Marketing Analytics

Panagiotis Angelopoulos

Persado

Persuasion Automation. Taking the profoundly human – language and persuasion – and machine generating for optimal performance

Marketers are, and have been, obsessed with communicating to the right person at the right place at the right time, using data, math and science to get there. Yet, marketers seem to leave the most crucial aspect of communication, the language itself, to chance. For the first time, data, math, and science are being applied to the generation of the most persuasive language to use in communications designed to drive action.Machine generating your marketing communications is now available, thanks to Persado. Marketers can double campaign response rates using the optimally persuasive language for digital marketing – emails, landing pages, text messages, social posts, Facebook ads, app notifications and the like. In this session you will get a glimpse into how math and language science are used to uncover the emotions and language that are most likely to drive a response from customers. The session will include case studies showcasing Fortune 500 companies leveraging persuasion automation to drive significant incremental value from their marketing activities.

Bio

Panagiotis is VP of Data Science at Persado, managing the data science department and overseeing the algorithmic research and development. Previously he was head of statistics at Upstream Systems working with major Telecommunication operators around the globe. He has been a lecturer at the University of Central Greece and has worked in many research programs funded by the EU. He is an expert in the design and analysis of experiments and in statistical modelling and machine learning.

Back to schedule

Assaf Baciu

Persado

Persuasion Automation. Taking the profoundly human – language and persuasion – and machine generating for optimal performance

Marketers are, and have been, obsessed with communicating to the right person at the right place at the right time, using data, math and science to get there. Yet, marketers seem to leave the most crucial aspect of communication, the language itself, to chance. For the first time, data, math, and science are being applied to the generation of the most persuasive language to use in communications designed to drive action.Machine generating your marketing communications is now available, thanks to Persado. Marketers can double campaign response rates using the optimally persuasive language for digital marketing – emails, landing pages, text messages, social posts, Facebook ads, app notifications and the like. In this session you will get a glimpse into how math and language science are used to uncover the emotions and language that are most likely to drive a response from customers. The session will include case studies showcasing Fortune 500 companies leveraging persuasion automation to drive significant incremental value from their marketing activities.

Bio

Assaf is SVP of Product & Engineering and co-founder of Persado. Assaf is responsible for the progression and foresight of Persado’s growing product portfolio and the management of all product advancements. Prior to joining Persado, Assaf was VP of Product for Upstream, where Persado’s core technology originated. Assaf previously worked for speech and imaging solutions supplier Nuance Communications as a senior director of product strategy where he was responsible for developing on-demand and mobile solutions. Assaf joined Nuance from BeVocal, following its acquisition in 2007.

Back to schedule

John Bates

Adobe

Bio

John Bates is the Senior Product Manager for Data Science & Predictive Marketing Solutions for the Adobe® Marketing Cloud. His core responsibility is to develop the software product roadmap for all advanced analytics and data science capabilities (i.e., including advanced statistics, data mining, predictive modeling, machine learning, text mining/natural language processing capabilities, and advanced marketing attribution and media planning solutions) found within Adobe Analytics. Before joining Adobe Product Management, Bates founded and managed the Predictive Analytics Consulting practice for Adobe Consulting – consulting with some of the world’s largest companies and brands using data mining and predictive modeling techniques in order to drive greater digital marketing success. Bates has a B.A. in Economics from Brigham Young University and a M.S. in Predictive Analytics from Northwestern University.

Back to schedule

Bob Bress

Visible World

The Future of Analytics-Driven Television Advertising

Historically, television advertising has been a blunt marketing tool.  Advertisers have generally targeted audiences by advertising in content that has high ratings for an age and gender combination appropriate to their customer base.  The measurement of ratings has historically been reported in this coarse manner due to the relatively small sample size of homes that viewership is traditionally measured on.  This leads to a tremendous amount of wasted advertising dollars as many of the viewers of a TV ad are not the intended target for the advertiser’s marketing efforts.

