Jeffrey Gilbert raised an important and controversial question on LinkedIn about whether CDN economics are “broken” and whether big streamers should care.
I agree with Jeffrey that the CDN industry is in a difficult position. The pricing deflation we have seen is good in the short-term for streamers, but it poses a big risk of more fragility amongst the CDN supplier base. I don’t agree that the market is broken, but it is certainly in a transitional phase.
Here is how I see things.
First, there is no doubt that content comes first in Media business investments, so that means acquisition and production get the lion’s share of the budget. In a tough commercial market for media businesses that must continue to produce good content, the rest of the business gets put under big cost pressure, no matter how counter-intuitive this is to those of us working in what we think is the fundamental delivery chain that brings that content to the audience.
Second, when I consider “big streamers” I draw a clear line between global streamers and national streamers (i.e., national broadcasters).
- Global streamers – and there are not many, but they include YouTube, Netflix, Prime, TikTok, Disney+, and DAZN to name the main ones – need global strategies with local tactics. I’m not surprised that open-caching and private CDNs (as a change away from public CDNs) and ISP performance indexes resulted from global streamers. They needed to make things happen globally on their terms and not wait for public CDNs. These streamers can be considered to have reached a large stable state of scale and are generally shifting from subscriber acquisition to profitability (especially in VOD services, even if Live streaming is a big area of investment).
- National streamers – and there are a lot of them, now entering their 10-year period of massive streaming growth – need national strategies. For now, streaming is a small part of their total business, both in terms of revenues and costs. Many national streamers in developed countries have less than 20% of their total viewership and viewing peaks on streaming. Yet they are still significant enough to drive CDN pricing down. These companies have traditional broadcast, CableTV and IPTV distribution to think about, and are adding streaming to the mix while the costs of the existing networks are not necessarily dropping. It’s a double-whammy on costs for now.
I also draw a clear line between different types of CDNs in the market.
- The biggest public CDNs have large global networks that the video streaming delivery price war has overstretched and driven the consolidation we have seen. They have also been hit by the insourcing activities of the largest global streamers that have had to control their own destinies and have often asked for more capacity at a price-point than any public CDN could consider offering.
- The private CDNs have been around almost as long as the public CDN model, supporting the CableTV and IPTV industry for well over 20 years. But they have generally remained close to the traditional Pay-TV operators, and have not ventured into competing directly with the public CDNs. They know the Telcos well and are poised to be at the heart of the future platform for national streamers.
- Regional public CDNs that have focused on national-level specialised media delivery services are a breed between the two previously-mentioned models. They are “OTT CDNs” rather than Cable/IPTV CDNs, and yet they can offer private CDN as a service as well, which provides an important alternative to the global public CDNs.
- The build-your-own approach has also been of huge help to broadcasters as their streaming has grown and they have had to keep their costs down. Not every broadcaster can do this, but those that have managed their core daily volume of delivery in this way have reaped the benefits.
Then finally there are the all-important ISPs that all our content is delivered over. They are evolving to monetize their networks for large-volume Media delivery, given they must also invest more in their networks to reach the distributed scale that will be required as national broadcasters shift across fully to streaming.
All these system dynamics point me towards the following conclusions:
- Public CDNs will do what they can do best to secure their financial futures – global content delivery and supporting mission-critical content delivery services for all industries, not just media.
- Media delivery for national streamers, due to the scale of bandwidth it requires (estimates are at least 10x today’s volumes and peaks of viewership), is a natural candidate to be supported by highly distributed ISP-based CDNs. National streamers could have service provided by a CDN service provider spanning multiple ISPs, or by a mixture of ISP-owned CDN services, or a hybrid of this model. It is not clear to me if a build-your-own model will survive this approach (see further conclusions below). The multi-CDN model in use today would become a new mix of CDN types.
- Broadcasters / National streamers will need to see their traditional broadcast costs reduce significantly over time to spend a more sustainable amount on CDNs. Broadcasters have a largely fixed distribution cost for broadcast as they maintain their reach to the national audiences for TV and radio services. So they are under pressure to keep their streaming costs under control, even as volumes of content delivered grows every year.
- To reduce traditional broadcast costs significantly probably requires a switch-off of DTT and DTH services. This is complex and political, requiring governments, regulators, and broadcasters to determine how to reach their populations so no one gets left behind or is unfairly discriminated against. I expect the timeframes for this in countries already obviously moving to a streaming first approach for national broadcasters could be 10+ years.
- If we imagine the world once DTT and DTH are switched off (or at least their costs are not borne solely by the broadcasters), then we will probably have a CDN delivery model that costs about 20-30% of today’s broadcast delivery model. This would mean a healthy cost reduction for streamers, a big growth for CDN revenues, an increase in revenue for ISPs, and a shut-down (at least for Media use cases) of nationwide DTT and DTH networks.
- We will probably have a CDN delivery model that is treated as Critical National Infrastructure (CNI) – just like telco and broadcast networks are today – because when the news or emergency alerts or national events must reach a whole population, it must be guaranteed. This also creates a very strict set of security requirements to protect these networks against cyber-attacks. It is highly likely that this CDN infrastructure/service will be regulated.
- To find economies of scale for streaming, national streamers could combine forces to share the costs and use of the CDN they heavily utilise between them, and often divide at different times of the day. This would look much like how the current broadcast networks operate.
- The CDNs that make it through the transitionary period will need to be focused purely on media needs, probably evolving to manage both content delivery as well as application delivery. They will need technical, operational, and commercial models that can manage the complex daily peaks and troughs of consumer viewing as well as the one-off high-peak needs that happen infrequently throughout a year.
For me the CDN model isn’t broken, but there are some behaviours in the industry that will lead to more CDNs going out of business.
I believe the Public CDN market is transitioning so obviously because streaming is growing so quickly and total video delivery costs are increasing for a media industry that is not making more money than before and yet must accommodate these extra costs.
We simultaneously have a telco industry that must scale out their core networks to accommodate a near-term future of 10x growth in video traffic for broadcasters which cannot be accommodated without significant investment.
The CDN market is most definitely under pressure. I think we are at a tipping point when the Public CDN and Private CDN models will start to take different shares of the total media streaming market, where national streamers follow the examples of the global streamers and create a new type of multi-CDN model. It will mean that Public CDNs limit their capacity for national media use cases and focus on global content delivery and the industries that meet their ideal customer profiles for the right ROI for them.
For the longer-term I think the real investment in national CDN platforms, that will then continue scaling up for higher bitrates and more immersive viewing experiences, will only happen when DTT and DTH services, and their associated costs, are significantly reduced or even stopped. Today, the signs are showing that we’re really starting to move in that direction, but the tunnel to reach that point could be long. Only those CDNs most focused on and geared for the needs of the Media businesses will make it through to the other side.