MARKETING EFFECTIVENESS ACCELERATOR (MEA) : 2026 (formerly Attribution & Analytics Accelerator)

the EARLY MOVER rate ends on Oct. 6

OVERVIEW

The Marketing Effectiveness Accelerator is the only event focused exclusively on attribution, marketing mix models, AI, and the science of marketing performance measurement. Top minds will present groundbreaking insights and empirically grounded case studies. The audience consists of strategic marketers seeking evidence-based solutions.

The 2026 Marketing Effectiveness Accelerator will be held on November 10 at People Inc., 225 Liberty St., New York, NY, 10281.

This gathering of modelers, marketers, researchers, and data scientists will accelerate innovation, deepen scientific understanding, and drive the industry toward best practices and better solutions. Here, B2C and B2B marketers from different sectors share their experiences and fresh perspectives, promote practical ideas, and explore innovative techniques and applications.

The Marketing Effectiveness Accelerator is the only event dedicated exclusively to attribution, marketing mix models, and the science of marketing performance measurement. On November 12, leading experts presented empirically grounded case studies that demonstrate how leading brands are solving today’s toughest challenges.

Marketing Effectiveness Accelerator - Call for Content Now Open Through May 30

We are considering work for the 2026 event and invite case studies, experimental results, or empirically validated best practices that can be linked to business success. Each submission must be carefully developed from the end-user perspective—no sales pitches. Only proposals that include participation from brands, end-users, academics, or a neutral third party will be accepted.

Topics include:

AI-Driven Modeling: Promise and Practicality
Data Complexity & Identity Resolution
New Methods and Data Sources in Marketing Performance Measurement
Impact of Changes on Consumer Behavior and Tools for Media Activation
Branding: Macroeconomics, Attribution, and Predictive Analytics

See full topic descriptions here.

What Makes a Strong Submission:

Showcasing how insights directly shaped strategic marketing decisions
Transparent methodology
Evidence of validation or triangulation
Acknowledgment of limitations, trade-offs, failed hypotheses, and lessons learned
Generalizable insights or learnings across verticals

Enter your work through the submission portal before May 30.

Sequent Accelerator Award - OPEN FOR ENTRIES THROUGH MAY 30

The Sequent Accelerator Award celebrates technical advances in key aspects of marketing analytics – spanning the modeling processes, metrics, data acquisition, AI applications, insight dissemination and organizational adoption.

In keeping with the spirit of the Marketing Effectiveness Accelerator (The Accelerator), this Award recognizes innovative solutions that deliver meaningful brand impact and advance the discipline of marketing analytics. It celebrates approaches that inspire progress and accelerate measurable effectiveness on brands and across the industry.

All marketing performance analytics innovations from the past five years are eligible.

The winner will be selected based on the strength of their innovative solution, proof of adoption, and impact of the new analytic approach, and will be recognized at the Marketing Effectiveness Accelerator event on November 10.

Download the award template here and send in your nomination through the portal by May 30.

For any questions about content submissions or award nominations, please contact Events Director Sara Serpe at sara@thearf.org.

