IT Sustainability Think Tank: The importance of measurement in closing the sustainability gap
There is mounting anecdotal evidence that enterprises are struggling to ensure their statements of intent on sustainability are matching their actions, so what steps can they take to bridge the gap?
It is now the expected norm for companies to have a sustainability strategy, with a YouGov survey finding that 70% of GenZ customers will pay more for sustainable products. However, there is sometimes suspicion that the stated goals do not match behaviour leading to accusations of greenwashing being levelled at organisations.
The tech sector generates between 2–4% of all greenhouse gas emissions. And the International Telecommunication Union estimates that digital technology alone is capable of reducing the entire world’s carbon emissions by 17%.
Additionally, work by Ericsson suggests the tech sector could further reduce its carbon footprint by 80% if all its electricity came from renewable sources.
Measurement is crucial to quantify the gap between where an organisation is and where it wants to be. That is, set a goal and then develop an incremental, continuous improvement strategy. Restating and clarifying an organisation’s strategy in measurable terms establishes progress.
What sustainability means
Fixing a consistent definition of what sustainability means need not be complex: for example, the common metric is the measure of an organisation’s CO2 (or CO2-equivalent) emissions. Especially hard to assess, however, is the entire value-chain emissions impact. Fortunately, there are now far more tools emerging in the sustainability domain; tools like Green House Gas Protocol’s Scope 3 Evaluator that target potential areas of emission reduction. This may mean optimising current hardware, migrating to the cloud, redesigning services, or all three.
There are a few potential quick wins with hardware. Organisations that still use servers and/or datacentres can measure both their electricity usage and the sources of that electricity.
Using publicly available data, this can provide a sufficiently accurate proxy for emissions. Once this is clear, there are several options available to reduce those emissions. For example, new servers can be procured and provisioned with greener components.
Alternatively, organisations should consider if they actually need new servers. That would avoid putting older servers in landfill, and instead reusing and repurposing them. Another option is to use hardware more intensively by putting more load on it than initially planned. For instance, a server running at 60% capacity has less impact than two servers running at 30%.
Public cloud providers have multiple incentives to use renewables and be sustainable, and they are almost always likely to be more energy efficient than a typical datacentre. Therefore, transitioning to the cloud will produce quantifiable improvements in performance on sustainability metrics. But moving to the cloud should be viewed as just the start. Managed cloud-native services can be many times greener than the equivalent application components running on IaaS server instances.
A further benefit is that cloud providers all have tools to produce quantitative emissions data that can be used to monitor progress, such as the Google Cloud Platform’s Carbon Footprint and Amazon Web Services’ (AWS) Customer Carbon Footprint tool; there are some excellent open-source tools out there too.
Finally, improving the sustainability of the engineering practises may be more complex and more contested. For example, the context the organisation is operating in will have a big influence on whether the choice of say Java is more sustainable than Python.
Good engineering practice, such as edge-caching, optimised data storage, reusability and code efficiency can almost always have a positive impact on sustainability. Applications that require less compute power use less electricity, which ultimately leads to a net reduction of CO2-like emissions.
It is important to take these factors into account when choosing architectural options and following green engineering best practices. The gains may be small at the level of the developer but become clearly significant when scaled-up to production levels. Quantitative measurement is essential to evidence that improvement.
Sustainability has also become part of DevOps vocabulary as GreenOps, focusing on improving continuous integration and delivery from the perspective of reducing emissions. A critical part of this role is adding sustainability reports to existing dashboarding approaches, giving organisations a real-time window on that closing sustainability gap.
The key is managing customer and organisational objectives throughout, and treating sustainability like a transformation programme. Organisations will need to think big, start small, and scale fast. They will need to take focused steps to set meaningful and measurable goals in an agile and iterative way.