We build data science communities both digitally and face-to-face. Organizer of Data Science Salon.
In advance of our upcoming event - Data Science Salon: Applying AI and ML to Media, Advertising, and Entertainment, we asked our speakers, who are some of nation’s leading data scientists in the media, advertising, and entertainment industries, to answer a few of our most pressing questions about the future of their industries.
Read on for their insights! Register here.
“There are two key trends that are having a significant impact in terms of how data is used in the industry. The first is related to data privacy. Now, more than ever, it is critical to ensure that data is stored, processed, and analyzed in ways that are consistent with consumers’ expectations of how their data is used which includes adherence to increasingly restrictive legal requirements.
This forces those in the analytics space to be more innovative in how data is used to drive impacts while ensuring privacy compliance.
The second key trend is the increasing level of analytical sophistication that is being built across companies in the industry. It is clear that any successful company in the media and advertising space is investing heavily in their analytical talent across all departments, including HR, Engineering, Product Management, Business Operations, and others. On top of that, advanced software applications now enable people from a variety of roles and backgrounds to run sophisticated analyses without needing an advanced degree in Data Science.
Analytical skills have gone from being something leveraged by a centralized team of highly trained analysts to something people across the company must bring to the table to be effective at their job.” -Bob Bress, Head of Data Science for Freewheel, a Comcast Company.
“Over the past ten years, the biggest difference I’ve seen is that the number of people asking for and relying on data. This is coupled with an increased understanding of what data can -- and can’t do -- and an awareness of what should inform different decisions.
As a result, I’ve seen product managers and engineers be more cognizant of integrating data collection into their products, creating a virtuous cycle.
Decisions ranging from design and user experiences choices, to what, where and when we’re distributing content are now data-informed while previously these decisions would have been reliant on instinct and anecdotes.” -Alyssa Zeisler, Research & Development Chief, Senior Product Manager, Editorial Tools, Wall Street Journal
“In the past, companies primarily monetized their data through the aggregation and sale of data to 3rd parties, usually for advertising or research purposes.
Increasingly, however, companies are beginning to realize the value of their data and are prioritizing more in-house applications. This is especially manifest in the pivot towards ever more targeted and personalized user experiences, crafted based on a user’s historical interactions with the brand.” -Daryl Kang, Lead Data Scientist at Forbes
“By far the biggest change and most pressing one is the industry steering away from 3rd party cookies and focusing on privacy and data issues. We are trying to reorient our entire data offering and advertising platform to not only address this but also make the advertising and user experience better.” -Amit Bahattacharayya, Head of Data Science at VOX Media
“Focusing specifically on Marketing Analytics - Data-driven marketing campaigns continue to grow annually. The biggest changes over time have been:
“The technology behind motion pictures has only a few transformational innovations since the beginning of filmmaking. First came the sound for silent movies, then there were colors for black and white frames, then everything went digital.
Similar to expanding to these new dimensions, now we are in the middle of another transformation: volumetric capture, converting 2D filmmaking to 3D. Capturing 270 GB/sec of raw data enables this revolution to record everything in 3D, truly digitizing all performances.” -Ilke Demir, Senior Research Scientist, Intel
“In the Entertainment Media space, there are two primary changes that we are seeing with data usage that requires creative thinking and adaption. First, is the changing dynamics around third party cookies, which is a well-documented roadblock to leveraging data across all industries. But, within the Entertainment media space, in particular, it's how to adapt to the disruption in Entertainment with the rise of streaming, e-Books, podcasts, digital music, etc.. How does an industry that has traditionally followed a print business model adapt to the fast-paced, quickly-evolving digital era? At Meredith, we believe that data is the key to unlocking the needs and interests of our users. We are approaching the changing landscape accordingly.” -Wes Shockley, Senior Manager - Audience & Analytics, Meredith Corporation
“In advertising, there are increasingly new and innovative ways to get a message across to specific target audiences. Addressable advertising on television is a great example of how advertisers are going beyond traditional digital advertising to reach relevant audiences at scale. These type of capabilities open up an opportunity to customize ads in a way that will most resonate with audiences. Rather than a one-size-fits-all type of ad campaign, an insurance company might, for example, target those in the market for motorcycle or boat insurance with an ad that specifically speaks to motorcycle or boat owners. These data-driven capabilities give advertisers more options in how they develop and use creative assets to maximize the impact with their audience.” -Bob Bress, Head of Data Science for Freewheel, a Comcast Company.
“Data generally has made us more nuanced in how we think about our audience and coverage. We no longer use a one size fits all approach, but look to offer customized and personalized experiences to our users.
