Product descriptions:
As competition in Australia’s streaming market intensified, Seven West Media needed to modernize how they engaged audiences and monetized content across their 7plus platform. Legacy systems made it difficult to access real-time insights, slowing down decision-making and leaving some revenue opportunities on the table. With Databricks Mosaic AI, Seven built an AI Factory that now powers predictive audience analytics, high-ROI re-engagement campaigns and a self-service chatbot used across departments. As a result, daily active users have grown 40%, re-engaging “lapsed” users has delivered $16 million in incremental ad revenue and marketing efficiency has soared for both below-the-line and above-the-line audience communication. Workflows that once took hours of manual effort are now possible in seconds, with AI driving real impact at scale.
Slow reporting and fragmented data stunted audience growth
Seven West Media is one of Australia’s leading media and entertainment companies, reaching over 91% of the population across broadcast TV, digital platforms and print. Their flagship streaming platform, 7plus, plays a critical role in connecting audiences with premium content — whether it’s breaking news, prime-time shows or live sports events. But in a fiercely competitive market where digital and streaming engagement determines advertising revenue, Seven knew they had to move faster.
“We had data, but we couldn’t be data-driven,” Andrew Brain, Director of Data and Growth at Seven, said. “We were sitting on billions of events and attributes from our digital and streaming platforms, but that was often hard to access through legacy tech. We couldn’t access them in real time to drive smarter decisions.” With the Databricks Data Intelligence Platform, Seven unlocked a unified data foundation to support the following high-impact AI use cases.
To improve customer satisfaction and retention, Seven built predictive models to forecast 7plus’ audience composition (by device, demographic, genre, minutes viewed, etc.) and their viewing behavior. What began as a seven-day forecast pilot focusing on video-on-demand content has since scaled to a four-week outlook and now importantly includes livestreamed content, enabling proactive, coordinated decisions across marketing, sales and programming. According to Andrew, “The seven-day pilot was the aha moment. Then we extended it to 28 days and still managed to achieve an impressive 94% accuracy. Now, we’re all working from one truth, with realigned audience behavior analytics across the business.”
With over 14.7 million registered 7plus users, Seven wanted to bring back “sleepy” users who hadn’t engaged with 7plus in over 90 days. After identifying and segmenting these quiet audiences, the team launched targeted campaigns that are now outperforming initial goals, reducing churn and boosting ad revenue. “If we could bring back 10% of lapsed users every time we pushed out a campaign, that would drive $16 million in incremental ad revenue. Well, we’re now above 13%,” Andrew explained. “And we’ve shortened the window of them coming back, seeing users return in 15–20 days instead of 45.”
Last but not least, Andrew’s team successfully built a self-service language query model powered by Databricks AI/BI Genie. Previously, any question about show performance or audience metrics required emailing various data and insights staff’s inboxes and, in some cases, waiting hours or up to a day for a response. Now, with Genie’s natural-language interface, employees across departments can get real-time answers on demand. “In the past 30 days, we’ve seen 400 unique business questions come through Genie. It has reduced the inbox by 35%, allowing the data and insights team to focus on more proactive, high-value work,” Andrew said.
These innovations weren’t possible with the company’s previous architecture. When asked about capturing data and insights in a pre-Databricks world, Andrew had this to say: “We couldn’t do these types of use cases. Take our audience predictions work. Before that work, we would wait until 10 AM today to understand how well we did yesterday in terms of audience and ad revenue, and that’s how the industry would report — in hindsight. With our predictive audience modeling, we can now understand what our audiences and ad revenues will look like tomorrow and the next day, all the way out to 28 days in advance — with foresight. Forward-focused insights now allow us to pivot our strategy when we need to. If we need more of a certain demographic to deliver ad campaigns, we can now make smarter decisions to ensure we meet our advertiser demands seven days a week. With the rise of streaming competitors and a growing need for operational intelligence, the business knew it needed to change — and fast.”
Building an AI Factory for enterprise-scale intelligence with Databricks Mosaic AI
To overcome fragmented systems and slow, manual analytics, Seven did more than just implement a few isolated tools. They reimagined their data strategy around a company-wide AI Factory powered by Databricks Mosaic AI.
