At a time when global tension over energy is regularly in the headlines and ecological pressure to reduce emissions is accelerating, the digital transition is seen as a means of reducing energy consumption in a large number of sectors (IT for green), and it is even considered that it will not be possible to control climate change without massive use of digital technology.
But what is "digital" doing to reduce its own environmental footprint? What are organizations doing to control their technological environments?
According to Arcep, the digital sector accounts for 4% of the world's greenhouse gas (GHG) emissions, with a 60% increase forecast between now and 2040. In France, it is estimated that the digital sector already consumes 8.5% of the country's electricity.
To try and control this increase, The Shift Project (report "For digital sobriety") proposes a "Lean" approach: "Buy the least powerful equipment possible, change it as little as possible and reduce unnecessary energy-consuming use". The aim of this approach is to counteract the fact that the production and use of digital equipment is a major contributor to greenhouse gas emissions.
But in this data mode, it's also time to face up to the ecological impact of data management. For years now, we've been witnessing an explosion in data flows (to the point where data can account for 80% of traffic on some operators' networks).
This increase in traffic is accompanied by a rise in the volume of data stored in data centers, which are as indispensable as they are energy-hungry: without them, it's impossible to run digital businesses, and in return, they will consume up to 13% of the world's electricity by 2030.
This headlong rush is contributing to a doubling in the volume of corporate data every 2 years. Double the data, but for what benefit? What are companies doing with this Big Data? What information is really useful to their business model, and is it easy to use?
Jean-Marc Jancovici, head of The Shift Project, often highlights three approaches to saving energy: efficiency, sobriety and poverty. Let's try to apply these principles to enterprise data management.
Efficiency. In concrete terms, this is the most obvious subject to tackle, but certainly the most "political". Data is everywhere in a company, often copied from one system to another, with problems of redundancy and quality. For example, how many systems in a company contain customer data?
To be effective, you need to make "political", structuring choices, prioritize needs, empower employees - in a word, "manage" data. Easy to say, not easy to implement.
Sobriety. It seems clear that, for many companies, the extraordinary Cloud offering from hyperscalers no longer requires a great deal of thought. Creating new digital environments, increasing storage, multiplying computing power - all this is dematerialized and can be done in a few clicks.
It is now simpler to buy new digital resources than to recycle old ones... to copy data to a centralized platform rather than optimize operational databases.
The surplus of resources thus created is then used for new consumption or even overconsumption (the famous rebound effect). All of this without questioning the use of data for business operations or environmental impact analysis.
To avoid "Poverty" (forced sobriety), it's certainly time to complement the "Lean" approach with a "by design" approach to digital architectures (and data in particular) to integrate ecological impact as a decision-making criterion in all organizational or structuring choices.
Digital technology is now an essential building block in the functioning of our economies and societies.
We can expect digital technology to have a positive impact on other industries (is digital the main factor in scope 4, i.e. emissions avoided in a carbon balance sheet, for other industries?), and for this to happen, digital sobriety is an important issue that requires a real change of Mindset.
It's time to move away from the logic of "always more", and start treating issues in terms of "right scale" and "data discernment". This will reduce the environmental and energy impact of digital technology, give organizations greater control over their data, and foster a more responsible and sustainable approach.
Samuel Descroix
Partner Data thinking