On the storytelling size, the abundance of data has been a boon to newsrooms and opened up new types of reporting and stories. At WSJ, we’re using data science to help us uncover relevant information faster -- stock market anomalies, rhetorical analysis of political figures, and changes in search algorithms to name a few. In other news outlets, machine learning has enabled incredible stories like the Mauritius leaks or tracking down spy planes. These are stories that would have been incredibly difficult if not impossible without advances in data science. We’re in an incredible moment where entirely new techniques are being developed that help us uncover and publish the news.” -Alyssa Zeisler, Research & Development Chief, Senior Product Manager, Editorial Tools, Wall Street Journal
“As data becomes more accessible, I see the content creation process becoming more formulaic, often gravitating towards established best practices or choices that have been A/B tested. While the benefits of this to the creative process are debatable, content is undeniably driven more and more by popular opinion. This affects the way headlines are written, the images that get picked, and even the publication times for articles. Perhaps subconsciously or otherwise, I believe creators are now more mindful of the impact of their pieces, not just on their readers, but on advertisers and SEO performance.” -Daryl Kang, Lead Data Scientist at Forbes
“Data and data integration provides creatives the opportunity to experiment with ideas and opens pathways to enhance personalization and engagement. Information about customers’ behaviors and preferences provide powerful clues about messages that are most likely to resonate.” -Denver Serrao, Sr. Software Development Engineer at WPEngine
“Data can help us understand not only what content resonates with people, but also give us signals on what works and what doesn't. For example, we have a multi-armed bandit algorithm to help surface the article title that resonates best with the audience. Not only does it optimize pageviews given several options, it can learn general properties of titles that perform the best.” -Amit Bahattacharayya, Head of Data Science at VOX Media
“Data has an influence in the content creation process at every stage of the process. From research to understanding your users and finding niche audiences to understanding missed opportunities and seeing digital traffic outcomes, in every step of content creation, there is a data input that can be leveraged. If the editorial team embraces it, creating a data feedback loop that they can rely on and build off of will lead to growth and engagement.” -Wes Shockley, Senior Manager - Audience & Analytics, Meredith Corporation
“This question leads to a sea of answers in my opinion. On one hand, automatic content creation (i.e., generative models) needs data, needs to learn from and with humans, in order to mimic some level of expressivity. On the other hand, humans are also influenced and enabled significantly by having many options, patterns, and freedom over the control of generative models. A high-level answer would be integrating AI workflows for immersive content creation, where we leave the trivial tasks to be learned and automated from data, and let artists go beyond the sample distribution by injecting coherent creativity” -Ilke Demir, Senior Research Scientist, Intel
“The world is telling us that advertising has gotten out of hand. We see 3rd party cookies on the verge of extinction. On the other hand users are willing to engage with the content they love and advertising that is somewhat personalized but not creepy will earn more trust.” -Amit Bahattacharayya, Head of Data Science at VOX Media
“We’re a very transparent organization in terms of our newsroom standards, because we believe in the importance of these practices in fostering trust with our audience. We publish methodologies (scroll to the bottom) for our computational journalism and are clear with bylines as and when we’ve used new technology. For instance, in the college rankings we published last year, which used NLG to create almost 1000 unique articles. I think a lot of organizations could learn from this and be more open in their methods, whether it be a reporting process, a paywall, or something completely different.” -Alyssa Zeisler, Research & Development Chief, Senior Product Manager, Editorial Tools, Wall Street Journal
“As scientists, sometimes we are hardwired to think quantitatively, leaving out everything that we cannot measure. We should start by internalizing that we cannot deterministically optimize, measure, or evaluate everything; especially human-centric issues. Once this is internalized by teams, companies, and individuals; the need for diverse ideas emerges as a consequence. Only then trust can be earned by shaping the guidelines around privacy and ethics following these diverse ideas from diverse people with diverse backgrounds.” -Ilke Demir, Senior Research Scientist, Intel
“I think the key to earning more trust is through transparency with consumers and clients in how data is used. Any steward of personal data should let consumers know how that data is used. This may include whether it is shared with other parties and how it supports different products and services. Additionally, companies with personal data need to give confidence to their consumers and clients that it is protected from hackers and other threats. -Bob Bress, Head of Data Science for Freewheel, a Comcast Company.
“Embracing best practices and caution when it comes to privacy will protect the user and protect the business. Generating opt-in experiences in order to better understand your users is great, as is the development of proprietary anonymized user information that can be leveraged to feed your technology and inform your analysis. Being transparent about your data usage for either of these use cases will create trust between your brand and the user. -Wes Shockley, Senior Manager - Audience & Analytics, Meredith Corporation
Hear from these speakers and more at Data Science Salon: Applying AI and ML to Media, Advertising, and Entertainment, September 22-25, 2020.