The shipping decisions you are making twice a day
Most manual shipping work is one decision, repeated. Writing it down once is the entire trick.
Watch someone process orders for twenty minutes and you will see the same judgement made over and over. German order, under five kilos, standard service. High-value order, signed for. Local order, collection point. None of it is difficult. All of it is attention, and attention is the thing in shortest supply on a busy afternoon.
Start with what you already do
The mistake is to design an automation scheme. The better approach is to notice the rules you are already following and write those down, in order of how often they fire. Four rules that cover eighty percent of orders is a better day than twenty rules that cover everything and that nobody trusts.
- Destination country decides the service more often than anything else.
- Weight brackets usually split into letterbox, parcel and freight.
- Order value tends to decide whether something travels signed for.
- Product group catches the awkward cases: fragile, hazardous, oversized.
Always have a default
A rule set without a fallback fails in the worst possible way: silently, on the orders nobody anticipated. Every rule set needs a default service and a visible queue of orders that matched nothing. That queue is also your backlog. When the same kind of order shows up in it three times, you have found your next rule.
Let the rule prefer a source, not a carrier
A rule that names one carrier ages badly. The carrier changes its surcharges, or a partner rate undercuts it, and the rule keeps firing anyway. A rule that says cheapest tracked service to Germany under five kilos keeps making the right call as the rates underneath it move.
This is also where having both partner rates and your own contracts in the same comparison earns its keep. The rule does not need to know which one won. It needs to know what you were optimising for.