DSS Blog

DSS Virtual: Media, Advertising & Entertainment Speakers on Big AI Trends and Challenges Impacting Their Industries

Written by Data Science Salon | Apr 27, 2021 3:48:07 PM

We are excited for next week’s DSS Virtual: Applying AI & Machine Learning To Media, Advertising & Entertainment on May 4-5. Ahead of the event, we asked a few of our speakers, some of the brightest leaders in the media, advertising, and entertainment in data science across the nation, to share their insights on the current state of industry trends and innovations. 

Sharing their thoughts with us are:



Get many more great insights at DSS Virtual: Media, Advertising & entertainment on May 4-5.

 

What are some of the biggest trends you’ll see being adopted by the entertainment and media industries?

Christopher Whitely: I think there are a few areas that we’ll see adopted by M&E industries in the coming months and years, including more contextual advertising, where advertising creative assets are matched to appropriate program content algorithmically. Federated learning is also a new trend, which refers to modeling using machine learning without sharing data sets. Privacy is important, so I expect we’ll see continued use of aggregated customer segments and clean rooms for marketing and analytics.  Also, lookalike models will help advertisers reach potential customers and optimize campaigns for the greatest effect.

Marketers are looking to improve targeting, efficiency, and automation. This will come through in cross-platform addressable decisioning, with unified advertising campaign planning, pacing, and execution across media channels, including linear addressable.

Jay Kachhadia: All the media companies have started realizing the power of data and wonders it could do as they have started moving towards powering their streaming business. So, If you love to work in entertainment, then this is a great time to enter the space as all of them are expanding their data science teams with new talent.

Vishal Juneja: More and more adoption of ML in not only the more primary use cases of content marketing, recommendations but also as an aid in content creation, finding new mediums for content consumption that enhance user engagement. Also solutions like Federated machine learning that are more compliant in terms of user privacy.

What will be the biggest challenges for those that already have dedicated data science teams?

Wayne Cheng: Data collection of high quality datasets and data engineering of complex datasets.

Christopher Whitely: The key will be determining how to organize projects to facilitate innovation and then get it adopted. Teams need to balance project-specificity with the ability to serve a broad set of needs.

Bonnie Magnuson-Skeels: I think continuing to update our methods as privacy laws change and less or different data is available for collection is important. We’ve always been privacy first at Samba TV and introduce privacy protections ahead of regulations, so we’re already strongly positioned with the data we collect.

Yi Kang:  Explainability. Not every model needs it but a lack of explainability will make it harder for models to get adoption and in certain industries run afoul of regulators.

Vishal Juneja: Alignment with the business teams and goals, support from the management with a clear data strategy roadmap will define success or failure of the data science initiatives within the organization.

What is something you wish you’d known about the media, entertainment, and advertising industry before you joined? 

Wayne Cheng:  The gap between research and industry is wide, with only a few organizations attempting to bridge the gap.

Christopher Whitely: I’ve found that this industry can be more complex than it seems at first and it’s easy to get lost in the weeds. For this reason, I think it’s critical to understand the big picture before diving into the details. Media is very broad – so my advice is to spend time talking with people that perform different functions in different parts of the industry (TV, movies, ad tech, etc.) before deciding on your passion.

Vishal Juneja: The fact that advertising is not an exact science and there are several blind corners of user behaviors owing to technical limitations and user privacy issues. Possibly a better way to look at it is as a Data Science challenge where we have to do our best and come up with the most appropriate solutions.

What kind of changes do you expect to see in the next 5 years for the industry?

Wayne Cheng: The adoption of AI technology as a content creation tool.

Christopher Whitely: I think a lot will change in the next 5 years! I expect consumers will want the ability to provide instant feedback to creators and to get instant, interactive recommendations based on what they feel like watching. The TV world will have much more of their TV advertising data-enabled, and AI will play a greater role – allowing more people to become creators, and making experimentation with advertising more effective.

Jay Kachhadia: I personally feel that media companies would be more into making power-house data science teams in the coming 5 years that can do everything from building ML models to data/ML pipelines to building data products for various internal stakeholders giving rise to more opportunities for Data Scientists, ML Engineers and Data Engineers to be a part of the same team.

Yi Kang: Speaking for digital advertising, it would be how to deliver relevant ad experiences to consumers in a post-cookie world.

Vishal Juneja: Increasing focus on user privacy and an evolving ML landscape that is able to work within the new limitations and slowly evolve to optimize both goals of privacy concerns and recommending useful products to the consumers through personalization by leveraging high quality first party data.

What would make your job easier as a data scientist in this field? 

Wayne Cheng: The availability of high quality datasets and easier access to machine learning hardware accelerators.

Christopher Whitely: We need more efficient ways to keep up with academic research in the field while also understanding important new findings deeply – while we’re doing the work itself.  And we need better ways to manage and simplify an industry that is getting more complicated, including faster querying and modeling tools, and automated data quality checks and tools that help explain metadata in complex data sets.

Bonnie Magnuson-Skeels: I would love to have more detailed demographic data available from the US Census Bureau. In general, the finer-grained data is, the more uses it can have.

Vishal Juneja: More support from Data Engineering is what a Data Scientist always wishes for.

That was a snippet of insights that our speakers will discuss at our upcoming DSS Virtual: Media, Advertising & Entertainment next week. Some of the topics covered are Content Personalization and Monetization, Personalization at Scale with AI, Cloud Automation and Machine Learning, Audience Targeting and Segmentation (across platforms), Data and AI for emerging platforms, Data Governance and many more! Be sure to register today to enjoy these and many more insightful sessions! See you there!