Have you ever wondered what goes on behind the scenes of your favourite websites? How do they deliver their content, protect their data, handle traffic spikes and collect analytics? In this blog post, we set out with a simple goal: to check out how different kinds of websites are set up and what kind of data they collect from us.
For this we’ll use Overmind, and their visualisation tool called explore that lets you explore the public network infrastructure along with your own cloud environemnts. To generate the graphs we will need to run a new query on the selected url or urls. We can then increase the link depth by double clicking to help expand the graph and see any related resources.
We’ve chosen four popular but very different websites: BBC News, US Gov weather, Facebook, and Dunelm.
BBC News
The BBC News site maintains an uncluttered digital presence, using only seven trackers. This indicates a measured approach to data collection. As they are public service broadcaster they will not rely on trackers and ads as much as commercialised broadcasters. Looking at the site infrastructure, we can see the core hosting handled in-house, while using AWS for the home of their analytics solution and Akamai for their (CDN) content delivery network.
Checking their uptime you can see that last experienced issues in November 2023, showing that they have a measured but controlled grip over their site with minimal reliance on third parties or vendors other than AWS & Akamai.
US Gov Weather
As expected, the US government's weather website operates with minimal tracking technology, employing only essential tools for performance and analytics purposes (universal-federated-analytics running in AWS & google-tag-manager in GCP) withe the actual webpage itself being served through Akamai. This aligns with regulatory standards and a commitment to user privacy.
Despite an outward appearance of zero trackers via Brave's browser detection, Facebook's network is actually a web of hidden complexity. This closed ecosystem, typical of leading tech companies, includes an integrated network that combines their various acquisitions, including Instagram and Oculus. Apart from DNS entries and HTTP’s we are unable to see if there’s any underlying third parties or vendors.
Dunelm
Dunelm (Largest UK homeware retailer) uses a lot of trackers, 74, to be precise. While this might sound excessive, these trackers cover everything from keeping track of visitor behaviour for analytics to managing secure payments and integrating with social media platforms. Each of these tools plays a role in creating a better shopping experience ensuring you get a personalised experience while shopping for homewares. Using tools like Datadog RUM to capture users’ journeys. Tools like Hotjar and Tealuimiq allow them to do things like A/B testing ensuring they continually optimise their strategies.
Even with all these trackers, which in turn are reliant on all three of the major cloud providers, Dunelm manages to keep their website loading quickly It's a good sign that they're able to balance a heavy set of tools without slowing down the shopping process for their customers.
Try the Free Tool for Yourself
Overmind's Explore is free for anyone to try, and it's all about discovering information like this on your own. It was built on Overmind's underlying graph tech & dependancy mapping and can be used on things like AWS & Terraform.
We’d love to see what you uncover, share your own graphs in our Discord community.
.gif)
