Eight years ago we built a site that weighed sixty-three times more than today's. We didn't measure it. The field existed. We weren't in it.
In June 2025, the Website Carbon Calculator measured circostudio.com — the site we kept for almost eight years without touching — and issued an F rating. 3.16 grams of CO₂ per visit, worse than 94% of sites measured globally. Hosting energy was renewable. The problem wasn't there.
In 2026, after rebuilding the site from scratch with explicit efficiency criteria, the same calculator measured circostudio.io. Rating A+. 0.05 grams per visit. Better than 95%. Same root domain. Same renewable hosting. Sixty-three times less CO₂ per visit. The difference is entirely architectural.
When we built the previous site, in 2017, we did not study sustainable web design. It was not part of the craft we practiced. The field already existed: the Sustainable Web Design methodology had been publishing versions since 2013, Wholegrain Digital had opened the public calculator shortly after, The Green Web Foundation had been building its directory of renewable hosting. None of that entered our technical decisions.
It's not that we thought it didn't matter. We didn't think about it. Active indifference is a stance — you look at the problem and dismiss it. Structural ignorance is something else: it is not knowing there is something to think about. And that is what happened to us for eight years.
The technical knowledge was available. What was missing was circulation within the professional sector. Sustainable web design did not appear in tender specifications, was not among acceptance criteria for deliverables, was not taught in systems engineering programs, was not part of the commercial proposals of neighboring consultancies. In the concrete practice of building software for companies, the weight of a corporate page was not a parameter anyone watched.
This does not exempt us. It explains it. And naming the explanation matters in order not to repeat the pattern with the next omission the sector hasn't discovered yet.
What changed was not technological. The technology to build lightweight sites existed in 2017. What changed was the criterion: starting any technical decision by asking what it consumes and what for.
The current site weighs sixty-three times less than the previous one because we made explicit decisions at every layer: plain HTML, no frameworks, no third-party libraries, no images except where indispensable, no trackers, fonts with strict latin subset, animations that respect reduced-motion. Each of those decisions is documented in the technical sustainability report. None is heroic. All are mundane. The only thing that took effort — because we had to decide it ourselves without anyone asking — was declaring that they mattered.
A consultancy that remakes itself in silence learns alone. A consultancy that remakes itself in public gives material to others so they can remake themselves too. If publishing this data — including the data of our previous ignorance — helps an IT team ask the basic question about the weight of their corporate site, it was already worth it.
We are not saying we now know everything. We are saying we learned this, that the next ignorance — the one we haven't yet discovered — will also surface at some point, and that when it does, we will publish it too. Self-criticism only matters when it stops being internal.