The pursuit of green web analytics

The answer to this question isn't as black and white as I'd wish. When I started Cabin in 2020, running entirely on renewable energy was one of our core aims. We relentlessly researched and tested different ways to power our service without fossil fuels. The available information about cloud infrastructure was hazy to say the least. The energy source also had to be reconciled with how much impact it would have on latency and network hops, which add their own energy consumption.

What does an analytics service need?

An analytics service needs to handle a lot of traffic. That traffic comes from all over the world and can be wildly unpredictable. It needs to reach a database with maximum efficiency and reliability.

Three things improve efficiency:

1. Edge nodes - positioned close to the visitors of your sites, reducing network hops and latency
2. A queue system - buffering writes to reduce the load on the database and ensuring data is processed reliably
3. A low-latency, reliable database - capable of handling bursty write patterns without dropping records

Firing up more machines across the world to collect traffic may seem like burning more energy, but it actually reduces the distance between the visitor and the collection point. Fewer network hops means a faster, lighter request. Overall, the net effect is a reduction in energy consumption compared with routing everything to a single server in one location.

What we learned about running lean

At Cabin we learned that these edge instances can be the lowest spec possible. Low CPU and memory instances, carefully optimised, can handle millions of requests without breaking a sweat. The code running on these machines is easy to benchmark, and any code changes can be analysed for their performance impact before they ship.

These nodes scale up and down automatically. If one goes offline, traffic is re-routed to the nearest available node within seconds while the original instance attempts to recover.

I receive pings on my phone if a server ever goes offline. Witnessing one of these alerts used to make me panic and jump on my laptop. I've since learned to do nothing - it fixes itself in minutes while traffic is seamlessly redirected.

Cost as a proxy for energy

Though it's hard to reliably track energy consumption in the cloud, you can use cost as a loose proxy. If you reduce costs in your cloud infrastructure, you're most likely reducing electricity consumption and therefore shrinking your carbon footprint. This has been our guiding principle: every cost optimisation is an energy optimisation.

At Cabin, our infrastructure cost per million pageviews has dropped consistently year on year, not because we're cutting corners, but because we're obsessively trimming waste.

The big cloud providers

The three major cloud platforms have very different track records on transparency.

AWS

AWS have historically been opaque about their energy reporting. They made broad claims that certain regions were "green" but didn't provide granular detail or independent verification. In recent years they've improved, publishing sustainability pledges and region-level carbon intensity data. But it still requires significant effort to verify that a specific workload is running on genuinely renewable energy.

Cabin runs on AWS, primarily in the eu-west-1 (Ireland) region for data storage and processing. Ireland's grid has a strong renewable energy mix, and AWS Ireland has invested heavily in wind and solar. Our backup region is London (eu-west-2), and our edge nodes use latency-based routing across four global regions.

Azure

Azure have been more transparent about their energy usage. Their sustainability dashboard provides region-level carbon intensity data, and they publish a detailed sustainability report with verified figures. They've committed to being carbon negative by 2030.

Google Cloud

Google Cloud have arguably led the pack on transparency. They've published per-region carbon-free energy percentages since 2021, and their Carbon Footprint dashboard gives customers direct visibility into the emissions of their workloads. Google also claims to have matched 100% of their global electricity consumption with renewable energy purchases since 2017.

Honesty over greenwashing

The uncomfortable truth is that no cloud service can guarantee that every electron powering your workload came from a renewable source. The grid is shared. What providers can do is purchase renewable energy certificates (RECs) and power purchase agreements (PPAs) to match their consumption.

At Cabin, we're honest about this. We choose regions with the highest verified renewable energy mix. We optimise our code and infrastructure to consume as little energy as possible. We track the carbon footprint of the websites we serve, giving our users the data they need to reduce their own impact.

We believe the pursuit of green web analytics isn't about perfection. It's about making deliberate, transparent choices and constantly improving. Our 1.2KB tracking script, our lean edge infrastructure, and our commitment to EU-based processing in high-renewable regions are all part of that pursuit.

What you can do

If you care about the environmental impact of your website:

1. Measure it - You can't improve what you don't measure. Cabin's Energy & Carbon dashboard shows your site's data transfer, CO2 emissions, and per-page breakdown.
2. Optimise heavy pages - Large images, uncompressed scripts, and unnecessary third-party tags are the biggest culprits. Cabin identifies your heaviest pages so you know where to start.
3. Choose green hosting - Use our How Green Is Your Website? tool to check whether your host runs on renewable energy.
4. Use lightweight tools - Every script you add to your site has an energy cost. Cabin's 1.2KB footprint is 77x smaller than Google Analytics.

The web is responsible for roughly 2-4% of global carbon emissions - on par with the airline industry. Every optimisation matters. And it starts with knowing where you stand.