM setting is cost-plus in which the advertisers are charged a markup of the inventory cost.  For such type of campaigns, the advertisers take the risk, so the nROI is defined from the advertisers' perspective as:

At the platform level, it is assumed that the return, cost, and budget of the i-th campaign is returni, costi, and budgeti respectively.

It is noted that nROI takes care of all the three aspects to evaluate a display ad campaign.  The three aspects are delivery, performance and cost.  The platform nROI actually quantifies how effective the platform can make profit for advertisers and/or platform given the amount of total budget that is managed and the total ad requests received.

The method and system focuses on the challenges that are posed by the Real Time Bidding (RTB) marketplace when conducting online A/B tests.  Although global model is easy to implement and/or deploy, the method and system advocates for line specific model for two reasons, the ad campaigns within a DSP are so diverse that a global model can hardly be always the optimal one for every line....