How can AI help counter a supplier price increase?

When a supplier raises prices, the first move is to see whether the increase is real. AI can help by rebuilding the cost stack from the quote's components — freight, handling, insurance, duties, surcharges — so the buyer can see which line actually moved. In the July 2026 PAIR-20 landed-cost task, all four platforms rebuilt the stack under the structured brief and none treated the headline freight quote as the final answer. That structural breakdown is what gives a buyer something to push back against.

Rebuild the cost stack before responding. Ask AI to decompose the supplier's quote into its landed-cost layers: base price, freight, terminal handling, insurance, customs, duties, and surcharges. In the July 2026 run, the platforms produced a line-by-line map that made each cost assumption visible. A price increase that sits in one layer looks different from one that appears across every line.

Draft the push-back around the lines, not the total. Once the stack is visible, the buyer can question each layer separately. A blanket "your price is too high" invites a blanket refusal. A targeted question — "your handling charge increased but the port's published tariff did not change" — requires a specific answer. AI can help draft that structure by comparing the current quote's layers against the previous period or against a second supplier's breakdown.

Verify every number by hand. In the July 2026 run, none of the four platforms matched the EUR 7,940 case answer for the landed-cost task. Two came in EUR 112 low, one EUR 87 low, one EUR 48 high — and one platform's table contradicted its own prose. The structure was useful; the arithmetic was not reliable. Before sending a push-back that cites specific figures, confirm that each line item traces to a source document the supplier will recognise.