RVA Data Hackers: Analytics for Online Journalism

Please join us to hear Andrew Montalenti (http://twitter.com/amontalenti) present on Analytics for Online Journalism. Andrew will walk us through how his company, Parse.ly (http://parsely.com), analyzes web traffic for major media outlets at scale.

Abstract

The rise of online content has simultaneously disrupted and empowered journalism. Disrupted: because now anyone can publish without the need of a printing press and army of editors acting as a gategeeker. Empowered: because content does not live and die on a static printed page, but instead flows in a evolving online ecosystem where it is viewed, shared, aggregated, bookmarked, queued, re-posted, linked to, commented upon, and reformatted for display on mobile phones, tablet apps, websites, widgets, RSS readers, e-readers and more.

"Analytics" for journalism used to have an easy, if imprecise, formula behind it: the print circulation of the publication you worked for adjusted by the estimated probability of any of your subscribers actually reading your specific article. Thus, every journalist's goal was to get on the "frontpage" of the highest circulation newspapers to maximize impact. But in the age of the web, the "frontpage" has a much more modest effect and impact can be much more precisely measured.

Parse.ly is building the definitive analytics system for online journalism. And in so doing, we are also building a rich data store. Currently, it clocks in at ~5 billion pageviews per month from ~200 million monthly unique visitors across ~10 million unique URLs.
Combining web crawling with 1st-party traffic information and data from 3rd-party social networks, we provide rich and detailed analytics to editors and writers at top media companies like Arstechnica, TheNextWeb, Dallas Morning News, Mashable, TalkingPointsMemo, and many more. We continue to grow our customer base, but more importantly, we are improving our understanding of this rich data set to better aid our customers and the media industry overall.

This talk will introduce our measurement and content modeling approach, discuss our most successful features to date, and dive into technical scaling challenges. I will also lay out our plans for the future of online media measurement.

About the Speaker
Andrew co-founded Parse.ly in 2009 in New York City. The company was incubated at Dreamit Ventures, a startup accelerator in Philadelphia, and since then has raised significant venture financing. At Parse.ly, he leads the engineering team as the Chief Technology Officer. He is also a mentor at HackCville, a hacker space in Charlottesville.