So, November 2024 had a couple of interesting events. The 45th POTUS won the 2024 US election to become the 47th POTUS, the world’s first wooden satellite was launched, Beyoncé became the most nominated artist in Grammy history and oh, there was a boxing match between a 27-year-old YouTuber and a 58-year-old renowned pugilist. That’s right, Jake Paul and Michael Tyson battled it out in Arlington, Texas. And the whole thing was being streamed on Netflix, which had live-streamed a couple of things, such as a Christopher Rock stand-up special and a roast of NFL player Thomas Brady.
But, this time was different.
This bout was hyped up, there was trash talk, there was a BTS show (not the K-pop band, the behind-the-scenes of the training), many had commented on how Tyson would eviscerate Paul & there were close to 110 million total live viewers around the world with about 65 million concurrent streams and roughly 70,000 physical spectators.
But, the biggest blow that night came from the streaming platform, not from either boxer. The live broadcast seemed to be marred by technical chaos.
The screen just wouldn’t stop buffering and there were said to be a lot of lags. Even though Paul ended up victorious that night, people felt like Tyson at the end: worn-out. It’s said that over 90,000 users reported problems on Netflix at its peak. Imagine if it was buffering whilst Tyson mooned the world. Many felt the fight was anticlimactic and maybe, so was the viewing experience. Someone in Florida even sued Netflix for what they deemed to be a shoddy performance.
Netflix had a similar issue when fans were left waiting for more than an hour to see a live reunion for “Love Is Blind”, which the platform was supposed to live-stream. And the NFL was said to have partnered with Netflix to stream its Christmas matchups. Whoa.
So, how could this happen? After all, Netflix is, like, one of the biggest streaming platforms in the world, if not the biggest. With all of its subscribers and money rolling in, how did it hinder a seamless viewing experience from taking place? Especially, when platforms, that may not have the size or scale of Netflix, were able to pull it off?
Peak concurrency is the highest number of viewers during a live-stream. The World Cup final between India and Australia was said to have a viewership of close to 60 million on Disney+ Hotstar. Before that, the league match between India and Pakistan was said to have a peak concurrency of 35 million. And during the CSK vs RCB match of IPL 2024, it was said that about 113 million viewers tuned in to see it being live-streamed on JioCinema. With RIL acquiring Disney+ Hotstar, who knows what future live-streaming numbers could look like?
All of these camera feeds have to come together to weave a cohesive narrative. How is JK Simmons orchestrating this symphony in such a way that the live-stream doesn’t rush or drag? For a cricket match, for instance, there are manned, unmanned and aerial cameras to capture every angle & these cameras send the raw footage to the production teams that are on-site. There, the teams work live to create the broadcast by choosing the best camera angles, adding commentary and overlaying the graphics. On top of that, there might have to be dubs and ad placement, which are said to be done by ad markers being inserted by human operators. To resume the gameplay seamlessly after an ad might not be easy and yet, it might be a split-second decision. And then, when the video is ready, it would have to be converted to formats for TV or phones or other devices. It might be automation x human oversight at play.
There’s something called CDN (Content Delivery Network), which is a couple of servers distributed across multiple places to deliver digital content to users in a reliable way. A single CDN might not be able to handle the demands of a marquee event, like an IPL or the Tyson vs Paul fight. This way, by having multiple CDNs, the traffic load is said to be spread out, so there’s backup, in case one network fails and video quality is maintained. Plus, there’s a war room staffed by engineers, content folks and leaders. If there’s a sense that a certain moment during the live-stream would cause a sudden spike in traffic, infrastructure would be scaled up beforehand to make sure the system is able to deal with the extra load. And then, there would be post-even analysis.
Sometimes, people don’t appreciate the work that goes into the things they effortlessly consume.
Somehow, Netflix still had trouble handling the traffic. Would it be fair to cite not enough co-ordination between technical teams or inadequate scaling of infrastructure?
Netflix may be used to content people consume at their leisure, but Disney+ Hotstar and JioCinema might have refined their offering to a base that needs that real-time viewing. Could it be that Netflix’s architecture is more optimized for static libraries? Maybe, being used to passive content doesn’t require “live resilience”.
After all, maybe, Netflix has created and shared new worlds for people to immerse themselves in, but maybe, the live arena is a leviathan of its own. Would something be done differently in terms of what live-streaming infrastructure is invested in? Or whilst Netflix has played some role in digital disruption, is live-streaming something of an overreach for the streaming giant? Maybe, it’s a tale of caution for them and this may be a forgotten mistake, as they continue to evolve. Similarly, are there times when a startup overestimates its readiness? Assuming they have a forte in a new sector, since they dominate another? As a startup diversifies, how can it make sure they’re not stretching resources too thin or diluting its core value proposition?
It’s not like this is a foreign concept to someone who’s built a startup. Sometimes, entrepreneurs may not be all that technically prepared or they may face customer dissatisfaction, even to the point where it hurts the brand. Is subscriber trust a fragile commodity for a streaming platform? Maybe, people aren’t going to get off Netflix just yet, after all, how would they rewatch “Breaking Bad” or “Seinfeld” or “Better Call Saul” or “House Of Cards” or “The Good Place”? But, could it mean that other parties may not want to go with Netflix for a live-streaming route, just in case?
Think of all the established players going the q-comm route after seeing the success of Zepto or Blinkit or Swiggy Instamart. But, just like Netflix’s live-streaming, these new q-comm entrants would have to weather high traffic, literally. Maybe, this is the blitzscaling you hear of: going for rapid growth without necessarily having the commensurate investments in backend infrastructure. Sometimes, it could work, like with Reliance’s Jio, other times, it may not, like with Zomato having to liquidate its subsidiaries and exit almost all major overseas markets.
Where will you like the next anticipated event to be live-streamed? And will you appreciate the intricate infrastructure a bit more the next time you press “play”?
