Last month, YouTube experienced an outage for about an hour and the internet got a taste of what would happen if the platform disappeared. The results were surprising: Internet behaviors shifted immediately and fiercely. There was a huge overall traffic surge, with some of the largest increases seen on app and search traffic.

Chartbeat analyzed the YouTube outage using global traffic data across a sample of more than 4,000 sites. Overall, this outage resulted in a 20% net increase in traffic to publisher sites. Just over half of this increase (11% of overall traffic) went to general articles on publisher sites, while articles about the YouTube outage comprised a 9% lift.

Huge surges in search and app Traffic

It makes sense that since 45% of the traffic lift came from articles about the outage, we saw a large boost in search traffic, with readers likely searching for answers regarding the outage.

Specifically, we saw a 59% increase in search traffic. Across other referrers types, we saw a consistent lift—notably also from platforms within the Google ecosystem:

• Direct +9%
• Links +28% – of total link traffic during the outage, Google Chrome Suggestions made up 19% and Google News 14%
• Internal +13%
• Social +11%

As we look at how and where people read during this outage, we saw app and Google AMP traffic seeing the largest surges, with 78% and 67% lifts respectively. We also saw boosts across desktop, but mobile and tablet saw even larger increases.

• Web +15%
• Facebook Instant Articles +6%
• Desktop +13%
• Mobile +26%
• Tablet +26%

Not all platforms are created equal

We compared this traffic boost to the Facebook outage on August 3rd, 2018, which brought a 2.3% net increase to publisher traffic in the 45-minute outage window. Similarly, the late August Reddit outage didn’t even make a blip to overall publisher traffic. In the Facebook case, only a negligible amount of that traffic went to articles about the outage.

There are a couple potential reasons for the vast difference in traffic increase in the YouTube outage vs. the Facebook outage.

1. YouTube is not normally a traffic driver to publishers: Unlike Facebook, YouTube does not refer a great deal of traffic to publishers. Thus, the effect of people moving from YouTube to publisher sites during the outage had a dramatic impact and was purely additive.
2. Day and time difference: Facebook’s outage occured on a Friday afternoon (US time) / evening (Europe/Asia time). Thus, the majority of people were likely at work or out for the evening. It wasn’t prime news-reading time. The YouTube outage, on the other hand, was on a Tuesday evening (US), which is prime couch time.

So far, we’ve seen there’s no single reaction when a platform goes down. Sometimes people are more apt to search for answers, sometimes they go directly to a news source they trust. The one thing we do see is that when Facebook or YouTube goes dark, the rest of the internet comes alive.