Flash News – Google and AWS Selling CDN Services for $2/TB

Categories

Live from NAB: Google and AWS are selling CDN services to a handful of companies for $0.002/GB ($2/TB). In other words, they are practically giving away CDN services to capture high-volume CDN media accounts. Google is aggressively pursing these accounts and winning the business. As usual, CloudFront is price matching Google. The $2/TB is below the cost structure of some CDNs even when settlement free peering is factored in.

The scary part – Google and CloudFront excel in media delivery, as in live and VOD. Google’s acquisition of Anvato and Amazon’s acquisition of Elemental Technologies are paying big dividends. We foresaw this happening last year, and this is the strongest indicator yet that media + video streaming pricing continues to decelerate.

Google’s strategy is simple. First, get experience supporting high-volume CDN media accounts at a low-cost structure. Second, roll out this cost structure to all Google Cloud customers. Finally, drop prices to the point where it drives a large chunk of the competition out of the market, which we believe happens at $1/TB. Google and AWS can afford to give away CDN services, because CDN is only a small piece of their cloud services portfolio. Plus, the cloud giants are introducing many high margin value-added cloud services like lambda, machine learning, GPU’s and edge computing.

Here are a few industry ramifications:

  • Google CDN and CloudFront will offer $1/TB by Dec 2017
  • Smaller competitors will be driven out of the media delivery business
  • Akamai will price match the cloud giants and offset it by selling other high margin features
  • Google and CloudFront will capture sizeable market share for VOD and live streaming by year-end
  • Next year, Google and CloudFront will develop sophisticated video processing capabilities, incorporating machine learning algorithms into video processing services

At this point, the best strategy to counter Google CDN and CloudFront in media delivery is to develop and introduce machine learning algorithms into existing video processing capabilities, which means going up the technology stack.

Scroll to Top