A split award divides one tender across two or more suppliers — by lot, volume share, or line item — instead of awarding the entire scope to a single winner. It trades some administrative simplicity for lower total cost, supply security, or both.
When splitting beats a single award
Three situations recur. No single bidder is cheapest on every line, so the lowest total cost is a combination. Volume exceeds what one supplier can reliably deliver, so concentration itself is the risk. Or the category is strategic enough that keeping two qualified suppliers active is worth a small premium on part of the volume.
Why split awards are hard to do by hand
With a handful of lines the best combination is visible. With hundreds of lines, several bidders, and constraints — minimum volumes, delivery capacity, incumbent transitions — the number of viable allocations explodes, and a spreadsheet comparison quietly settles for a plausible answer rather than the best one. In dmp’s bid evaluation, the lowest-total-cost allocation is calculated across normalised bids, and the evaluation committee reviews and can override it — with the override logged like everything else.
The takeaway: a split award is an optimisation problem wearing a procurement name. Treat it as arithmetic on clean data, not as horse-trading across a table.