Our client is a leader in the production of fine chemicals, counting most of the most recognizable consumer brands in the world as its customers. They are on a road to digitalization and collaborating a lot with start-ups and other innovators to bring the organization forward.
Their goal in supply planning was to bring in a best-in-class planning algorithm to achieve the following:
- Reduce inventories, which has been a long-term objective but always challenging as the company was already well optimized in this respect, especially towards the end of the year
- Reduce the number of deliveries
- Foster planning excellence
- Don’t change the existing workflow, make it seamless for the planners
- Longer term: automate the update of purchasing parameters such as minimum order quantities
- Integrate into the existing IT infrastructure compliant with security guidelines
- Conduct first a proof of value with manually extracted data to validate the business case
- Define the KPIs for a success of a pilot. They were measured on 200 SKUs over a period of 6 months. They included:
- Order cost ratio: Track the cumulative number of PO (purchase orders) placed over the whole period
- Month-end inventory trend
- Monthly average inventory ratio: Inventory divided by total demand
- Monthly snapshot of cost savings in the GenLots dashboard
- % of recommendations followed by the planners
- Net promotor score: A qualitative, anonymous questionnaires on the planners working with GenLots
- Identify which developments are still needed according to the customer, such as:
- Single Sign On (SSO) with the existing identity provider such that planners were logged automatically into the GenLots application
- Automatic data exchange between the client ERP (SAP) and GenLots
- Position GenLots in the existing IT landscape and some projects in the works, such as the acquisition of SAP IBP (Integrated Business Planning)
- Conduct a penetration test, whereby we mandate an independent organisation to try to intrude into the GenLots application and try to extract data
- Create a way to weight different objectives to set the focus more on inventory or delivery reduction and create scenarios for different parametrizations
- Integrate an inventory optimization specific to the reporting period at the end of the year into the algorithm
- Provide a customized reporting to measure the KPIs relevant to our client
- Integrate feedback by the planners into our interface
- Several planners working on the portfolio were then planning according to GenLots recommendations. We had weekly meetings with the head of the planning teams to get feedback and integrate it directly in the application. The monitoring was going over a period of 6 months. We did a validation with a site in Europe, as well as with another site from another product division in the US to make sure to cover all use cases.
The results from the Proof of Value were confirmed, meaning an inventory reduction of around 10% while at the same time reducing the number of POs placed by 25%.
During all this time, 97% of recommendations generated by GenLots could be followed by the planners. In the 3% of the cases where they decided differently, it was mostly because of context that was impossible to integrate into the algorithm.
The survey results among planners were overwhelmingly positive, especially with regards to user friendliness and visualization. But we also identified some areas where we could improve, which we did with new features released every two weeks.
Collaboration with the digital department was particularly productive and we implemented all the developments we set out to do. Reading data from the ERP took around 5 days to set up. Writing back into the ERP took longer and was new for both sides.
There were also some unforeseen benefits. The customer mentioned repeatedly that the project helped them to understand their own cost structure and current processes better and shone the spotlight on some improvement projects for the future. The collaboration and team spirit were excellent, people from different departments got into contact for the first time on this project and lasting relationships were created internally. The GenLots team was able to learn a lot from the customer along the way, and our new feature, the simulator, was a direct result from the many simulations we did for this client, which is now accessible to all customers.