Content vs. Eyeballs

For years, there's been a chicken-and-egg between eyeball networks and content providers, each saying there's no point in enabling IPv6 until the other side did. Users don't need IPv6 until there's content. Content doesn't need IPv6 if nobody will see it. And occasionally the consumer electronics folks would blame both of them.

World IPv6 Day in 2011 gave content providers a chance to try IPv6 together, without worrying about losing customers to the other guys, because they were also trying it. It worked well enough that we tried World IPv6 Launch in 2012, when content turned on IPv6 permanently, and ISPs tried to get enough users enabled for 1% of them to reach content over IPv6. Finally, content and access were working together to enable IPv6.

Since then, ISPs have continued to turn up users, and more ISPs have added themselves to the list, to see where they stand in the relative rankings.

Google has also continued measuring eyeballs, and Dan Wing has measured content. I'm also indebted to Eric Vynkce, for tracking Google's per-country data over time.

So, where to point the finger now?

Globally, ISPs have a long way to go. Some countries are doing very well. The U.S. is better than average, and significantly, the percentage of eyeballs using IPv6 in the U.S. is about the same as the percentage of the top 25,000 web sites using IPv6. Seems like neither industry sector can point at the other one.

More of the top 1 million sites have IPv6 enabled than the top 25,000. It looks like some domain squatters and a few web hosting companies have just enabled IPv6 for all of their domains or customers. To get more content, that will be important.

I'm not going to say that either content or eyeballs can ease off the pedal. But with the announcement of IPv6 support in the Xbox One and some blu-ray players, we're looking at you, consumer electronics.

Lee Howard
18 November 2013