Speakers

Agenda

8:00 – 9:00am	Breakfast & Registration
9:00 – 9:10am	Opening Remarks Scott McDonald, Ph.D. – President & CEO, ARF
	New Methods and Data Sources in Marketing Performance Measurement
	Sessions
9:10 – 9:30am	From Cult Favorite to Category Contender: The Measurement Story Behind Duke’s National Launch Duke’s Mayo has built one of the most compelling growth stories in CPG: a beloved regional brand targeting 25% national household penetration by 2029, up from under 15% today. Getting there requires more than media spend. It requires knowing, with real confidence, what the spend is actually doing. Brainlabs and Duke’s built a measurement ecosystem that triangulates methods – MMM and matched-market testing via synthetic controls – to answer a question the industry still struggles with: what media actually causes versus what it merely correlates with? This case shows how calibration between modeling and experimentation works in practice, how a growing CPG brand uses monthly MMM updates to close the loop between forecast and result, and what happens when you treat measurement as a strategic asset rather than a reporting obligation. Andy Littlewood– Chief Planning and Data Officer, Brainlabs Rebecca Lupesco – Director of Marketing, Dukes, Mayo
9:30 – 9:50am	AI-Powered Synthetic Control for Incrementality in High-Reach Campaigns IQVIA has transformed how healthcare brands measure the true impact of high-reach media campaigns. Campaigns now span linear TV, Connected TV, and digital channels, often reaching 90%+ of measurable audiences. This leaves too few unexposed patients to form reliable control groups, making traditional test-versus-control measurement biased and limiting accurate cross-channel incrementality reads. IQVIA’s patent-pending AI-driven Synthetic Control methodology solves this by using a tailored diffusion model to generate high-fidelity, privacy-safe synthetic patient profiles that augment limited real controls, correct bias, and enable unbiased 1:1 matching. Across brands, it reduced bias by 67% and cut measurement error by 76% revealing impact that traditional methods missed and helping brands optimize the media mix, reduce waste, and make better decisions. Ronnie Choudhary – Group Director, Marketing and Customer Experience Analytics, Oncology, AstraZeneca Neelam Hinduja – Director, Product Management, IQVIA Yanping Liu – AI Scientist Director, IQVIA
9:50 – 10:10am	InstaPoll: Simulating Human Decision-Making at Scale with Synthetic Populations and LLMs While LLMs have demonstrated remarkable reasoning capabilities, accurately simulating complex human decision-making remains an open challenge. Modeling authentic survey responses is notoriously difficult because of evolving behavioral landscapes and high biases in source data. This talk introduces a theory-driven, multi-step reasoning framework that addresses this by integrating Arima’s Synthetic Society with LLMs. How do we generate synthetic survey respondents at scale by calibrating personas to real-world data and incorporating temporal factors? See a robust performance analysis comparing Arima’s synthetic framework with parallel human surveys and real-world outcomes, demonstrating significant improvements in simulation accuracy and bias mitigation. Learn technical insights into the promise, architecture, and limitations of this approach. T.S. Kelly – Managing Director, US, Arima Winston Li – Founder, Arima
10:10 – 10:30am	Small Variables, Major Impact: Insights From a Brand Lift Meta-Analysis The measurement team at People Inc. is continually advancing the quality and accuracy of measurement to help clients understand real consumer actions and outcomes. In this session, they’ll share insights from more than a decade of brand lift data, showing how seemingly small campaign and business decisions can drive meaningful shifts in performance.  They’ll explore key factors—including KPI selection, methodologies, and the ways individual KPIs can influence one another within a single study—and demonstrate their impact on results. Neil Napolitano – Executive Director, Measurement Innovation, People Inc. Mike Lichter– Executive Director, Measurement Innovation, People Inc.
10:30-11:00am	Morning Break
11:00 -11:20am	Closing the MMM Actionability Gap: How Google Meridian Turns Insights into Decisions Marketing mix modeling adoption is accelerating, but implementation alone doesn’t guarantee results. A 2026 Harvard Business Review Analytic Services study of 547 marketing leaders found that 87% of organizations consider MMM important for data-driven decision-making, yet only 28% report effectively translating model outputs into timely action. This session examines the MMM actionability gap: the structural, organizational, and operational barriers that keep insights from reaching the people and processes that can act on them. Drawing on Google’s HBR research and Adswerve’s hands-on Meridian implementations, this session offers a practitioner’s view of what it takes to move from model to media plan, with a real advertiser sharing how MMM outputs are now embedded in their media planning and business processes. Stephen Mangan – Head of X-Media Outcomes Measurement, Google Dani Brandtjen – Senior Director of Sales Engineering, Adswerve
