What role can Machine Learning play in social procurement and economic development as the economy slowly comes back on?

As we think of ways to support local businesses personally these days, does your organization have similar goals? If so, how have you tackled the classic data challenge of knowing which of your partners along the supply chain are ‘local,’ ‘small/medium size enterprise,’ or any other sustainability dimensions your organization values?

While buzzwords such as Artificial Intelligence (AI) and Machine Learning (ML) have been trending over the last few years, their value for answering this exact question has come to the forefront of conversation as we gradually dial the dimmer back up on our economy. 

In light of this, we want to share our team’s experience helping our client learn more about their suppliers and the social impacts of their procurement practices.

The Challenge:

To transform their Corporate Social Responsibility (CSR) program, our client wanted to monitor and understand the impact of their procurement activities on small/medium businesses (SMBs) and Indigenous Businesses. 

Key issues our Client was facing:

1.    Necessary data was not captured or tracked;

2.    Lacked database integration from multiple sources; 

3.    Lacked investment in Business Intelligence (BI) tools; and

4.    Existing processes to develop and validate business insights were highly labour intensive, inefficient, and prone to human errors.

Our Solution: 

We set out to leverage the power of machine learning and predictive analytics algorithms. Our team developed a machine learning model to profile suppliers by only using 5% of total vendor records with 90%+ accuracy. And we were able to deliver this solution in only a matter of weeks, showing that use of machine learning can deliver immediate and valuable results much faster than traditional ways of working!

Our Impact: 

Our analysis found that the client’s current spend patterns were delivering social impacts in their local communities. Further analysis helped them understand where (i.e. what categories) they could further focus their efforts in developing the supply market with emerging players.

Through the power of machine learning and predictive analytics, our client could now make data-driven decisions to enhance and be more targeted with their CSR initiatives. 

Lessons Learned: 

Far too often, we see organizations jumping into the deep end with heavy investments such as significant process changes to expand data collection and data cleansing. Using machine learning techniques to understand the current state and establish an on-going analytics mechanism is not only effective but a safe approach before investing in fit-for-purpose process changes. This also allows entities to be agile in times of ambiguity.

The solution highlighted above was one successful case of machine learning and predictive analytics modelling in a social procurement setting, but the use of this model is not limited to the specific information the client wanted to know. For example, our solution could enhance procurement policies for indigenous businesses, environmentally friendly/green products, as well as industry sectors that are driving economic growth (e.g., local retooling, digital procurement, supply chain automation). Overall, the implications of the role of machine learning and predictive analytics in social procurement are significant for the public sector and beyond.

As the procurement landscape changes in the post pandemic world and we start looking at new suppliers to support economic activity, managing risk becomes key. Our Machine Learning model helps provide the foundation for supporting the activity. 

Stay tuned as Alicia Stanfill shares our thoughts on supplier risk management in our next month’s edition. Stay safe!