Demand planning in aquaculture – a practical collaboration with INVE Aquaculture
Case summary
INVE Aquaculture operates in an international and complex planning environment where demand is influenced by feeding cycles, customer behavior, seasonality, and frequent order timing shifts. In this context, demand planning is not about predicting individual orders. It is about establishing a reliable reference that remains stable despite volatility.
At the start of this collaboration, demand planning at INVE was largely manual and sales driven. Globally spread teams created forecasts each month based on local insight. The process required significant effort and still resulted in structurally biased forecasts, limiting their usefulness for supply planning and S&OP.
The initial step was deliberately limited in scope: introduce a system-generated demand baseline and benchmark it against existing sales forecasts using INVE’s own rolling accuracy metrics. There was no integration, no change to roles, and no sales input - only validation.
The results were clear. The system baseline consistently reduced bias and improved forecast accuracy across most product categories. Improvements typically ranged from mid-single-digit to low double-digit percentage points, while bias dropped to low single-digit levels. These results were achieved with a fraction of the manual effort previously required.
An important insight emerged around order timing. Orders frequently move between months, which can distort accuracy measurements after the fact. By explicitly separating the system forecast from post-order timing adjustments, the team could distinguish forecast quality from execution effects - avoiding misleading conclusions.
Only after this validation did sales teams become involved, initially in a limited number of regions. Their role shifted from producing forecasts to reviewing and refining an already stronger baseline. In many cases, sales input aligned closely with the system proposal. Where customer-specific insight added value, it led to further improvement. Where it did not, the baseline stood on its own.
To maintain discipline, sales and planner adjustments were tracked separately rather than overwriting the baseline. This made interventions visible and reviewable, and helped identify where manual input consistently improved outcomes.
This was not a tool rollout. Success depended on the people involved: clear direction from supply chain leadership, strong ownership from planning and S&OP, and close collaboration with IT. Automation and SAP ECC integration only began once the logic, roles, and value were proven.
Sven De Vlieger, Supply Chain Director, set the direction. Piyachat Ditpakdithorn, Business Systems Manager, reviewed forecast behavior and pushed for clarity on which numbers should be used. Georgia Toumleli, S&OP Manager, linked forecast behavior to how the business reviews volumes and commits decisions. Kanokporn Yotjana, Demand Planning Manager, provided continuous feedback from an operational angle. Frederik Nagels, Senior IT/IS Manager, prepared the integration with SAP ECC and set up SSO.
Today, sales input is used to enhance a strong demand baseline rather than to compensate for a weak one. Manual effort has been reduced, bias has been addressed and forecast accuracy has improved in a way that supports more reliable supply and S&OP decisions.
That sequence - validate first, involve people second, automate last - is what made the collaboration effective.
