How important is AI for the future of the video industry?

The video industry has transitioned from a growth-at-all-costs mindset to one defined by retention, engagement, and profitability. It’s clear that AI, with its ability to deliver greater automation and efficiency, as well as higher levels of user satisfaction, will be central to what comes next. Whether it’s surfacing content that viewers didn’t know they wanted, improving ad targeting, or quietly adjusting the user interface to suit someone’s behavior in real-time, AI is set to become the behind-the-scenes engine powering the viewing experience.

However, while appetite for using AI-enhanced analytics is strong, its application remains relatively elementary. According to a new research study carried out by Accedo in collaboration with Professor Serguei Netessine of The Wharton School, while video services understand the potential of AI-powered data analytics in driving personalization, predicting churn and improving user engagement, extracting the maximum value from data remains a challenge. Video providers often struggle to turn raw data into meaningful, actionable insights, and many are placing bets on AI to help them do exactly that.

Current state of play

With the right data, AI will theoretically empower services to optimize the viewing experience, tailor every part of the UX and UI, anticipate potential churn risks, make content more discoverable, and enable data-driven decision-making. In practice, few video providers have yet reached this level of sophistication.

Video services typically collect various viewer data points, including data around streaming performance, content consumption, quality of service, app behavior, and onboarding. Out of these, streaming services ranked content consumption analytics as the most important factor, and surprisingly, quality of service analytics as the least important. This is particularly interesting given that user retention and satisfaction hinge as much on quality as on content relevance. The research also found that the majority of streaming services collect and report data in real-time, though not all providers have reached this level of maturity yet. Additionally, only 27% of streaming services used predictive analytics to forecast user behavior and trends. 

The ecosystem of analytics tools remains fragmented. As identified in the study, streaming services use a mix of third-party analytics tools, with Google Analytics leading the pack, followed by Adobe, Conviva, NewRelic, JUMP, among others. Many also rely on in-house systems or vendor-provided proprietary tools. The use of disparate tools combined with the fact that data is at times inconsistent and held in silos is hugely problematic. Without unified and consistent, reliable data, confidence in data is undermined and data-driven decision making is challenging if not impossible. This is a real problem for video services, especially when executives are potentially swayed more by external market expectations than internal insights.

Data is the foundation

AI is only as good as the data that is fed into it. Quality insights that help video services optimize the viewing experience and improve engagement are only possible if the foundational data is solid in the first place. Data should therefore be considered as a strategic asset, rather than a byproduct of user interaction. If a dataset is inconsistent, gappy or out of date, no algorithm, no matter how advanced, will deliver the right results. Instead, it may well have the opposite effect than was intended. It could end up driving viewers away by serving poor recommendations, irrelevant ads, ineffective personalization or an out of touch UX/UI. This is why data cleansing and validation processes are critical to the effective application of AI for data analytics.

Once the data foundation is in place, AI-driven analytics holds immense potential for video services. Predictive analytics can be used to predict and prevent churn, identify inefficient monetization strategies, identify revenue leakage such as fraudulent activity and unauthorized account sharing. Additionally, AI-driven recommendation systems can be used to dig into individual user preferences, viewing behavior, and contextual data to surface content that feels tailor-made for each viewer. And for ad-supported services, AI-driven ad targeting, placement optimization, and audience segmentation are all seen as areas with high potential for improving ROI, provided once again, that the underlying data is reliable.

The onboarding process is another area that stands to be refined by enhanced AI-driven data analytics. As identified in the study, video services already use data analytics to streamline signup flows or identify friction points during onboarding. If this process can be further improved by implementing better onboarding and conversion AI-driven analytics, video services stand to benefit from measurable gains in conversion. When data insights are applied thoughtfully, even small tweaks can produce meaningful impact.

Advertisement

Ushering in a more predictive, proactive, and personalized future

It’s clear that video services are no longer satisfied with conventional analytics. They’re actively exploring AI technologies that do more than just analyze existing data; they want systems that can anticipate user behavior, drive engagement before it drops, and adapt experiences in real-time. According to the study, personalization is the top priority for analytics usage, and there is demand for AI-powered tools that can enhance recommendations through AI-driven metadata enrichment.

There’s also a real appetite for intelligent AI-driven analytics that can flag churn risks early and suggest meaningful ways to keep viewers interested. Services are also looking to better understand how viewers engage with content, which is creating a demand for AI-driven tools that can refine content engagement analytics at a granular level for deeper insights.

Additionally, there’s a growing interest in merging insights from multiple sources and channels including social media, as well as service interactions and feedback, to build a fuller picture of viewer sentiment. On the ad front, the push is toward applying AI to optimize ad placement dynamically based on real-time viewer behavior. Content acquisition decisions have also become data-driven, and there is demand for AI-powered tools that can provide real-time performance analytics to shape licensing and investment decisions.

In a landscape defined by rapid change, the services that have the AI capabilities and tools to turn raw data into relevant, real-time insights will set the standard for what streaming can become.

Mrugesh Desai, AccedoMrugesh is a technology and media executive with 17+ years of global experience leading high-impact teams and scaling digital businesses.