How YP Use Their Mobile Data

Interview with Eric Farng, Technical Lead Data Scientist at YP


Eric Farng is the technical lead data scientist at YP Mobile Labs. He is responsible for managing and optimizing performance for thousands of campaigns across mobile and display. He plays a key role in the strategic deployment of ads for national brands.

Eric joined YP in November 2015, bringing with him more than a decade of experience in data and engineering. Previously, Eric was senior data scientist at GrubHub, and before that worked at ad tech and eCommerce brands including Bluefly, Yodle and Quigo. Eric has degrees in Computer Science, Machine Learning, and Statistics from Carnegie Mellon, Cornell, and Columbia. We sat down with him ahead of the Marketing Analytics Innovation, taking place in Chicago this May 19-20.

How did you get started in Marketing Analytics?

My path in marketing analytics came with stops and starts. It began at Quigo Technologies, an ad exchange for publishers which was eventually acquired by AOL, where I got my first taste of what the future could hold. I was intrigued with how Natural Language Processing applied to contextual ad matching and the myriad of challenges that resulted, whether it was calculating CPM or CTR or adding new ads and new ad locations. All of these problems led me to pursue a career that focused on the application of data science to marketing.

How is the approach to Marketing Analytics different at your company/ what do you do differently?

At YP mobile labs, we use mobile phone data to create detailed profiles for each device. We combine phone GPS locations, census data, and mobile app data to create anonymous user profiles that allow us to target more effectively. We are also combining mobile data with desktop data to create a more complete user profile that can better determine purchase intent.

How have you seen analytics within marketing develop over the last 10 years?

The shift from simple daily automated bidding to real-time bidding has created a wealth of data for advertisers to use to better inform marketing strategies and ways to measure success. There are literally billions of pieces of data that can be analyzed through algorithms that suit an advertiser’s needs. When combined with mobile location data, this information can be used to track the effect of mobile advertising spend on real conversions and in-store revenue. At YP, we share this information via a Store Visit Report, and it’s something that was unimaginable 10 years ago.

Are there any innovations in the data science space that have been real ‘game-changers’?

We're all excited about how Deep Learning created huge gains in speech recognition, natural language processing, and image recognition. There are lots of people trying to find the right place for it in marketing, and I think it can have great success in marketing analytics, possibly click fraud detection or behavioral targeting.

Another game changing innovation has been the introduction of new big data tools in marketing. Currently, there is a resurgence by the SQL database community in creating New SQL databases that can handle big data (Vertica, Redshift, VoltDB, SciDB, etc...). I wonder if these new tools will minimize the need for engineers to help analyze data and change the players again.

What will you be discussing in your presentation?

The rise of digital advertising has opened a new world of possibilities with respect to measurement. Yet at the same time, there is a lot of work still to be done. We’ll be talking about ways to combine online marketing with offline sales, the pros and cons of different methods and what the ultimate solution could be.

You can hear more from Eric, along with other industry leaders, at the Marketing Analytics Innovation Summit, taking place in Chicago this May 19-20.

University lecture small

Read next:

How Are Higher Education Institutions Using Analytics?