The AI Factory is a centralized initiative that unifies data, models, governance and business applications into a single operating model. It’s the foundation for how Seven delivers AI-powered solutions for agencies, advertisers and their own employees. According to Andrew: “All of our achievements — our production use cases, the collaboration, the speed — live under the umbrella of the AI Factory. It’s a major business enabler.” In the same place, internal teams can optimize content scheduling, and Seven’s advertising partners can make data-driven decisions on where to allocate spend.
Seven was the first company in Databricks’ history to launch an AI Factory. The initiative was developed in close partnership with Databricks Professional Services and Mantel Group, which provided local implementation support and helped build the foundation for scale. Several key components of the Databricks Data Intelligence Platform enabled the AI Factory to deliver rapid value.
For starters, the previously mentioned language model chatbot that runs on Genie is the media company’s most widely adopted AI application. “Genie has completely changed how the business can access data,” Andrew said. “It democratizes insights. People are getting answers to audience, device and content questions in under 25 seconds, so adoption has just exploded.”
For creative teams generating promotional scripts, Seven is piloting a retrieval augmented generation (RAG) solution using Mosaic AI — including Model Serving, Vector Search and Agent Framework — with query functions and a Databricks Apps front end. There were sliders for factors like tone of voice, temperature and content length so that creatives could have a more accurate starting point.
Andrew explained: “We ingested 10,000 promo scripts. What used to take 45 minutes for any given show gives a great first draft in under 4. This isn’t in production yet, but the creative team says that these scripts read like something a human would write.” That’s because with RAG, the model doesn’t just generate content from scratch — it pulls in relevant examples from trusted source material (like past scripts) to ground its responses in real-world context.
MLflow also supports the AI Factory. Since Seven runs multiple machine learning models per use case, MLflow manages and supports the full model lifecycle, from experimentation to deployment. Andrew’s team can scale AI without needing a dedicated MLOps person. MLflow and Unity Catalog help them maintain consistency and quality with fewer people. With Unity Catalog, Seven now has clear visibility from source to dashboard.
Whether building predictive models, launching engagement campaigns or fielding executive queries, the AI Factory ensures that insights are fast, trusted and accessible across the organization.
Outpacing the market, redefining the future of streaming
In the last two years, the fight for viewer attention (and advertising dollars) has never been more intense. But while the industry adjusted to new levels of fragmentation, Seven surged ahead. Thanks to AI-powered Databricks solutions built within their AI Factory, the company has seen significant gains in platform engagement, marketing efficiency and strategic alignment. According to Andrew, “We’re hitting all the right engagement metrics. We’re outpacing the market.”
Daily active users on 7plus grew 40% in the past financial year, with stream starts up 42% and total minutes watched rising 48%. In a market growing at 20% overall, Seven more than doubled that pace and reinforced their leadership in a fast-changing streaming landscape. Marketing efficiency also improved dramatically. By using AI-driven insights to guide digital campaigns, the company cut above-the-line spend. Meanwhile, click-through rates rose 40% and cost per click dropped 50%, delivering better results with fewer resources. These savings enabled smarter targeting and freed up budget for more strategic initiatives.
The AI Factory is also reshaping how Seven is perceived externally. Advertisers and agency partners are taking notice. “We’re having new conversations with stakeholders. It’s no longer just about campaign costs, but how we can deliver smarter, data-driven outcomes together,” Andrew said.
Looking ahead, Seven is exploring how to expand their AI capabilities into the buy-side advertising space. The goal is to make it easier for small businesses and nontraditional advertisers to access premium video inventory on 7plus. “The adage was, if you don’t have a million dollars, you can’t advertise on TV,” Andrew explained. “But with agentic AI, we want to change that. We want to make it as easy as buying a Facebook ad — for the mom-and-pop mechanic business who wants to spend $1,000 a month on local area marketing.”
Using intelligence to simplify ad buying, reduce friction and expand reach is foundational to Seven’s next chapter. “I want to talk about the internal and external democratization of data and AI-driven insights, not just about CDPs and cloud migrations,” Andrew concluded. “This is about using intelligence to expand the business and reach new audiences. That’s what sets us apart.” As Seven continues to lead innovation, Databricks is proud to be a trusted partner on their journey.