Marketing Analytics for the Transforming Media Landscape
Innovations in marketing analytics offer marketers the potential to transform their media strategy. MSI is assembling a group of forward-thinking experts — media company executives, advertisers, and data scientists from business and academia — to discuss these new methods, approaches, and challenges. Based on practice and research, they will present the latest knowledge and practical insights on these topics:
- Measuring and improving advertising response and other KPIs
- Integrating marketing mix modeling and attribution with causal methods and experimentation
- Navigating martech and advertising markets
- Delivering the right content at the right time in the right channel
- Optimizing communication budgets across media and channels
This conference is for business leaders in analytics, digital media/content, media buying, advertising analytics, and advertising research.
This event is open to members of MSI corporate sponsor companies and qualified academics. If you aren’t a member and interested in learning more, please email Earl Taylor (firstname.lastname@example.org) for more information.
|8:00 – 9:00 a.m.||Registration and Networking Breakfast|
|9:00 – 9:30||Welcome and Introductions|
|9:30 – 10:00||In Search of the Perfect Decision
Christopher Frank, Vice President Marketplace Insights, American Express Company
Information is essential to making intelligent decisions, but more often than not, it overwhelms us in today’s data-rich environment. Decisions are being made at faster rates, however these decisions are often made by consensus—reflecting a flight to safety rather than serving as a catalyst for change. The root cause are teams aiming to make decisions with minimal uncertainty or said another way – the perfect decision. This MSI presentation will discuss a fresh approach to tackling a core paradox – how do you improve the speed and confidence of decision-making in an environment of incomplete information?
|10:00 – 10:30||Making Sense of it All: How Business Leaders use Market Research and Big Data Together to Achieve Business Outcomes
Radha Subramanyam, Chief Research and Analytics Officer, CBS Corporation
|10:30 – 11:00||Break|
|11:00 – 11:30||Linking Clicks to Bricks: Spillover Benefits of Online Advertising
Vibhanshu Abhishek, University of California, Irvine
While email ads allow firms to target consumers with promotions, are they effective in increasing offline revenues for firms that predominantly sell in brick-and-mortar stores? Vibhanshu Abhishek will present research that addresses this question, using a dataset from one of the largest U.S. retailers that links each consumer’s online behaviors to the item-level purchase records in physical stores. Using a doubly robust estimator (DRE) that incorporates nonparametric machine learning methods and that enables causal estimation on observational data, the researchers find that receiving email ads can increase a consumer’s spending in physical stores. Further, this effect is heterogeneous across consumer segments. These results suggest that receiving email ads from the retailer can generate awareness and remind the consumer of the retailer’s offerings, gradually increasing the consumer’s purchase probability in the retailer’s physical stores.
|11:30 – 12:00 p.m.||Spending the Next $1 in an Omni-Channel Business
Tom Tang, Vice President Marketing Analytics, Walmart
Over the years, a variety of measurement approaches have proven to help marketers deploy their dollars more effectively across specific channels and/or strategies. With today’s ever-changing and dynamic retail shopping behavior, the need has arisen for flexible measurement solutions that can effectively evaluate strategy across stores, ecommerce, and an ever-growing number of marketing touchpoints – a challenge the team at Walmart faces on a regular basis. Tom Tang will discuss how MMM and MTA solutions can be used to set a solid omni-channel marketing strategy by complementing the strengths and weaknesses of each solution. He will talk about elevating efficiency in existing channels through the use of marginal return curves. In addition, he will discuss an approach on how to identify and test into new marketing channels. Ultimately, he will help understand how all these elements work together for a best-in-class measurement program where the next dollar is spent effectively.
|12:00 – 12:30||Mobile Advertising and Cross-device Effects
Michelle Andrews, Emory University
Should marketers advertise on mobiles even though consumers purchase on tablets and desktops? Michelle Andrews will use geo-experimental evidence to explain whether, how, and why increasing the amount of money marketers spend on mobile advertising can generate sales on larger devices. She will also explain how marketers can get around consumer- and device-level identification issues and how the type of mobile keywords marketers bid on can shape consumers’ multi-device path to purchase.
|12:30 – 2:00||Lunch|
|2:00 – 2:30||Putting Analytics into Action: Using Cupcakes to Improve Business Outcomes
Martyn Crook, Vice President Advanced Analytics & Marketing Science, The Coca-Cola Company
The best analytics in the world are useless if we don’t get the right information to the right people at the right time, and then ensure that they take action. That is challenging in any organization but trickier still in a highly matrixed global organization. Martyn Crook will talk about the new approach he is taking to improve business outcomes through the planning and execution lifecycle by combining new forms of marketing analytics and behavioral approaches to create cupcakes that are enticing and easy for the businesses to consume.