11:20am – 12:00pm	Reconciling MTA and Incrementality: A Unified Framework for Diagnosing Attribution Bias in Digital Advertising Multi-touch attribution (MTA) and causal incrementality often yield conflicting channel effectiveness estimates, leaving advertisers uncertain about budget allocation. This research introduces a diagnostic framework that explains WHY methods diverge and WHEN each is reliable. Using simulations with known ground truth, we show that MTA systematically over-attributes to demand-correlated channels (e.g., search), while causal methods under-estimate channels with long carryover (e.g., video). The framework is being piloted with Amazon Ads’ largest advertisers. Rather than choosing one method, we show that the gap between estimates is itself diagnostic, identifying where experiments are needed. This transforms attribution from a “which number is right” debate into an actionable measurement strategy. Vivian Qin – Principal of Analytics, Amazon Ads Srinivas Osuri – Associate Principal, Amazon Ads
12:00 – 12:20pm	Why Measuring Real-World Business Impact Is Harder Than It Looks Some of the hardest measurement problems are not technical – they are operational. In enterprise environments, systems rarely roll out under clean experimental conditions for long. Teams adapt workflows, business priorities shift, overlapping initiatives emerge, and strong early KPI movement can quickly make causality difficult to interpret. This session discusses practical lessons from measuring impact in real-world environments where clean holdouts, stable baselines, and simple before/after comparisons often break down. The focus is on how organizations can combine experimentation, observational methods, and measurement governance to make more reliable decisions under imperfect conditions. Aman Tripathi – Data Science Manager, Experimentation, Hyatt Hotels Corporation
12:20 – 1:20pm	Lunch
	Measurement & Modeling for Marketing and Product Growth
1:20 – 1:40pm	Attentive Reach: Integrated Actionable Measurement Beyond Reach and Frequency Not all impressions are created equal. Today’s planners, buyers, and sellers need tools that reflect both the quality and quantity of media to ensure advertiser dollars are valued appropriately. Learn about attentive reach, which provides that foundation by creating a shared understanding of quality and scale and delivering an effective, outcome-driven view of campaign impact. The research enables advertisers and publishers to apply insights for strategic planning, mid-flight course correction, and post-campaign evaluation, ensuring every dollar spent maximizes impact. This research is a collaboration between Nielsen and Adelaide, focused on creating metrics that combine Adelaide’s AU attention scoring with Nielsen currency reach and frequency metrics to provide integrated, actionable measurement. Claire Browne — VP, Research, Adelaide Metrics Shweta Shah — VP, Data Science, Nielsen
1:40 – 2:00pm	The Missing Variable in MMM – Measuring the ROI of Creative In this session, learn about three key points: 1) Why it’s critical to measure creative within MMM; 2) How to measure creative; and 3) A practical example: a client MMM use case showing the ROI of creative. Amritansh Sahai — Manager, Marketing Data Analytics, Lenovo Garth Viegas – Measurement Architect, Analytic Edge
2:00 – 2:20pm	Conversation to Conversions: How Data-Driven Measurement Moves the Needle for M&E This session combines a visibility analysis of select M&E advertisers with a meta-analysis of entertainment tune-in studies from Samba to explore the relationship among paid advertising, community conversation, and viewership. Learn general best practices for advanced analytical solutions while uncovering specific insights from first- and third-party engagements. Together, these studies offer a new framework for understanding how conversations may influence what audiences ultimately choose to watch. Rob Gaige – Senior Director of Global Insights, Reddit Alyson Sprague – VP of Measurement Science, Samba Liz Strait – Senior Measurement Researcher, Reddit
2:20 – 2:40pm	Reconciling MTA and Incrementality: A Unified Framework for Diagnosing Attribution Bias in Digital Advertising Multi-touch attribution (MTA) and causal incrementality often yield conflicting channel effectiveness estimates, leaving advertisers uncertain about budget allocation. This research introduces a diagnostic framework that explains WHY methods diverge and WHEN each is reliable. Using simulations with known ground truth, we show that MTA systematically over-attributes to demand-correlated channels (e.g., search), while causal methods under–estimate channels with long carryover (e.g., video). The framework is being piloted with Amazon Ads’ largest advertisers. Rather than choosing one method, we show that the gap between estimates is itself diagnostic, identifying where experiments are needed. This transforms attribution from a “which number is right” debate into an actionable measurement strategy. Vivian Qin – Principal of Analytics, Amazon Ads  Srinivas Osuri – Associate Principal, Amazon Ads
