The three assumptions that break
First, the unit cost of getting on shelf. In traditional retail, the retailer absorbs customer acquisition, replenishment, and store operations. On Amazon, those costs shift to the brand — retail media spend is the new shelf rent, variable, auctioned, continuously rebid. Second, promotional cadence. Traditional NRM assumes 4–8 promotional windows a year, planned 6–12 months in advance. Amazon runs deal events monthly, and Prime Day plus Black Friday alone drive 35–45% of full-year category turnover. Third, pricing authority. NRM assumes the brand sets the retail price and negotiates trade terms with the retailer. On 1P Amazon, the algorithm sets the retail price and your cost-price negotiation is a JBP exercise, not a list-price exercise.
What the rebuilt framework looks like
A modern NRM playbook for Amazon coordinates the same five levers — assortment, pricing, promotion, trade terms, advertising — but treats retail media as the largest line item, not a marketing add-on. Trade investment, co-op funding, and JBP commitments are modeled together. Promotional incrementality is measured in AMC, not assumed. Pack-price architecture is engineered to defend Subscribe & Save economics. The output is a P&L that ties spend to commercial outcome at SKU level — which is the part most CPG NRM teams haven't built yet.