How We Test VPNs: Methodology
By this point, you’re probably wondering what qualifies us to review the best VPNs in France and what methodology we used to determine which VPNs made the cut. The first thing we do is start at the beginning: we make sure it has all the features you’re looking for. For the best VPNs in France, in particular, we considered the amount of servers in the country each VPN provider has. Variety of server locations is important because proximity matters when it comes to the speed of each server; the closer you are to a server location, the better the connection will be.
We don’t believe in simply repeating information we’ve gleaned from scrolling through each VPN provider’s website. Instead, we personally test each provider across iOS, Android, MacOS and Windows devices to ensure they can live up to advertised expectations on all platforms. While researching the features a provider offers is essential, we also go deeper than that. We examine each individual company, where it’s based, and whether or not the headquarter’s country is a member of Five Eyes, Nine Eyes or Fourteen Eyes, those pesky intelligence agencies that can force VPN providers to share their customers’ personal information.
As we do with all of our reviews, we tested each VPN in our Brooklyn office on a private Optimum network. Using one device at a time, we set a baseline by judging our internet speed without a VPN, and use the findings as the control group for our research. We understand that most people are firmly planted in the Apple or Windows camp, so we test the VPNs using a Macbook Air and a Windows Vivobook so we can tailor our recommendations based on your operating system of choice.
Our testing method is quite simple: we test download and upload speed along with ping (latency) with and without a VPN. While download and upload speed are measured in megabits per second, latency is measured in milliseconds. Once we have these figures, we calculate the difference between the two measurements as a percentage, because of how greatly the speed differs between the operating systems. Though VPN speed varies based on factors such as distance to the server, operating system, device used, etc., VPNs should have no more than a 40 percent difference in any of the speed categories. While we try to provide the most transparent results and information possible, we urge our readers to conduct the speed tests themselves so the results are more accurate.
Most people who opt for VPNs are looking to protect their web traffic, which is composed of domain name servers (websites visited) and their respective IP addresses. We take that a step further and want to make sure that users’ private IP addresses are not leaked due to WebRTC, which essentially enables default browsers, like Chrome, Microsoft Edge, and Opera, to directly communicate with each other.
We use DNSLeakTest.com to test for DNS leaks by comparing our computer’s IP address with the IP addresses that come up on the website. If we’re using a VPN and our personal IP address pops up, the VPN is leaking our data. After that, we test for WebRTC leaks on ExpressVPN’s website, so we can look at the local and Public IPv4 IP addresses to determine if there were any WebRTC leaks.
Once we’re finished testing for DNS and WebRTC leaks, we focus our attention on each VPN provider’s privacy jurisdiction and data-logging policy. We try to avoid companies that are based in areas that fall under Five Eyes, Nine Eyes, and Fourteen Eyes jurisdiction, international surveillance alliances that could legally force companies to hand over customer information.
In terms of data-logging, we look for companies that don’t hold any overly-sensitive information and only keep it to a user’s basic account information, such as their name, email, and payment information. We don’t think that it’s necessary for VPNs to keep any information about when customers use their VPNs, for how long, or their web activity and IP addresses.
For our longer reviews, we examine the VPN’s encryption methods and Internet protocols to ensure they adhere to industry standards, which include AES-256 and OpenVPN, the most secure methods around. Additionally, we search for anonymous and dynamic IP addresses, because they’re difficult to track. Finally, we make sure that each VPN has a kill switch, which will shut down all Internet browsers in the event that the VPN fails.
We all love streaming our favorites shows and films, so we look for VPNs with Netflix and torrenting access. We also look for split tunneling, which gives users the opportunity to access their VPN and public network at the same time. This can help increase speed by using lower bandwidths. We also prefer double or multi-hop VPNs, as they encrypt data multiple times through multiple servers.
In addition to providing readers with unbiased information about the subscription options each company offers, we compare their pricing to that of their competitors so we can determine if a certain company is charging a reasonable rate or not. The price range of VPNs varies, but it’s nothing too extreme — VPNs can go from free options (yes, free!) up to $15 a month, but on average it’s around $5 to $10 a month.
While monthly plans offer users more flexibility, annual plans often come with discounted prices. Additionally, we provide information about any trial periods or money-back guarantees that a company offers. Finally, we put a subscription’s offerings into the clearest terms possible to let you know what each package entails, in terms of the numbers of server switches, simultaneous connections, and devices allowed. Typically, VPNs allow for unlimited server switches and devices and anywhere from one to unlimited simultaneous connections.
Most VPN companies provide a free live chat feature, in addition to any email information or ticket slips to feel out. Though the customer service typically isn’t the best with VPN companies, we prefer 24/7 live chat and award extra points if the company provides a phone number.
We test the company’s VPN app on both iOS and Android-run devices, as well as learning their ratings; anything below three stars on the device’s respective app stores doesn’t get recommended to users with that device.