2024零售媒体中的增量测量与数据协作白皮书

Retail Media Advanced Measurementand Data Collaboration Table of Contents Table of Contents2What is Incrementality3Understanding Advanced Measurement Techniques9Randomized Controlled Trials (RCTs)11Match-Market Market16Machine Learning Counterfactual Models19Synthetic Control Method (Technique)21Shadow-Mode Testing (Technique)22Causal Models23Media Mix Modeling (MMM) in Retail Media25 The Role of Data Collaboration in Advanced Measurement27Challenges in Measuring Incrementality in Retail Media30Steps for Measuring Incrementality in Retail Media34A Path Forward with LiveRamp36Collaborating on Common Standards36Understanding Incrementality with Pacvue37Acknowledgement38Contributors38About IAB40Contact Information41 What is Incrementality Incrementality is not a new concept to marketers; it has long been used as a way tomeasure the impact of marketing efforts. However, it has become increasinglyprominent in retail media due to the growing need for precise and accountablemarketing strategies. Retail media's closed-loop measurement capabilities enhance theimportance of incrementality by allowing for direct tracking of the consumer journey fromexposure to purchase. However, the walled garden nature and questions aboutcomplaint data usage in retail media channels poses significant challenges. Thisconfinement makes it difficult to achieve accurate cross-channel attribution, as datawithin these channels is often siloed. As a result, a comprehensive view ofincrementality measurement necessitates advanced modeling techniques to bridgethese gaps and provide a more complete understanding of marketing effectivenessacross different channels. “Incrementality measurement can be accomplished in various ways, andthespecific methodology you choose is dictated by the advertisingenvironmentand data you have available to you,”said ChristineGrammier, VP of Global Measurement Products at LiveRamp. “Dataclean rooms provide a privacy-compliant way to gain access to higherquality exposure data from publishers and audiences and conversionsfrom retailers.” TheIAB/MRC Retail Media Measurement Guidelinesdefine incrementality as thepotential causal impact of marketing. Specifically, for retail media organizations, itrepresents the causal impact of marketing and is often linked to outcomes like sales orattributed to advertising campaigns or exposures. In other words, incremental sales arethe sales you would not have gotten without showing the ads In retail media measurement, there are several operational levers that go beyondtraditional testing methods to distinguish incrementality. These levers are used to isolatefactors such as seasonality or natural sales growth within the analysis to separate theoutcomes from "organic" v. "ad-attributed" efforts.. Without clearly separating theseoutcomes, there’s a risk of overestimating the effectiveness of marketing campaigns bymistakenly attributing organic sales to paid media efforts. By incorporating thisdistinction, incrementality models provide a more accurate evaluation of how marketinginvestments contribute to additional business outcomes, beyond what would haveoccurred naturally. This precision is essential for optimizing marketing strategies andensuring that resources are effectively driving true incremental value. These models are crucial for understanding the true effectiveness of retail mediastrategies, distinguishing between correlation and causation, and ensuring thatmarketing spend contributes to tangible business outcomes. Selecting the right modelrequires a nuanced understanding of various advanced machine learning approaches,their challenges, limitations, and opportunities, as well as considerations around datacadence, quality, and sources. This paper aims to provide an overview of incrementalitymodels in the context of retail media performance measurement, focusing on theintricacies of leveraging advanced machine learning techniques. “Ourapproach to incrementality is collaborative and flexible.Weunderstand that investing in incremental measurement tools can powergreatoutcomes for both retailers and brands,”said Nicolas Pirog,Director,Measurement Strategy,Analytics at Roundel.“We alsobelieve in having more than one way to approach incrementality andmatching the tool to the use case.” Brands and agencies are increasingly demanding transparency and justification of retailmedia investment by proving out performance. Incrementality provides the rigor neededto prove the value of marketing initiatives and make informed decisions about budgetallocations. “There’s a lot of myth and mystery surrounding incrementality, but let’s clear thefog: large CPG brands have been mastering incrementality measurement foryears, proving that measuring for significance is achievable when done right. Acombinatory approach that leverages geo and ID-based measurement isespecially effective for omni-channel campaigns. Key elements for successinclude accommodating baseline sales, aligning media weight