The way people discover what to watch is becoming increasingly complicated. Not because there isn't enough content, but because there's too much of it, spread across platforms, feeds, and recommendation engines.
In the latest episode of After the Download, RealityMine® CEO Chris Havemann sits down with Rohan Jacob, founder and CEO of Nokio, to discuss how audiences navigate that abundance and how recommendation systems might evolve. Several themes come up that go well beyond streaming: the challenge of content discovery, the role of social recommendations, the limits of algorithmic suggestions, and what those shifts could mean for platforms and creators.
One of the central topics in the discussion is the fragmentation of discovery. Today's viewers receive recommendations from many places: streaming homepages, social feeds, messaging apps, trailers, and increasingly AI tools.
Jacob describes how recommendations often appear informally in everyday conversations. "Most of these used to happen on WhatsApp, Facebook Messenger, Instagram — it's all over the place," he explains. "And then you're like, where is that title you sent?"
The idea behind Nokio emerged from that behavior: rather than creating another content platform, the goal was to build a structured place where recommendations from friends and communities could live in one space.
The episode highlights a useful distinction between discovery platforms and engagement platforms. Many digital services are designed to maximize time spent within the app. Metrics such as session length, scrolling behavior, and daily usage dominate platform design, but discovery products operate differently.
Jacob describes Nokio's role as a starting point rather than a destination. "We would be the platform you open before opening Netflix or YouTube," he says. "The goal is to help you decide what to watch and then leave."
That positioning places discovery tools at a different stage of the digital journey. Their value lies not in prolonged engagement, but in helping users make a decision quickly. It also explains why usage patterns can look unusual compared with typical social platforms. People may not open discovery tools every day, but they return when they are actively deciding what to watch.
Jacob introduces a framework he calls the 1-9-90 model, which is worth unpacking for anyone thinking about audience behavior more broadly.
In this framework:
Most digital platforms are designed around the first two groups while the 90% largely remains invisible. The majority of viewers watch films and series regularly, yet rarely write reviews, comment publicly, or follow critic culture closely. For this group, discovery often still happens through informal word-of-mouth recommendations. That insight shapes how Jacob approaches discovery tools: rather than focusing primarily on critics or influencers, the goal is to make everyday recommendations between viewers easier to track and share.
Havemann asks how AI might influence content discovery in the coming years, and Jacob's answer is more measured than you might expect. Generative AI systems can quickly produce lists of films or series, but they still require context about the viewer's preferences or mood.
"You can ask AI for recommendations," he notes, "but it still needs a starting point."
Social signals — recommendations from friends, communities, or trusted voices — may continue to provide that starting point, with AI helping to filter and organize options rather than replace the human layer entirely.
For businesses studying consumer behavior, the themes raised in this episode extend well beyond streaming media. Algorithms are powerful tools for organizing information, but social trust still plays a major role in how people make decisions. From entertainment to travel, restaurants, or product purchases, many choices are still influenced by recommendations from people we know.
The brands with a genuine edge are the ones that can see beyond the visible signals and understand what's actually driving decisions for the majority — including the 90% who never show up in the data.
