How do you detect when and where products and services might be inaccessible?
For every product, Google logs the number of visits it receives in a given time period, along with an approximation of the geographic region where the request came from. If there is a significant decrease in traffic to a product or service from a geographic region, that might mean that users can’t access that product or service. Our code looks at worldwide traffic patterns over time and detects significant changes. We also receive inquiries from journalists, activists or people on the ground, prompting us to look into our graphs.
Why do product outages or disruptions occur?
Users may be unable to reach our products for a variety of reasons, ranging from network outages to government-mandated blocks.
Are the lists of current and known disruptions comprehensive? Are they updated in real time?
We add events to the lists after we detect or discover significant drops in our traffic graphs and obtain related information from government sources, news outlets or ISPs. Since we update these lists manually, they are not updated in real time. There may also be events we haven’t discovered in the graphs yet.
If traffic to a product is disrupted in a geographic region, does that mean that no one in the region can use it?
Not necessarily. Depending on the cause of the disruption, it may prevent the vast majority of people in a region from accessing a product (like YouTube in China) or restrict access for certain segments of the population (like all of our products in Kenya in February 2012).
How do you generate the graphs?
The graphs give a representation of the ratio of that region's request rate to the worldwide request rate. For each time period and product, we divide the traffic for each geographic region by the total worldwide traffic, giving a number between 0 and 1. Then we multiply all numbers by a constant, which normalizes but does not change the shape of the graphs.
How do you know what geographic region traffic is coming from?
Determining the geographic origin of Internet traffic is not an exact science. We use a combination of third party data and our own technology.
Do the graphs reflect actual traffic volumes?
No, the numbers are only for comparing time periods within each graph. If the number goes from 10 to 20, that means it's doubled, but gives no indication of absolute volume. Numbers between different graphs cannot be compared. Keep in mind that what is being graphed is a geographic region's portion of Google's worldwide traffic. If it appears that traffic in a given region is decreasing over a long period of time, that doesn't necessarily mean that the amount of traffic we receive from that geographic region is actually decreasing. It just means that the selected geographic region's traffic is growing more slowly than the worldwide average. The graphs show product use over time, so normally you will see a similar pattern every week. For example, in this graph you'll notice that on weekdays there's usually more traffic than on weekends to Gmail from France. What catches our attention is when there's disruption to the normal graph pattern—for example, Google products were largely inaccessible in Egypt in January 2011 and in Syria in November 2012 when their respective governments halted Internet traffic during periods of political unrest.
Why is data available for some geographic regions but not others?
We don’t display traffic graphs for regions where there is not enough traffic to our products.
Why are you releasing this information?
By providing real-time information about traffic to our products, anyone can see if our products are accessible in a country or region. We hope people will be able to use this data to analyze and better understand the state of information flows online.
How do I get a copy of this data?
We currently don't have a way for people to download the data in our Traffic feature. If you believe the data would be useful for your research, please let us know by outlining how you'd benefit from it by using our "Send Feedback" link. Before sending your request, please review this FAQ which explains how we normalize the data displayed in our graphs so you can decide if it will serve your intended purpose.