The Good Tech Companies - Why MP4 Video Is Broken for the Modern Web

Episode Date: January 5, 2026

This story was originally published on HackerNoon at: https://hackernoon.com/why-mp4-video-is-broken-for-the-modern-web. MP4 video is static and outdated for a dynamic web. Discover how runtime video models like MP5 enable interactive, personalized experiences. Check more stories related to media at: https://hackernoon.com/c/media. You can also check exclusive content about #video-for-modern-web, #runtime-video, #composable-video, #mp4-limitations, #data-driven-video, #mp5-video-model, #personalized-video-tech, #good-company, and more. This story was written by: @jonstojanjournalist. Learn more about this writer by checking @jonstojanjournalist's about page, and for more stories, please visit hackernoon.com. MP4 video was built for broadcast, not for the dynamic, data-driven web. As video becomes the primary interface online, static files can’t personalize, adapt, or respond in real time. Blings proposes a runtime video model (MP5) where video is assembled on the fly like software—interactive, composable, privacy-first, and personalized for every viewer.

This audio is presented by Hacker Noon, where anyone can learn anything about any technology. YMP4 video is broken for the modern web by John Stoy and journalist. The modern web runs on systems, not files. Pages and apps today are assembled at runtime. They react to user context, data, logic, and state. Video, however, while it's the most powerful medium online, is still delivered as a static artifact, an MP4 file that plays the same way for everyone. Blings was built around a simple observation that its CEO, Jonathan Schreiber, articulated during an interview with a J. Schneider of Loyalty 360 for leaders in customer loyalty. Industry Voices. One of the things I realized while working on video products over the

last decade is how not suitable for the web videos. That mismatch is no longer a creative inconvenience. It is an architectural problem. The dominant video format on the web was designed for broadcast, not computation. If you think about the video format behind the web, Schreiber continues. This is a black box. You can create it with any tool you can imagine, but once created, it's just a static black box. You cannot change anything in the video after creation. That model made sense for television.

It does not make sense for web where video is becoming the primary interface. By 2025, video is expected to account for roughly 82% of global web traffic, making it the dominant form of online content. The structural limits of static video. The limitation is not engagement. Video performs extremely well. Viewers retain up to 95% of a message delivered via video, compared to about 10% for text. The limitation is structure.

MP4 assumes a linear playback model, one timeline, one audience, one output. Any personalization occurs upstream, through pre-rendering variants or stitching assets. That approach does not scale and cannot respond to real-time context. How does a normal HTML website work? Schreiber contrasts it with the rest of the web. It's coded. This means I can program it to be relevant to each and every user. Video, by comparison, remains passive.

You stream a video, and everyone sees the same thing, Schreiber continues. It's not interactive. You're passive, a new layer in the web stack. Blings approaches this problem by treating video not as content, but AS infrastructure. Instead of exporting a finished file, video experiences are defined. as modular components like layers, scenes, text, and interactions combined with logic. The video is rendered at the moment of viewing on the user's device. Schreiber describes it as fundamentally changing what video is.

We're trying to build a new kind of video that is more data-driven, more live, more relevant to each and every user, more interactive. This shift turns video into something closer to software. It can accept inputs, respond to data, and branch dynamically, exactly like a web application. Blings refers to this execution model as MP5, a runtime approach where video IS created in real time rather than pre-rendered. The implications extend beyond flexibility into privacy and security. The video is created on the edge device, on the mobile phone that the user isopening. The video is created in real time when he's opening the video, Schreiber explains. We have

the ability to create those videos and we're not been exposed to the data, because of this, Sensitive data never needs to pass through or be stored by the video system itself. Composable video, assembled at runtime. In this architecture, video is no longer a single asset. It is a system-of-reaseable components assembled with logic. Schreiber offers a simple analogy. You can say that we do mail merge for video. But unlike traditional mail merge, this applies not just to text, but to structure and visuals. Schreiber continues, you can have like 10 different Lego blocks and create different videos as a combination based on some business logic. The result is scale without duplication.

A single template can generate millions of unique experiences without pre-rendering or acid explosion. AI fits naturally into this model, not as the core value, but as an accelerator. AI today is great in creating content, states Schreiber, but once you get to the level of thousands or millions, you cannot create today with AI millions of versions yourself. The leverage comes from the runtime, not the generator. Loyalty is a downstream effect. Loyalty programs benefit early from this shift because they are inherently stateful. Points, tears, milestones, and life cycle moments change continuously, and when video becomes part of the system, it can reflect that state instantly. Shriver notes how this plays out in practice. The video builds the ability

for any brand to send to massive users, but let them feel VIP each and every one of them. In controlled experiments, this infrastructure-level change has produced materially stronger results, including multi-fold improvements in conversion versus control groups. The outcome is not driven by novelty, but by relevance are rendered at runtime. From media to infrastructure, video already dominates attention and delivers ROI. Pages with video are 53 times more likely to rank on the first page of search results than those without. However, simply adding more video without changing its structure will produce diminishing

returns. What Blings represents is not a new campaign format or marketing tactic. It is a new layer in how video operates on the web. Shriver summarizes the shift implicitly by comparison. You don't even ask this question about a mobile app. The app is creating the page for me and for you. MP5 applies that same logic to video. MP4 will continue to exist where broadcast makes sense. But wherever video is expected to respond to users, data, and systems, a runtime model becomes inevitable. Thank you for listening to this Hackernoon story, read by artificial intelligence. Visit hackernoon.com to read, write, learn, and publish.