Conference Schedule

When optimization is framed as a sequential decision problem, it can unify testing, targeting, and even attribution, as interrelated subtasks. Learning algorithms are designed for a real time, high-performance environment – like marketing – and feed off a continuous stream of user interaction and feedback. Matt introduces Reinforcement Learning as a general approach to solve multi-step, multi-touch problems.

Growing a team is a delicate dance between the work load, the budget and a continuing lack of unicorns. Should you hire the best and brightest and blow the budget? Should you cross-train people already in your organization? How about hiring fresh grads and training them from the ground up? With a combined 38 years of experience, these digital analytics professionals have been responsible for a wide variety of teams. Between mad technical skilz, project management wisdom, and superior political talent, building and running an analytics team requires walking that razor’s edge of being a supportive confidant and a slave driver. Rusty, Amy and Tim share their best advice on making sure you can get the best out of the people you have and how to recognize when you’re missing something.

Machine learning (ML) is a rigorous procedure in statistical science that helps data scientists choose the best model to fit a data set. Michael provides an operational understanding of what ML really entails and thoroughly examines the esoteric field of feature engineering. Feature engineering is unlike other aspects of data science because it is not learned from disciplines like math, statistics and computer science. Instead, it is learned on the job and requires substantial domain knowledge of marketing in a given business. Therefore, feature engineering knowledge is typically not transferable across business domains and it is the most valuable and least replicable part of any data scientists’ work. Companies that understands feature engineering are those who can really turn their big data into business results like magic.

Knowing the difference between a 301 and a 302 redirect, and why it matters to your SEO team will make you an analytics star. Knowledge of Broad Match, Broad Match Modified, and Phrase Match, and why it matters to your Paid Search team, as well as what header bidding is, how your data can change for sites using header bidding, and why your display team shakes their head in frustration when they speak of it will make you a trusted member of the greater marketing team. Andrew delivers knowledge on these channel-specific marketing ideas, as well as how your increased fluency of these and other aspects of search will make your data shine, and make you personally indispensable to your organization. He also shares best practices from nearly a decade of agency experience, and how he trains his staff in both the breadth and depth of agency marketing.

While these other data sources do have some value, behavioral data is at the top of the predictive list. Peter shows how to can use behavioral data to find your best customers and market to them across multiple channels, decreasing wasted marketing spend and increasing profits. Using real-world examples and talking about perspectives and methods that go against the grain of traditional product-centric marketing wisdom, Dr. Fader provides insights to help marketers find and create more value.

Everyone talks about the importance or marketing measurement, but is Multi-Touch Attribution (MTA) really the answer that we have all been waiting for, or too good to be true? This session will demystify this extremely complicated area and speak practically as to how a marketer can utilize multi-touch attribution. Understand how to define MTA, find out why it is so damn hard, and learn what do other marketers and retailers tell us about how they are using it. Attendees will walk away with the confidence and knowledge they need to properly select and apply MTA approaches to their own organizations.