n many jewellery retail optimisation projects, it’s easy to blame automation when recommendations feel uncomfortable or counterintuitive.
But in most cases, the issue is not the algorithm, the logic, or the replenishment model.
The real problem is the absence of clarity on what success actually looks like.
High-performing systems require clear direction. They need a well-defined hierarchy of KPIs—what matters most, what comes next, and what should take priority when trade-offs are unavoidable.
Should inventory turns be maximized, even if it impacts display depth?
Should ageing be minimized, even if it reduces assortment variety?
Should margins take precedence over availability?
When these priorities are not explicitly defined, the system ends up optimizing against shifting expectations. And when targets keep changing—from turns to display, from ageing to margin, from autonomy to control—no recommendation will ever feel “right.”
Automation is not a guessing engine.
It simply optimizes the target it is given.
The solution is not to keep adjusting the system.
It is to clearly define and own your KPIs upfront—so every decision, human or automated, moves in the same direction.
