Data science

Sustainability from a data mindset (2)

April 30, 2024 - 5 minutes reading time
Article by Frank De Nijs

A mature data mindset provides the starting point for an organization to deal with data in a coherent manner. It drives the intelligent acquisition of sustainability data, among other things, so that it can then be adequately and cost-effectively transformed into actionable insights. The data mindset also drives the realization of barrier-free application of sustainability data and insights into processes, decisions and integral ways of working. Making the data mindset discussable with other players in the ecosystem creates integration of sustainability efforts that will extend beyond the organization. In this way, they can ultimately make the ecosystem, our society, more sustainable.

Part 2

This is part 2 of a two-part series on sustainability from a data mindset.

In the first part of this article, we concluded that digitization in ecosystems for sustainability is going further and further. Beyond digitizing processes and linking applications to exchange transactional data with other applications or organizations. The concept of an organization-wide data mindset was introduced in Part 1.

Data is increasingly becoming a consciously managed asset for any organization. In fact, it always has been, yet we still often see data scattered here and there without any records being kept on it, apart from access rights. This is strange. If data is so important, why don't organizations manage data like they manage finances? Data, like money, is a core asset. After all, data is equivalent to money.

The sustainability-stack

The data mindset naturally leads to thoughtful governance of data sources to be used and the resulting insights. Aspects such as transparency in data quality and origin of data are crucial here. Not only from a privacy perspective, but also from an economic perspective. Instead of collecting data sets and securing algorithms in separate places for each issue, it is important to realize an appropriate governance and organizational structure. Only relevant and reliable data should be used, so that decisions are made at the right levels and sustainability will actually pay off.

Only in this way are tasks, work processes and resources with the greatest sustainability impact and potential quick wins adequately identified. It provides the organization with an integrated data-driven roadmap for short-, medium- and longer-term improvement opportunities.

‘The data mindset leads to thoughtful governance of data sources to be used and the resulting insights'

Sustainability as a Service

Providing data and process insights are needed to integrate sustainability considerations into decisions. Sustainability needs clear direction and guardrails from the top down, but at the same time must be activated from the bottom up throughout the organization. Employees need decision-making power to act where it matters. To embed sustainability, it is essential to encourage people throughout the organization to integrate sustainability into their daily tasks and decisions. Therefore, it is obvious that organizations should start feeding the analyses back into business applications. This is also true for the other parties in the ecosystem with whom they are collaborating. In addition to direct interaction with manufacturers, transporters, customers and other parties in relation to current conditions and sustainability goals, data also represents business value for analyses within other organizations. Data and analyses will increasingly be exchanged. Data as a Service, in other words, for other stakeholders in the ecosystem.

Data exchange in the form of Sustainability as a Service for obtaining sustainability insights, unlike for transactions between ecosystem partners, will prove strategic, whether for a fee or not. In any case with a view to Scope-2 and 3, and not least to promote cost efficiency in our society as a whole. The business applications of totally different players in the ecosystem also benefit from better sustainability data, so that integral improvements in sustainability can be realized there too.

From insight to outlook

Artificial intelligence (AI) has all the attention right now. It is now perceived as special and sometimes even threatening. But in the long run, AI, like the Internet today, is common good. The disruptive nature of the Internet on business and society in the past will be no different with respect to AI. A characteristic of disruptive techniques and practices is that they require new knowledge and skills for those who wish to apply them.

Safely mining numerous data sources, cost-effectively setting up and operating a data warehouse and developing models based on AI require a pretty steep learning curve from established organizations.

So what path should you take when using AI on your business data and decision-making? On the one hand, the ambition of one's own organization is important to achieve highly effective business solutions to complex issues. Such as solutions that go beyond what we can achieve with classical analytics within a given time frame. On the other hand, the pace of developments in the market is a driving force in applying AI techniques within the ecosystem. By contrasting these two fundamentals, roughly four possibilities emerge:

  1. Transaction-driven data exchange without analytical capability (little ambition, few market requirements)
  2. Connecting own applications with AI models of others by using their domain knowledge (little ambition, many market requirements)
  3. Developing own simple AI models on a project basis (high ambition, low market requirements)
  4. Deliver AI powered services based on own domain knowledge and corporate intellectual property (high ambition, high market requirements)

Step 4 will not prove to be a requirement for every enterprise. Nor does the same path need to be followed for every insight question. A hybrid approach is obvious. The question for now, however, is for which type of organization and activities it is crucial to be able to survive over time without the need for applied AI models to belong to one's own intellectual property. The moment the need for use of AI materializes, the race is on. At the root of all this is data, as a valuable asset, which must be the first thing to get right at minimal cost.

Every man for himself?

Should each company then build its own intelligent "sustainability stack"? After all, market forces are a great thing, and intellectual property within analytical models should be able to be developed in confidentiality. Also in terms of waste reduction, energy transitions and sustainability in general. But where it comes to generically available multi-tenant cloud-based solutions, where data and analytical models can be kept confidential, there is no need for everyone to reinvent the wheel for themselves.

There is also the permanently exponentially growing sea of data. How long do organizations want to continue bringing in data that is also stored elsewhere? How many parties in an ecosystem think so cost-effectively, and sustainably, to store their data in a cloud environment? How often will data be replicated in the cloud over time, purely because each organization wants the security of fast and reliable data access for itself?

Data hunger will accelerate with the rise of artificial intelligence and training Machine Learning models. Fortunately, technology is increasingly enabling us to securely and quickly access external data sources, allowing the classic Extract & Transform principle to increasingly take place "on demand" and limiting replication of data sources with the deployment of good "Data as a Service agreements" between parties in the ecosystem.

Confidentiality plays a crucial role here. The need for central control over datasets and analysis results is therefore growing, with issues such as rights structures, pseudonymization and pricing being managed.

'Data hunger is only increasing with the rise of artificial intelligence and training Machine Learning models'

Thoughtful data-driven approach

Sustainability with strengthening a company's competitive position and social responsibility is not workable without a thoughtful data-driven approach. Detailed analytics on various data sources enable organizations to measure our environmental footprint, reduce waste, manage risk, engage stakeholders and drive innovation. Only the development of data and analytics strategies to integrate sustainability into myriad business processes provides the opening to improving performance and reputation of organizations in an increasingly sustainable society.

Contributor(s)

As an enterprise innovator at Centric, Frank de Nijs helps organizations get the most out of their data, by connecting technology, data science and business challenges. He writes about business development, data-oriented ecosystems and the many business opportunities of data science.

Topics
Data science Retail Logistic

Insights on technology