Data science – it’s without a doubt a buzzword in today’s digital publishing world. Until recently though, it was something associated with only a handful of large publishers: Hearst and The New York Times to name a few. For some time, these behemoth publishers have been investing resources into developing robust data science departments to inform their content strategies and fuel their digital advertising revenue.
Yet economy of scale has kept the majority of publishers out of the data science game – leaving them stuck with outdated or low-quality ways to bring new, valuable traffic to their site. In the programmatic world; however, data is undeniably the path forward for these same publishers to get their content in front of the right audience and increase their advertising revenue.
At Freestar, we specialize in creating technology for publishers that will increase their revenue and ultimately grow their site. Freestar’s VP of Data Engineering, Peter Jaffe, leads our efforts in creating revenue generating insights from our hundreds of digital partners. Peter previously worked at Hearst as Director of Data Engineering where he owned the pipelines that processed hundreds of millions of webpage hits daily. We did a QA with Peter on why data is so important in the digital space today.
Q. Why is data science in general so important for publishers today?
PJ: Today consumers have nearly unlimited media options to spend their time on. The obvious upside here as a reader, is you have access to quite a lot of content that’s going to interest you. But there are downsides to so much content being right at our fingertips. For readers, it’s “How do I know where to begin (or where to find what I’ll like)?”. For publishers, it’s “How do we get our content in front of the right audience?”.
For consumers is the answer Google? Friend recommendations? These only get you so far in finding content you’re truly interested in. For publishers, the fact is you can’t rely on people to find your content. They won’t. You have to go out and find the right readers. This is where data science comes in to bridge the connection.
Bigger publishers have been building their data science competency for years because they understand how critical it is to seek out traffic in the digital era. We’re all sitting on massive datasets – some more massive than others – but even a small publisher has the potential to log every event that happens on every page of their site. By doing this they can use this data to expand their reach, grow their audience, and ultimately their revenue.
Q. So how does data science increase revenue? And why aren’t all publishers utilizing it?
PJ: When you combine first-party user cookies with data science you get access to a treasure trove of data. Who is reading what? How much time are readers spending on what content? What ads are they seeing? What sort of revenue are you generating from different types of content? All tremendously valuable data if you want to build audiences you can use to go out and find more traffic that your content will resonate with. The more engaged traffic you can bring to your site, the more valuable that traffic is from an ad revenue standpoint.
The difficulty for smaller publishers is they can’t afford to hire a data science team to realize the value of all that data. For starters you need to hire both engineers to collect and scientists to analyze the data. Then there’s storing the data, which depending on the volume you’re collecting, can cost tens or hundreds of thousands of dollars a year to store – if not more. Many of the publishers we work with are staffed by a handful of people (or even one person) who may not be especially technical in the data space. So going out and hiring qualified people to collect, store, and mine data into actionable insights would stretch the majority of publishers to their financial limits.
Q. How can data science improve the advertising tech landscape?
PJ: There are a few obvious ways that data science can improve ad tech. You can take a system like header bidding and apply machine learning to optimize the process of soliciting bids from ad networks. You can use natural language processing to build audiences out of your user data and content. You can use deep learning to predict the performance of ad content. In a nutshell, data science is improving the ad tech landscape by improving the efficiency of the market and in the process, it’s helping get the right ad in front of the right audience at the right time.
The fact is, machine learning and artificial intelligence are rapidly moving toward becoming commoditized, thanks to aggressive pushes by Google, Amazon, and Microsoft. In a few years A.I. methods will be as common as A/B tests are now. Data science is already a part of the ad tech landscape, and will only continue to grow its’ footprint.
Q. What excites you most about working at Freestar?
PJ: Freestar is positioned in a really great place, being one of the companies that’s building the next generation of ad tech and providing a huge value to publishers. I love working with real-time streaming data and massive datasets. There are so many exciting cutting-edge technologies in the big data space right now, and I love getting to experiment with them. It’s great being able to do that at a company that’s really knocking it out of the park when it comes to growing publisher revenue (a real win-win situation).
About Freestar: We engineer cutting-edge monetization solutions for digital publishers that dramatically increase revenue. By combining industry-leading technology, data, and massive scale, we enable busy site owners to seamlessly maximize revenue while freeing themselves of the hassles of ad operations. Publishers then have more time to do what they do best: create content. Visit www.freestar.io or email [email protected] for more information.