Seven West Media and Databricks are working together to accelerate the development of personalized ad experiences and audience engagement strategies for Seven West Media’s customers while fostering a culture of data democratization across the company.
Seven West Media’s 7plus is one of Australia’s leading streaming platforms for broadcast VOD (video on demand), enabling audiences to livestream broadcast content while also offering a premium VOD service with over 20,000 hours of first-run and exclusive content from global distributors, plus local channels. In the past few years, the platform has grown from a place to catch up on missed programming to a premium, VOD and live streaming service.
Working together with Databricks, Seven West Media is driving innovation with its AI Factory, a dedicated generative AI and machine learning team that works closely with Databricks. This initiative brings together a team of data scientists, solution architects, and machine learning engineers from both companies to accelerate the development of personalized ad experiences and audience engagement strategies for Seven’s customers while fostering a culture of data democratization across the company.
The AI Factory team is launching several AI-based initiatives powered by Databricks Mosaic AI that are designed to boost Seven's digital engagement, audience understanding, and advertising strategies while upskilling Seven's internal divisions by embedding GenAI at the core of business operations.
Powered by Databricks, the AI Factory is about democratizing access to data, machine learning, and generative AI across the business. The AI Factory’s goal is to improve customer satisfaction by delivering tailored content that enhances their experience with our platform. Since beginning this initiative, our customer engagement has increased significantly.— Andrew Brain, Director of Data and Growth, Seven West Media
At the heart of the AI Factory's mission is the creation of custom solutions that can enhance viewer engagement and optimize ad delivery for the 7plus platform. For instance, the team is exploring how personalized ad loads can be tailored to individual viewing patterns. By analyzing session durations and engagement scores, Seven aims to tailor ad experiences that resonate more with each viewer, ultimately improving both user satisfaction and ad revenue.
Another key focus is re-engaging the "sleepy users" among the millions of 7plus customers. Before the AI Factory, Seven’s data team spent much of their time managing ad hoc queries. Now, non-data scientists are able to quickly access new insights with the help of automated tools. With these AI-driven insights, Seven plans to create personalized content recommendations and communications that bring back around 10% of these inactive users. This approach is designed to revitalize audience engagement while increasing revenue. The marketing team is also able to access Tableau Dashboards directly through the Databricks Data Intelligence Platform to track audience metrics such as recency and frequency, for insights to help identify which audiences to re-engage.
The AI Factory isn't just about internal innovation; it's a collaborative hub for agencies and advertisers. Planned trials include predictive audience modeling and content scheduling optimization, which have the potential to refine how 7plus targets and engages its audiences. Predictive models are able to show what audiences are likely to watch in the next 28 days, as well as how many minutes of content they will consume and how big the audience will be. Seven is also linking partner data to their forecasts, leveraging AI to give insights into what viewers are likely to purchase so they can provide their advertising partners with timely and actionable insights.
Seven’s integration of Databricks Mosaic AI tools to manage their machine learning workloads and MLOPs is already delivering powerful results. Before using Databricks, Seven would wait up to 48 hours to receive audience insights and numbers, operating in hindsight. Now, Seven has become a data-driven, forward-looking company, with a 93% accuracy rate in predicting daily viewership and a 92% accuracy rate in predicting daily content revenue for 7plus. Seven’s Group Managing Director, Digital, Gereurd Roberts, emphasized the impact: "As we place data, insights, and AI at the heart of our operations, our collaboration with Databricks in the AI Factory marks a significant milestone in the digital transformation of our business."
One example is Seven’s use of Databricks Genie to develop a natural language-based query engine designed to democratize data access across all business units, empowering non-technical users to obtain real-time insights from comprehensive datasets. In a typical scenario, a business user accesses the query engine via a secure web portal and asks a question such as "How many viewers, minutes, and revenue did Farmer Wants a Wife generate yesterday compared to the week before?". The NLP engine then retrieves the relevant data from Seven’s extensive databases, makes comparisons across specified time periods, and processes the information to generate actionable insights by identifying differences and trends.
The results are presented in a clear, easily understandable natural language format, complete with tables and visualizations, allowing users to review the insights for informed decision-making or further analysis. Andrew Brain, Seven's Director of Data and Growth, highlighted the advantages of this Databricks Genie-powered NLP engine: "This groundbreaking tool will empower all business units and non-technical users to query Seven's extensive data repository, providing real-time insights to drive strategic decision-making and operational excellence."
These initiatives underscore Seven's vision of embracing cutting-edge data and AI technologies to maintain a competitive edge in the digital media landscape. As Roberts explains, "By leveraging GenAI and machine learning, we are not only enhancing our ability to predict and understand audience behaviors, we are also revolutionizing the way we acquire, personalize, and deliver content and engaging advertising campaigns to our viewers."
This blog post was jointly authored by Andrew Brain (Seven Network), Gereurd Roberts (Seven West Media), Philip Mackertich (Databricks) and Otto Li (Databricks).
