Jason Scheller, Eyeview | Spark Summit East 2016
01. Jason Scheller, Eyeview, visits theCUBE!. (00:19) 02. Background of Eyeview. (00:57) 03. Indicators of the Spark Impact on Eyeview. (02:34) 04. Why Eyeview Infrastructure was Compatible with Spark. (03:60) 05. How Customer Workflow Changed with Spark Speed. (07:04) 06. Customer Examples of Integrated Analytics Advantage. (08:13) 07. Advertising Attributes Being Measured by Eyeview. (09:23) 08. Enabling Advertisers to Work Efficiently. (10:47) 09. Necessity of Customizing Advertising Data. (12:14) 10. The Databricks Notebook. (14:45) 11. Looking Down the Road this Year with Eyeview. (15:58) https://siliconangle.com/2016/02/18/spark-gives-eyeview-speed-vision-on-ad-performance-analytics-sparksummit/ #theCUBE #Spark #SparkSummit #Eyeview #SiliconANGLE #Databricks --- --- Spark gives Eyeview ‘speed vision’ on ad performance analytics | #SparkSummit by Betsy Amy-Vogt | Feb 18, 2016 Eyeview, Inc. is focused on increasing ROI for their clients. The video advertising technology company uses Spark as the basis for its personalized video advertising platform to speed access to usable data and allow marketers easy access to analytics. In an interview at Spark Summit East 2016 at the New York Hilton Midtown in NYC, Jason Scheller, director of data and analytics of Eyeview, talked with Jeff Frick and George Gilbert, cohosts of theCUBE from the SiliconANGLE Media team, about Big Data and how it affects ad performance analytics. It’s not your grandfather’s marketing Big Data has revolutionized marketing, giving advertisers access to data on an individual level and allowing ads to be customized to target very precise demographics. “It’s about focusing the right ad on the right consumer,” said Scheller, who described how a campaign could include a million individual ads, each with a slight variation that makes the product more relevant or appealing to the viewer, thereby increasing the chance of a sale. Making the move to Spark The move to Spark required Eyeview to redesign its data architecture to a small extent, but the overall process was “pretty seamless,” said Scheller, and Spark was up and running within a few weeks. The small inconvenience more than paid off in the difference of ease and speed. Before Spark, accessing a days worth of data took 24 hours; now six months worth of data takes only 10 minutes. “It used to be ‘Let’s kick this off before I go home and hope it’s done by the morning,’” he said. Now he is working on two or three notebooks at the same time, and there is “no walk away anymore.” Slicing the data lake “We have to be able to slice and dice the data in every possible way to be able to work out what is working in different areas,” said Scheller. He quoted a recent case study where the client achieved a 4:1 ROI: For each dollar the client invested in advertising, Eyeview was able to show how they gained $4.00 in actual sales. Scheller credited this to the ability through Spark and real-time access to not only view the entirety of data, but to follow the progress of every individual ad, zooming in and allowing fine-tuning to the level where they can target an audience of one. What tech stack? Asked by Gilbert to describe the tech stack above Spark, Scheller responded that Eyeview’s analysts are able to work directly on Spark because of the notebook product in Databricks, eliminating the need to build an additional stack on top. The built-in visualizations allow the analyst to directly see results and have an instant visual to show the account manager. Bigger, faster, easier Looking to the future, Scheller sees Databricks only getting bigger and faster. “We’re getting to the point where it’s hit a button and done,” he said. “We’re taking less and less energy to get to the end point.” @theCUBE @SiliconANGLE theCUBE @Databricks #SparkSummit