|2:30 – 3:00||Advertising and Brand Attitudes: Evidence from 575 Brands Over Five Years
Kenneth C. Wilbur, University of California, San Diego
Kenneth Wilbur will present research that investigates the relationships between three brand attitude variables (perceived quality, perceived value and recent satisfaction) and three types of advertising (national traditional, local traditional and digital). The data represent ten million brand attitude surveys and $264 billion spent on ads by 575 regular advertisers over a five-year period, approximately 37% of all ad spend measured between 2008 and 2012. Inclusion of brand/quarter fixed effects and industry/week fixed effects brings parameter estimates closer to expectations without major reductions in estimation precision. The findings indicate that (1) national traditional ads increase perceived quality, perceived value, and recent satisfaction, (2) local traditional ads increase perceived quality and perceived value, (3) digital ads increase perceived value, and (4) competitor ad effects are generally negative.
|3:00 – 3:30||Break|
|3:30 – 4:00||Inferring the Correlates of Musical Success Using a Bayesian Modeling Framework
Khaled Boughanmi, Columbia University
Khaled Boughanmi will present research, conducted with Asim Ansari and Rajeev Kohli, that uses a nonparametric Bayesian model combining online data of different modalities (e.g., metadata, acoustic and textual data) to explore the dynamics of musical success of albums over the last half century. One model component uses a supervised hierarchical Dirichlet process to summarize the perceptual information in crowd-sourced textual tags. Another time-varying component uses dynamic penalized splines to capture how different acoustical features of music have shaped album success. By analyzing how various subjective and objective acoustic measures have waxed and waned in importance over the years, Boughanmi, Ansari and Kohli uncover a number of themes that categorize albums in terms of sub-genres, consumption contexts, emotions, nostalgia and other aspects of the musical experience. Their model estimates can be used to compile new albums and playlists to appeal to listeners with different preferences and requirements.
|4:00 – 4:30||Optimizing Marketing Experiments throughout the Customer Lifecycle
Eric Schwartz, University of Michigan
Experimentation has become an important part of the marketer’s toolbox. But as “test and learn” marketing becomes part of regular operations, how should those tests best be run inside the organization to improve sustainable firm profit? Eric Schwartz will cover a range of approaches to improve “earning while learning” from marketing experiments, considering the full customer lifecycle. By leveraging machine learning approaches and modern randomized testing tools, managers can balance the desire to learn the best performing versions of marketing actions — ads, offers, product offerings, messages, website / email content, etc. — with the need to efficiently allocate resources, and deploy the best versions more widely. Attendees will learn (1) cutting edge methodological approaches for optimizing experiments, and (2) a framework for choosing the kinds of variants as well as choosing the types of decision rules for allocating resources during and after the test.
|4:30 – 4:45||Wrap Up|
|4:45 – 6:15||Networking Reception|
|8:00 – 9:00 a.m.||Registration and Networking Breakfast|
|9:00 – 9:15||Opening Remarks|
|9:15 – 9:45||Causal Measurement: Not for the casual scientist
Neha Bhargava, Manager, Advertising Research, Facebook, Inc.
Facebook has been working with the research community to advocate for causal measurement within the advertising industry for over a decade. In this presentation we will discuss the realities of implementing experiments, industry barriers and what we think the future of measurement will look like. Join Neha as she shares lessons learned and how Facebook is preparing for the future.
|9:45 – 10:15||How Online Display Ads Work: Lessons from 432 Field Experiments
Garrett Johnson, Boston University
Garrett Johnson will discuss lessons from 432 online display ad field experiments on the Google Display Network. These experiments featured 431 advertisers from varied industries, which on average included 4 million users. He will answer three fundamental questions: (1) How well do online display ads work? (2) What is the carry-over effect after the campaign ends? and (3) How does the lift at the top of the funnel translate into the lift at the bottom of the funnel?
|10:15 – 10:30||Break|
|10:30 – 11:00||The Treachery of Convenience: Redefining Reach
Kevin Whitcher, Vice President of Product Management, Oracle Data Cloud
Measurement is only as valuable as the outcomes it can change for the marketer or their business, but not enough time is spent making sure the metrics we use matter for those outcomes. Instead measurement is often a product of the metrics that are available and convenient –plus a long-standing belief that those convenient metrics drive the right outcomes. Kevin Whitcher will talk about returning to marketing’s first principles as the marketing world evolves and how that exercise can redefine the way we think about metrics and how we measure them.
|11:00 – 11:30||Competitive Advertising on Brand Search
Andrey Simonov, Columbia Business School
Andrey Simonov will discuss research that examines the effectiveness of competitive advertising on brand search using a large-scale randomized ad allocation on Bing. Results show that competitors can steal traffic from the focal brand, and they steal more (6%-20%) if they occupy the top paid position on the results page than if a focal brand’s paid link occupies the top position (1%-3%). However, the stolen clicks are of low “quality,” with 42.6% of consumers who have clicked on competitors’ links coming back to Bing in less than 30 seconds. Andrey will discuss the implications of these findings for the focal brands, competitors and search engines.
|11:30 – 12:00 p.m.||Closing Remarks|