There are a number of key technological changes in the television industry that are fundamentally changing how TV advertising is measured, analyzed, and purchased.  The collection of data from Smart-TVs and set-top boxes has provided insight into how millions of households are viewing television.  Many of these homes can also be associated with deeper levels of segmentation beyond just the age and gender demographics historically collected.  Today, advertisers have the ability to look at viewership for detailed segments like: people who have bought specific brands of dog food, people who used their credit card heavily in the past 6 months at a particular department store, or people who have made recent campaign contributions.  This tremendous amount of viewership data provides a great opportunity to use advanced analytics and optimization methods for determining how to most efficiently deploy media dollars and allocate advertising across available television advertising inventory.  The opportunity exists for both buyers of advertising as well as sellers.  The buyers are motivated to use optimization methods to determine how to most efficiently deploy their media dollars.  Specifically, they often want to minimize their cost per targeted impression.  The inventory supplier (e.g., a network or pay-TV services provider) has to come up with a schedule of advertisements which meet all advertisers’ targeting criteria while adhering to various inventory constraints.  Additionally, inventory suppliers will look to schedule advertisements to market their own programs and products.  Networks for example must promote their own shows while cable and satellite companies may look to sell ancillary services like high speed internet or phone service.  Allocating all of these advertisements in a way that maximizes revenue can be a very challenging analytical problem.  This presentation will consider how bringing analytical rigor to these problems is shaping the future of the television advertising industry.

Specifically, in this session we will cover:

  • How the use of data from Smart TVs and Set-top boxes is changing the TV advertising industry
  • How third-party data matching enables the ability to target very specific audiences on television
  • How advertisers are using data and optimization methods to efficiently develop media plans to target specific audiences
  • How advertising inventory sellers can use optimization methods to efficiently allocate many advertisers across a schedule to maximize revenue
  • How progress is being made to measure the effectiveness of these analytics-driven media campaigns

Bio

Bob Bress is Senior Director of Product Management and Analytics at Visible World. In that role he has applied his expertise in data and analytics to lead development of the next generation of advanced targeted advertising products for television. Bob has over 15 years of business analytics experience including work in the energy industry, providing advanced analytics services for innovative demand-side energy programs and at GE’s Global Research Center in the Applied Statistics Lab where he worked on cutting edge statistical applications for a variety of GE businesses. Bob holds undergraduate and graduate degrees in Industrial Engineering and Operations Research & Statistics from Rensselaer Polytechnic Institute.

Back to schedule

Fahner

Gerald Fahner

FICO

Combating Attrition Through New Developments in Transaction Analytics and Customer Dialogue

“Silent” attrition, i.e. indefinite lapses of customer purchases without contract cancellation, remains a costly problem. Marketers require fast detection and insight into the nature of attrition to create effective retention programs. Our case study with credit card transactions illustrates how tree ensemble models, instrumented with low-latency transaction features, rapidly and reliably detect card-level and merchant category-level attrition. Although sometimes described as “black boxes”, these models can be inspected: We explain the function of our models, which capture strong and interpretable interactions, and relate their predictive performance to profitability. We demonstrate that complex models trained on large data volumes are far more profitable than simpler models for this problem.

We also discuss how marketers can boost persuasiveness of offers (another key driver of profit) by engaging customers in dialogues to learn about their preferences before customizing offers. Using a Bayesian framework and simulations we illustrate the value of customer dialogue.

Bio

Dr. Gerald Fahner is Senior Director in FICO’s Research division where he leads the Predictive Technologies Group. He specializes on innovative algorithms that turn data and domain knowledge into superior predictions and decisions. Gerald is also responsible for the core algorithms underlying FICO’s Scorecard development platform, and used to develop the FICO Score. His work on causal modelling won the Best Paper award at the Credit Scoring and Credit Control XI conference. Prior to joining FICO in 1996, he researched machine learning and robotics at the International Computer Science Institute in Berkeley, and earned his Computer Science doctorate from University of Bonn.

