The talent and game-changing data science research coming out of academia never ceases to amaze me. There’s such a tremendous spirit of curiosity and exploration in faculty and students, and we encourage our research team for Adobe Marketing Cloud to embrace this spirit. When we launched Adobe’s first ever Adobe Digital Marketing Research Awards back in June 2014, this was the heart of the idea. By exchanging ideas and research between Adobe and universities, we knew we could uncover incredible advances in data science, and that we could push the boundaries of how data science is used in digital marketing.
Adobe sits on a wealth of data with Adobe Marketing Cloud. Adobe Marketing Cloud powers 41 trillion transactions and 4.1 trillion rich media requests for our customers, annually that includes two-thirds of the top Fortune 50 brands. This presents a tremendous opportunity to develop theoretical and empirical solutions to problems in marketing that can improve consumer experiences.
Adobe’s annual Data Science Symposium at our San Jose campus is an opportunity for academic researchers to learn about our Digital Marketing business. Professors also present research proposals for the Adobe Digital Marketing Research Awards, which provide up to $50,000 in funding support from Adobe. Research institutions including the likes of UC Berkeley, Stanford, Massachusetts Institute of Technology (MIT), and Northeastern University participate at the Symposium.
The academic institutions we work with benefit from the ability to bring together conceptual ideas with real, industry problems and data. Here are a few projects from past recipients of our Adobe Digital Marketing Research Awards:
- UC Santa Cruz Professor Lise Getoor’s cross-device graph-based entity resolution research, which was awarded a grant in August 2014. The research matched users across devices and platforms by understanding contextual features or signals (e.g. location, login info) across the different platforms, and developing scalable algorithms to perform the matching. Dr. Geetor’s research enhanced personalization performance by 10-15 percent, suggesting that the ability to match users across their various devices would provide brands with better targeting and insights into consumer behavior.
- Professor Ramesh Johari from Stanford and his team, Adobe Research Award grant recipients in August 2014, hypothesized that predictive modeling – or building correlation insights from existing data – isn’t sufficient to determine the value of potential interventions on individual’s purchase decisions. So take for example a marketer sending a digital coupon to a customer. It’s not simply about sending a coupon; it’s about how much more likely the individual is to convert with the coupon. Dr. Johari combined machine learning and causal inference techniques, so in this case, it is not about simply measuring how many people who received a coupon made a purchase but rather what influence did the coupon have on the individuals that led them to make a purchase to evaluate the impact of interventions to individuals. This is especially important when optimizing ad spend across multiple channels, where the goal is to maximize the ‘bang-per-buck’ matching specific marketing content with particular customers.
- Professor Anindya Ghose from New York University and his team won a grant in February 2015. They proposed a new mobile advertising strategy leveraging information from mobile phones on consumers’ offline activities. The research offered insights into how businesses can leverage both consumer mobile in-app and offline physical behaviors to better understand and influence customer preferences.
- Rutgers University’s Professor “Muthu” Muthukrishnan won a grant in August 2015. Professor Muthu has an advertising background, having previously worked at Google. His project examined issues with analytics, in particular detecting anomalies (e.g. a sudden drop in people of certain demographics visiting a page or a sudden increase in clicks from a specific region) and offering insights into the cause of change, especially when the data is large in size and the number of possible explanations are numerous. The goal of his research was to develop algorithmic solutions to circumvent these challenges.
Students from Professor Getoor’s UC Santa Cruz program and Professor Johari’s curriculum at Stanford have interned at Adobe. We’re always on the lookout for both undergraduate and graduate level interns to work with our product and operational teams, as well as with our Adobe Research organization. In addition to the Adobe Digital Marketing Research Awards, we offer grants to universities to continue our work with interns and their professors with students returning to school following a successful internship.
I’m excited to share that on May 26 2016, we’ll host our third Data Science Symposium at our Adobe San Jose campus. Stay tuned for a look into the inventive ideas and research discussed and presented at this year’s Adobe Data Science Symposium.
Adobe will also award academic institutions with research grants up to $50,000 for data science projects in the areas of digital marketing in February and August. If you’d like to submit a proposal for the Adobe Digital Marketing Research Awards, submission details can be found here.