Lets talk Digital Sustainability

One of the primary reasons that the people behind Wholegrain work here is a want to do better – that means build more sustainable websites, consider the needs of people more centrally to our decisions and play a leading role in moving forward our understanding of what sustainable web design really means.

I can imagine this to be true for anyone working in any sector where they consider and challenge how to do it more sustainably. It can be frustrating, constantly looking for the definitive end line that says ‘this is how we do it perfectly’, but we all know that we are not there yet and to be truthful, may never be there. Not that this should stop us trying.

At Wholegrain (and in all other businesses and communities we stand beside in this space) we are continuously reading, writing, listening and sharing the latest ideas and frameworks that can push us all collectively closer to that golden state of understanding. Learning from each other and communicating new ideas is what has built the understanding we have today and it often leads to the need to pivot or unlearn what was thought to be true but now is proven incorrect. Again, not that this should stop us trying.

This past week the overall approach to Digital Sustainability was challenged, sadly not through conversation.

The following is the thoughts of Andy, who manages our Curiously Green newsletter and spends his time in and around the wider thinking on Digital Sustainability with input from Bailey, our new business Manager and Dan, our Technical Team lead – who both speak internally and externally about what Digital Sustainability looks like on a tangible level, as people on the front line of actually designing and building sites to be more sustainable. 

The ever evolving nature of digital sustainability means that we’re constantly reading and consuming articles, reports and ideas on the topic. We keep a careful eye on developments in coding, file formats, data gathering, storage optimisation, changes to calculation models and philosophical ideas about digital usage Our website carbon calculator, Website Carbon often features in these articles in these topics, sometimes acting as a lightning rod for thoughts and findings about low carbon web design.

The most recent example is a report from the BBC called “Does what you scroll burn coal? Mythbusting energy consumption on the web”. The post on the Research and Development subsite has caused some consternation in digital sustainability circles.

The article looks at two areas:

• The effect of dark mode on device electricity usage and user behaviour
• Correlation between performance metrics and data transfer and energy usage

The dark mode research is interesting and goes against some of the sustainable web design advice and methodologies. In their research a small sample size of participants turned up monitor brightness levels up when using dark modes, regardless of overhead lighting conditions.

This rebound effect is potentially a useful finding. If it can be replicated across a larger sample size it would certainly change some conventional wisdom and design methodologies. We would add that any dark modes implementation should be in line with accessibility guidelines and based on user preference rather than prescriptive.

As the researchers note, further research would be required using devices with OLED screens, which is something that is noted in the Web Sustainability Guidelines.

The rest of the report has frustrated us at Wholegrain and others in the community. The sections on performance and data transfer feel to have taken an over simplified view on areas that the digital sustainability community treat as nuanced and complex.

The report states that “performance and energy efficiency aren’t always linked, so optimising for performance may not yield the desired energy results”. This is something that we would agree with. Indeed we wrote about how strong performance metrics don’t always correlate with low site emissions in November 2023. 

Without knowing what sites were tested, it’s hard to comment further. What we can say is that we and other reputable agencies know that “Developers cannot therefore simply optimise for performance and hope that energy savings occur as a happy by-product”. Suggesting otherwise feels harmful to the industry and undermines trust and engagement at a time when digital emissions continue to rise.

This feeling of being unduly undermined continues when the researchers discuss data transfer. The researchers question whether a link can be made between data transfer and device energy use. The report raises issues where “data is used as a direct proxy for energy” going further to state that it presents a “completely fictitious narrative of how emissions are generated on the internet”, citing Website Carbon as one place this narrative is shared.

Them’s fightin’ words.

It’s widely known and acknowledged within the sustainable web industry that data transfer is a flawed proxy for carbon emissions. I think Fershad Irani’s piece on this issue is the most eloquent and persuasive on the topic. Put simply, we know there are issues and that in certain cases it might cause some paradoxes. Where those paradoxes are proved we take note and adjust best practices.

The report discusses testing “top sites” but not what they were, how they are built and what platforms or frameworks they use. Picking holes in the sustainable web design practices without testing sites built using them doesn’t feel fair or representative. Indeed they tested the COP28 site which did an infamously bad job of implementing the “low carbon mode” mentioned in the report. The site (which was not built by a sustainable web specialist) was criticised heavily at the time and revisited when updates were put in place. Citing this website as a good example of best practice is incredibly frustrating.

In the conclusion of the BBC piece the researchers write – 

This is pretty insulting to the careful thought and dedication brought by contributors to the Web Sustainability Guidelines. Agencies like MightyBytes and Wholegrain, organisations like the Green Web Foundation, SustyWeb and individuals like Alexander Dawson, Fershad Irani, Michelle Baker, Hannah Smith and many others work tirelessly to bring nuance and suitability to the interventions prescribed. 

We’ll end this response in the same way as the report does, with some suggested next steps. The researchers express a desire to collaborate with the industry to develop “tools that allow them to accurately estimate energy”. A good place to start might be to collaborate with an organisation like GreenSpector who share similar aims. 

They could also ask to collaborate with any of the organisations and individuals mentioned above. We’re an open bunch who desperately want to do the best thing for people and planet in every way. We’re open about what we do and why we do it. We showcase our sites and think they hold up to scrutiny. We’re open to making changes if our methods turn out to be flawed. We want to iterate and improve where we can. We want accurate measures just as much as anyone else. Maybe even more so.

Our mission to “accelerate the shift to an Internet that’s good for people and planet” needs collaborators so we can continue to do the work we do for the companies we work with – one being BBC Good Food, just for reference.

Let’s talk.