Back to schedule

Bill Harvey

Bill Harvey Consulting

Marketing Increasing Its Profit Contribution through Big Data Analytics

Now there are impressive results to show for analytics in marketing. Marketers since 2006 have been matching household records across diverse databases with identity protection. This one innovation led to a more general buzzword “Big Data”. From results summarizing dozens of case studies, Key Takeaways will include:

  • How brands that have stopped growing can re-start sales and profit growth by identifying geographic areas of growth opportunity with predictive analytics.
  • The last bastion holding out from analytics – Hollywood content – and how a new form of content analytics is finally bringing an analytic perspective to this auteur seat-of-the-pants business.
  • What data and analytics work, when.
  • Which ones don’t in specified situations.
  • Typical marketing questions, general approach to each question, typical results.

Bio

Bill Harvey is a well-known media researcher and inventor who co-founded TRA, Inc. and is its Strategic Advisor. His nonprofit Human Effectiveness Institute runs his weekly blog on consciousness optimization. Bill can be contacted at bill@billharveyconsulting.com

Back to schedule

Loree Lash-Valencia

WGSN INstock

Better Decisions though Data – A Retail Challenge

The retail industry is moving faster than ever before. As retailers become more agile, adding new items and discounting others, it’s critical to have the right merchandising strategy and the ability to respond to changing conditions. Understanding the competitive landscape in regards to assortment and pricing is critical to strategic and tactical decision making.

This talk will outline:

  • The challenge retailers face today to improve sell through, understand competitive pricing, maintain margins, and avoid terminal stock
  • The problems they currently have collecting actionable data to face those challenges
  • A solution to this problem that captures this data and presents it using visualizations that make it actionable for the business.

Bio: 

With senior roles in the retail and technology industries, including YESIMO and MyShape, Loree began her career as a designer and merchant then progressed into e-commerce and the fashion tech space. She is one of the USA’s leading figures in brand building and fashion retail.

Back to schedule

Daniel McDuff

Affectiva

Large-Scale Emotion Analytics and Predicting Consumer Behavior

Emotions play a huge role in our everyday lives. They influence memory, decision-making and well-being. Using new tools in computer vision and machine learning we are able, with consent, to quantify emotional responses more accurately and on much larger scales than has been possible in the past. In this talk, I will present the state-of-the-art in affective technology and insights from analysis of the worlds largest dataset (1B+ data points) of consumers’ emotional responses. The talk will include a number of case studies, from how emotions reveal political preferences to how they can be used to predict a viral video.

Bio

Daniel McDuff completed his PhD in the Affective Computing Group at the MIT Media Lab in 2014. He is building and utilizing scalable computer vision and machine learning tools to enable the automated recognition and analysis of emotions and physiology. His work has received nominations and awards from Popular Science magazine as one of the top inventions in 2011, South-by-South-West Interactive (SXSWi), The Webby Awards, ESOMAR and the Center for Integrated Medicine and Innovative Technology (CIMIT). His work has been reported in many publications including The Times, the New York Times, The Wall Street Journal, BBC News, New Scientist and Forbes magazine.

Back to schedule

Sanjog Misra

UCLA

Attribution Done Right.

Multi-Touch Attribution (MTA) uses individual level marketing exposure and outcome data to ascertain the impact that various marketing investments have on consumer outcomes (e.g. sales) and ascribe credit to each investment at the most granular level possible. In this talk I will introduce cutting edge methods related to MTA and talk about challenges and the efforts being undertaken to address these.   The talk will outline the basic ideas of MTA, the methodology behind it, some implementation details and firm outcomes using real world cases.

Bio

Sanjog Misra is Professor of Marketing at UCLA Anderson School of Management. Professor Misra’s current research involves the use of structural econometric methods to study consumer and firm decisions. His particular interests involve understanding firm decisions pertaining to pricing, distribution and salesforce management issues and how consumers make discrete choices.