2:40–3:10pm	Afternoon Break
3:10 – 3:30pm	Beyond the S-Curve: A Multi-Seasonal Hierarchical Diffusion Framework for New Product Launch Measurement New product launches are often modeled with a single S-curve to capture the slow-start, rapid-growth, and saturation phases of adoption. In practice, this approach fails when campaigns start months after launch, ad spend varies across advertisers, and categories such as fashion exhibit strong seasonality, trends, and retro-revival cycles. We extend the classic sigmoid diffusion framework with (1) a piecewise pre/post-campaign formulation, (2) a Bass-style decomposition that separates paid-media drivers from organic momentum, (3) a hierarchical logistic specification that borrows strength across sparse brands, and (4) multi-seasonal components that model annual, trend, and retro-cycle effects. The result is a category-aware measurement of launch performance. Vivian Qin – Principal, Amazon Ads Gabe Fishbein – Director of Product for Ad Tech, Gopuff
3:30 – 3:50pm	Measuring Dynamic Product Complementarity and Substitutability with Temporal Graph Learning AI shopping assistants increasingly need to understand not only what consumers buy but also how products relate to one another in real time. This research introduces a scalable temporal graph-learning framework that dynamically distinguishes product substitutes from complements using streaming basket data. Unlike traditional recommendation systems that collapse all product relationships into a single similarity score, our approach separates complementarity and substitutability into interpretable relationship spaces. Using over 20 million grocery co-purchase events from Instacart, the model identifies evolving product relationships with strong predictive accuracy while preserving managerial interpretability. The framework enables AI-driven merchandising, cross-selling, substitution planning, and intelligent shopping assistance at the SKU level. Arvind Rangaswamy – University Distinguished Professor of Marketing, Penn State University Wenlan Yu – Ph.D. Student, Penn State University
	Impact of Changes on Consumer Behavior and Tools for Media Activation
3:50 – 4:10pm	Plausibility Is Not Validation: When Is Synthetic Data Decision-Ready? Synthetic respondents, digital twins, and AI-generated simulations can generate outputs that appear credible and internally consistent. But plausible isn’t the same as valid, reliable, or fitting for a particular decision, and not every use requires the same level of validation. Drawing on research and industry experience across MSI and ARF, this session introduces a practical approach that starts with the decision itself. Attendees will learn how to match validation rigor to decision risk, and how to determine when synthetic results should be benchmarked against relevant human, behavioral, historical, or holdout data before they can be trusted. Keith Smith, Ph.D. – Managing Director, MSI
4:10 – 4:30pm	Leveraging MMM Response Curves to Guide Global Decentralized Decision-Making There are surprisingly many ways people currently use MMM response curves to determine optimal spend. For a single brand, despite the pitfalls, most marketers use MMM to make reasonable spend decisions. When using MMM to guide the total corporate marketing budget, particularly when some spending decisions are decentralized in local markets, it becomes important to have good rules of thumb. We discuss various rules of thumb for using response curves, including one we call “Efficient Spend,” which we feel helps both global and local market decision makers leverage MMM response curves to optimize profit and growth. Igor Levin – Senior Director, Global Media Analytics and Data Science, Kenvue Ross Link – CEO, Marketing Attribution LLC
4:30 – 4:50pm	Sequent Accelerator Award The Sequent Accelerator Award honors breakthrough innovations in marketing analytics, including advancements in AI, modeling, metrics, data, and insight delivery, that have driven measurable brand impact and advanced the field within the past five years. The winner will be announced soon.
4:50 – 5:00pm	Closing Remarks Scott McDonald, Ph.D. – President & CEO, ARF
5:00 – 6:30pm	Cocktail Reception

2026 Advisory Committee

We express our gratitude to our Advisory Committee members. This esteemed group of practitioners and academics combined their collective knowledge and curiosity to generate the topics for this year’s MARKETING EFFECTIVENESS ACCELERATOR (MEA) Call for Content.