Using Data, Analytics & AI to step-change your Sustainability

Platforms as a foundational building block

Firstly, you’ll need a platform (or several platforms). There are a wide range of off-the-shelf platforms available, covering generic use cases and more specific (e.g. industry specific) use cases. Where possible, buy, don’t build. But it’s a nascent market out there so it’s quite possible that you’ll need to do a bit of both.

You can pull data into the platforms via APIs from your other systems e.g. ERP, Accounts Payable. Suppliers can, in turn, upload data to the same platforms. Ideally, via APIs from their systems, but the prevalent (while less efficient) means currently is via file upload. Some platforms also go out and, using publicly-available information, automatically generate a supplier Sustainability profile – this can be a really useful starting point to then collaborate with your suppliers to build upon.

The upshot is, your suppliers’ data is going to get sucked into this platform, meaning you can then access it, as a centralised repository – a single source of truth. It will power all your usages of that data: from strategy, to operations, to reporting. Regarding the latter, you should look to use that data to efficiently generate multiple reports for each of the ESG reporting methodologies you subscribe to.

Data hustling

Look, this process to gather data can take A LOT of work. But with an informed, well-coordinated approach, the pay-off in downstream efficiencies and effectiveness will be worth it.

Below are some important considerations:

Actual (or Primary) data – rather than estimated data. Actual data can be traced back to an original source, which could be a sensor or an invoice. Reporting methodologies including the EU’s CSRD legislation want you to use as much primary data as possible. Don’t use secondary data unless there’s no other way, as it can be widely inaccurate and you might then have to re-baseline later, which is a bad thing.

Accurate (error-free) data – relying on having a robust system of data preparation and hygeine, ensuring that context is not lost as data is transferred from one system to another.

Audit-ready data – your data needs to be prepared so that it is auditable, meaning considerations like monitoring and validation of chain of custody and data lineage are accounted for.

Business-ready data – the data needs to be utilised by the business (strategy, operations) as well as for reporting (avoiding highly awkward disconnects) meaning that you need a system that combines business data with Sustainability data.

Complete data – avoid inadvertent omissions and resulting risk of greenwashing claims – you need the time and effort to build a rigorous methodology.

Fresh data – use automated systems where possible to support a higher frequency of updating and reporting – important given you’ll be using the data more than once a year, whether for monthly stakeholder updates or for generating various ESG reports throughout the year.

Granular data – you need sufficiently detailed data that you can have confidence in your baselines, your targets, and the overall business case. Consider using IoT for asset-level usage data.

Done? Great job! With all that lovely data now sitting there, what better next step than to…

Deploy analytics and AI

Massive data sets like the on you’ll be sitting on are almost audibly crying out to have analytics and a smorgasbord of AI techniques deployed on them. Doing so will allow you to gain so much value from all of that data. Here are few examples and ideas:

Use machine learning and a variety of data science techniques to, for example:

• Recommend actions that suppliers can take to reduce their impact (based on what is working for other suppliers).
• Reconfigure your supply chain so that sourcing, production, distribution have a reduced impact (e.g. reduce the emissions from transportation by locating production closer to retail).
• Calibrate production runs (for example of freshly baked goods to your retail outlets) to usage, thereby reducing waste at the end of each day.
• Identify opportunities to optimise your clothing manufacturing process, to reduce water usage.
• Detect pollution via data that comes from an array of sensors including monitors, IoT and satellite information, combined with machine vision.
• Speaking of machine vision, use it to analyse the types of waste leaving your production facilities, and then find ways to reduce that waste.

Use LLM Co-pilots to provide a friendly, at-scale interface for your suppliers to learn and improve from. Imagine being able to access a knowledge base of what you want them to give you data-wise. And to collaborate, at scale, on the process. But what about also if you could give those same suppliers a load of help to become more sustainable – actually building your supplier capability to become more sustainable, in turn improving your own metrics? Talk about win-win.

Use deep learning to surface hitherto unknown causal patterns from the data, so in turn you can make changes based on those patterns. For example, you might find out that specific weather patterns or types of event result in a particular change in consumption, which in turn allows you to plan to meet demand with lower resource usages.

Use predictive analytics to combine various data sources and recommend courses of action to reduce impact…or perhaps even have a regenerative effect, using excess resources in a net positive way.

Use data visualisations with your stakeholders (employees, impacted, communities, partners, investors) to increase transparency, raise awareness and build trust.

Set up automations to:

• Monitor and adapt production and/or distribution reduce impact (e.g. material/energy/water usage, or staff overtime requests) within thresholds
• Monitor and adapt the physical movement of materials, reduce waste, reduce employee travel and therefore emissions
• Pull in external data, like economic/country/industry risk factors – for example wildfires, adverse weather, or social factors that could have an impact on your business. Remember – there’s the outside-in (‘financial materiality’) aspect of Sustainability, as well as the inside-out (‘impact materiality’).

Depending on where you are on your Sustainability journey, the above set of examples and ideas might seem a little intimidating. The point is, you need to get started somewhere, and you can improve over time, according to a roadmap. Also, tech is improving quickly, as is people’s knowledge of how to best deploy it. Surf’s up – paddle out on your board and catch that wave early!

Also, the good news is there are lots of solutions and companies out there (like Riverflex) that can help you. Not only can we help with choosing the right platforms and tools, but with getting them set up correctly, preparing the data, and doing bespoke analytics and AI work. We can also help with the management of what can sometimes be significant programmes of work, necessitating not only new technology and data management, but also changes to processes, roles, governance, mindset and culture.

A final word: if the opportunity to modernise business and address injustices done to planet and people aren’t enough, there’s more. The shifting of these tectonic plates, and the resultant eruptions, will also present a perhaps once-in-a-lifetime opportunity. New islands – nay, continents – of value will emerge; deeply fertile soils of opportunity will be laid down; vibrant new business ecosystems will flourish. Prepare to innovate, build new markets, gain market share, refresh your brand positioning, discover powerful new means of differentiation, energise your workforce, and unlock new levels of customer engagement and loyalty.

In summary, a Sustainability paradigm shift is underway, so get ready!

Thanks to Anisha Passcuran, Mohamed el Harchaoui and Shobhendra Devries for reviewing and providing valuable expert feedback on this piece.

Get in touch with the Riverflex team to find out more and how we can help you become a Sustainable business.