George Sadler heads up the global Marketing analytics team at eBay, leading a team of passionate data scientists and business analysts to measure the impact of eBay’s global CRM, IM and Loyalty activities via A/B tests, econometric and multivariate models and other techniques. They uncover actionable insights to optimize operations in email, mobile notifications, SEM, display, paid & free social, affiliate and loyalty programs. We sat down with him ahead of his talk at the Marketing Analytics Innovation Summit.
How did you get started in marketing analytics?
I fell into it, really. I was doing operational strategy and BI work when Dell started heavily investing in social media marketing. They soon ran into the tough ROI questions, and since there were no ‘quants’ in corporate marketing at the time, they looked to the BI teams in the regions. And so, a colleague and I started the social media analytics team (I was employee #2 of what was eventually a 20 person team). Then I inherited a newly consolidated market research team with the logic that listening to customers in social is similar to listening to them in primary and secondary research. This team soon grew its charter again, as Dell launched its first enterprise brand campaign, and we got into econometric models, growing into what is now the marketing sciences team at Dell.
How is the approach to marketing analytics different at your company/what are you doing differently?
Scale and sophistication (some would say complexity too). We are hyper-focused on causality (correlation is not good enough at eBay). When I talk to others about what we do and how we do it, I often hear ‘you guys are on another planet!’
How have you seen analytics within marketing develop over the last 10 years?
The bar is continually being raised re: demonstrating impact. We used to say that ‘half of our marketing dollars are a waste, I just don’t know which half.’ Now we are asking ‘which dollar is a waste… and why?’ We can’t always answer this question, but that’s the bar.
Are there any innovations in the data science space that has been a real ‘game-changer’ (or that you think will be in the future?)
Uplift models are a game changer. No longer are we satisfied with knowing ‘who will buy?’ Now marketers know ‘who will buy, but only if I send them an email?’ Machine Learning will become a game changer. ML will replace much of the decisions that marketers make today (manually), so that they can be made at scale and in real-time. This will lead to increased personalization and ultimately customer-centric marketing at scale. It will also relegate marketers to campaign and copy design (until the machines take that over too).
You can hear more from George, along with other experts in marketing analytics, at the Marketing Analytics Innovation Summit, taking place in Chicago this May 19-20.