Artificial Intelligence (AI) is revolutionizing every industry, including media and advertising. AI-driven tools and technologies are reshaping the way media and advertising professionals work, making their jobs more efficient, targeted, and personalized. With the increasing demand for AI-driven solutions, the media and advertising industry is rapidly adapting to this technology.
With regards to the upcoming Data Science Salon NYC focused on applying AI & machine learning to media & advertising, we had the chance to talk to leading data scientists from the industry who shared with us their AI use cases, main challenges of their daily work, as well as look at the top AI/ML techniques to explore in 2023.
AI Use Cases in Media & Advertising
AI is transforming the media industry by enabling personalized content creation and delivery. AI-powered algorithms analyze consumer behavior, preferences, and trends to create tailored content that resonates with target audiences. By understanding consumer behavior, media companies can create content that is more engaging, leading to increased traffic, engagement, and revenue.
Personalization
Di Wu, VP of Data Science at Jellyfish mentions that Jellyfish has been using and developing AI for many years. According to her, “AI enables us to create more personalized content and experiences, by analyzing customer data and preferences. We apply this to client solutions including recommendation engines, chatbots, and many more.”
Related article: Using Machine Learning for Content Recommendations: Q&A with the New York Time’s Data Science Director
Predictive Analytics
Predictive analytics is a powerful tool that leverages historical data to uncover patterns and trends, empowering brands to make informed predictions about future customer behavior. By identifying these trends, businesses can optimize their advertising and marketing strategies to better engage with their target audience. To accomplish this, predictive analytics models such as propensity modeling, lifetime value analysis, cluster modeling, and marketing mix modeling are developed. By utilizing these models, businesses can make data-driven decisions that lead to improved customer engagement and increased revenue.
Content Classification
According to Bob Bress, Vice President and Head of Data Science at FreeWheel, FreeWheel has several interesting use cases for ML supporting their products in the video advertising ecosystem: “Some of those applications include content classification in which attributes of video content such as genre and content rating, are modeled from video metadata using deep learning techniques. This allows video content to be classified to make it easier for advertisers to find their target audience.”
Biggest Challenges of Applying AI in Media and Advertising
Data Quality
According to the State of AI in the Enterprise Report 2023 by Data Science Salon, data quality continues to be one of the top challenges when implementing AI solutions. Privacy regulations make it even more challenging from a data collection and usage perspective, says Bob Bress: “As powerful as AI models can be, they are still only as good as the data used to train them. Ensuring data quality across disparate data sets in the media ecosystem can be a challenge. Extra effort must go into ensuring models are not built with inherent biases that may skew results. Additionally, those companies relying on AI for critical decision-making will need to find ways to provide transparency and interpretability to what are otherwise black-box methods.”
Identifying causal inference
Besides data quality, Ruixin Zheng, Lead Marketing Analyst at DirecTV highlights causal inference as another major challenge data scientists are facing when using algorithms in the advertising industry: “For example, we build AI models to predict if customers buy the product after they receive the email advertisement. But some customers also buy the product without clicking the emails. So, we also need to figure out if the emails take effect in increasing the purchases.”
A/B Testing
Alex Eftimiades, Data Scientist at Penguin Random House adds A/B testing as an additional challenge when using AI in the industry: “Since measuring performance requires spending money, it is important to act on new information as quickly and efficiently as possible to redirect spending. There has been new research in multi-armed bandit problems over the last couple of years that can be leveraged for certain scenarios, but it doesn’t seem like a solved problem.”
Generative AI and Advancements in Computer Vision will further Disrupt the Industry
When asked about innovations in the industry with AI in 2023, Di Wu mentions Generative AI as a main driver: “It has a wide range of potential applications, from generating realistic images and videos to creating new music and even writing coherent text. Gen AI will revolutionize areas in creative asset production and copy development. It will help with creative assets generation in making, creating/editing images more easily. In terms of content creation - product description and ad copies; help our client to build more tools like virtual assistant and virtual try-on.”
Bob Bress adds that using AI to create new creative content will help generating content that resonates better with customers and allow greater precision of ads: “It will allow for enhanced audience inferences through audience data and create greater relationships between content and the ads that air in that content. It will drive immersive experiences through improved recommendations, shoppable ads, or even augmented reality tied to advertisements or content. This has the potential to fundamentally change how ad impacts are measured and valued.”
Novel computer vision techniques are likely to further improve ad generation, highlights Alex Eftimiades: “We need to generate bounding boxes with orientation to superimpose new books on ads with existing books. There has not yet been research on object detection with complete perspective and orientation (though Nvidia published work on generating bounding boxes with one angle of rotation), but surely the innovations are around the corner.
Conclusion
In conclusion, AI will continue to transform the media and advertising industry in 2023. By enabling personalized content creation and delivery, providing insights into consumer behavior and preferences, and optimizing ad spend, AI is making the industry more efficient, effective, and personalized. As AI technology continues to develop, we can expect to see even more exciting developments in the media and advertising industry.
Staying up to date with the most recent advancements is crucial for the successful implementation of AI initiatives. Join top data science experts from the media and advertising industry at Data Science Salon NYC on June 8th for a full day of networking and content sharing. The program includes insightful talks from the speakers featured in this post and a panel about “The New Age Data Storytelling with Generative AI in Media & Advertising”. Sign up now and learn from those working on the frontlines of machine learning in the industry.
