How AI Improves Forecasting and Detects Anomalies for Media Finance Teams
SymphonyAI Media and IABM teamed to for a webinar that examines how artificial intelligence (AI) and machine learning (ML) can help media and entertainment companies‘ greatest revenue challenges. In this post, we share key takeaways for finance teams seeking to deliver greater value through these technologies. View the full webinar on demand here.
Anyone working in media finance today knows just how fast the video industry is changing. New OTT platforms are entering the market and business models are evolving. We are seeing an increased focus on licensing models such as AVOD and the introduction of new ones like PVOD. Meanwhile, companies are still trying to maximize revenue from linear TV licensing deals. Revenue is becoming increasingly fragmented, which brings unprecedented complexity to distribution deals and revenue streams.
Industry finance teams are grappling to manage and maintain larger and larger volumes of data from a variety of sources. Trying to manage new business models (and complicated revenue streams within existing ones) with tools like Excel is testing organizational limits. Finance teams need abilities like accurate revenue forecasting, but sophisticated analysis and reporting are operationally unthinkable to teams that rely on manual workflows and legacy tools.
With multiple licensing models and so many data points to manage, how can AI help finance teams detect anomalies in financial reporting and predict revenues accurately?
Media and entertainment revenue models are especially vulnerable to hidden anomalies, as data is dispersed from a variety of different platforms such as distributors, subscriber management systems, content delivery networks, and streaming and payment platforms. While supervised ML answers human defined business questions, unsupervised ML can surface valuable answers to questions you don’t think to ask. Unsupervised machine learning (ML) excels at pattern detection to locate anomalies hidden in complex data sets – the unknown unknowns.
To take a real-world example, the Revedia AI platform detected a segment of streaming subscribers who had signed up for a promotion that, by mistake, gave them indefinite free access to an SVOD service. The provider was unaware of the issue, but unsupervised ML doesn’t require humans to pose specific business questions; it automatically finds revenue leaks and calls them to users’ attention (in this case, in a matter of hours after initial system deployment).
AI software can also detect revenue leakage in the form of distributor underpayment, such as a video service inaccurately reporting subscriber counts or paying below its contracted rate. At SymphonyAI Media we have error rates upwards of 25%, which would have gone undetected without deeper analysis. Enterprise AI significantly reduces the risk of finance teams missing subtle data anomalies and revenue leakage that have a profound impact on business revenue.
AI and data analytics provide finance teams with a single, comprehensive overview of their revenues and subscribers for accurate financial reporting. With access to visualizations, dashboards and “explainable AI” (the ability to describe how a conclusion or prediction was reached), it allows clear visibility across all metadata.
Supplementing financial analysis with predictive forecasting delivers even greater value. For example, is subscriber loss a seasonal normality or indicative of a forthcoming downward trend? By blending historical and third-party data and applying highly accurate predictive algorithms, AI can process data at a scale humans cannot. AI forecasting tools therefore deliver much more accurate predictions. Humans, for their part, can adjust variables in AI models to better understand how changes in the business could positively or negatively affect future revenue.
The media finance environment is much more complex and nuanced than in other sectors.
Many commercial AI solutions focus on fixing problems higher in the value chain, such as AI-assisted content production and AI-assisted marketing. These common applications no doubt add value, but organizations must correctly understand the root cause of revenue challenges. If a marketing team sees that revenue is falling, they may use AI to determine the best discounting promotion to increase upgrades. But finance teams with AI capabilities can help the entire organization understand what’s driving revenue shortfall in the first place; suppose it’s due to inaccurate subscriber reporting. In that case, investing in customer acquisition (potentially at a high cost to the business) addresses the wrong issue.
Why aren’t more media finance teams using AI, then? The main issue is that mass market “horizontal” AI solutions are not built to spec for the industry. They require extensive customizations to be useful in the unique data landscape of media revenue. While these solutions could serve some core needs, like automating report generation, they ultimately limit the value that AI can deliver for all stakeholders across the business. A horizontal AI solution provider isn’t going to dedicate 100% of its resources to understanding and staying ahead of the constant disruption in the world of streaming media.
To keep pace with this ever-changing industry, finance executives are turning to vertical AI solutions that are built for their requirements and can be future-proofed for their ever evolving needs. Specialized AI software can provide finance teams with not only far superior and more efficient workflows, but can also structure and normalize revenue data for more effective anomaly detection and forecasting.
The value from AI is evident, with 73% of corporate finance teams seeing revenue lift from AI adoption, a 23% YOY increase. AI is an essential technology for finance teams to make informed decisions. Finally, AI solutions are available that address the specific pains of managing and optimizing revenue in media and entertainment. Expect it to become the norm in the near future.