Professor Misra is currently the co-editor of the Quantitative Marketing and Economics journal and has served as an Associate Editor at Marketing Science and International Journal of Research in marketing. He also serves on the editorial board of the Journal of Marketing Research and has acted as Associate Editor for special issues of Management Science and the Journal of Marketing Research.

Professor Misra is actively involved in partnering with firms in his research and has worked on various projects with companies such as Eli Lilly, Adventis, Mercer Consulting, Sprint, MGM, Bausch & Lomb, Xerox Corporation and Lucent Technologies with the aim of helping them design efficient, data based, management systems that result in better decisions. Professor Misra currently chairs the research committee at Convertro (a division of AOL) and is an Advisor to AOL Platforms on methodology related issues.

Prior to joining UCLA Anderson Professor Misra was Associate Professor at the Simon School of Business at the University of Rochester and has taught (as visiting faculty) at the Johnson School of Management at Cornell University and the Graduate School of Business at Stanford University.

Back to schedule

Daniel Porter

BlueLabs

Pinpointing the Persuadables: Convincing the Right Customers and the Right Voters

Marketing, political campaigning, and healthcare have one major thing in common: millions of per-person treatment decisions must be selected in order to drive positive outcomes. Prior to President Obama’s reelection campaign, standard practices for persuading voters—that is, changing their minds—were unscientific and driven by long-standing assumptions and hunches. This mirrors outreach efforts by other companies and organizations, which know that a certain percentage of their marketing efforts will inevitably be wasted on people who are not going to be receptive to it. Daniel Porter of BlueLabs, who served as the Director of Statistical Modeling for the Obama Campaign, will discuss his experience using the results from a large-scale randomized, controlled experiment to target persuadable voters for the Obama Campaign, as well as ways these cutting edge statistical modeling techniques can be applied to influencing behavior in realms ranging from health outcomes to customer retention. Attendees will walk away with greater insight into the capacity predictive modeling to identify individuals likely to be persuaded, and the necessary ingredients for experimental testing.

Bio

Daniel Porter is a cofounder and head of Data Science at BlueLabs, a Washington DC based analytics, data and technology company formed by the lead practitioners from the Obama 2012 Analytics Team. At BlueLabs, Daniel’s team works to develop predictive analytics solutions across a variety of sectors, including Education, Healthcare, Political Campaigns and Consumer Marketing. The goal of his team’s work is to use data science to help clients across all these sectors expend their resources as efficiently as possible by measuring the actual ROI on any expenditure.

On the Obama campaign, Daniel was Director of Statistical Modeling, where he and his team of 8 data scientists developed individual level statistical models that were used throughout the campaign for fundraising, media buying and state strategy.
Daniel holds a BA from Wesleyan (CT) University and and MA in Quantitative Methods from Columbia University.

 

Back to schedule

Pathik Tanna

WGSN INstock

Better Decisions through Data – A Retail Challenge

The retail industry is moving faster than ever before. As retailers become more agile, adding new items and discounting others, it’s critical to have the right merchandising strategy and the ability to respond to changing conditions. Understanding the competitive landscape in regards to assortment and pricing is critical to strategic and tactical decision making.

This talk will outline:

  • The challenge retailers face today to improve sell through, understand competitive pricing, maintain margins, and avoid terminal stock
  • The problems they currently have collecting actionable data to face those challenges
  • A solution to this problem that captures this data and presents it using visualizations that make it actionable for the business.

Bio: 

Coming with a background in Finance and Computer Science, Pathik has worked in various industries including Telecommunications, IT/E-Commerce and currently in the Fashion Industry, as the Director of Software Development. Pathik leads a team of various skilled software engineers, developing creative technology solutions for the Fashion, Interiors and Lifestyle Industries.

Back to schedule