Overview

Amazon will revamp the way UK sellers run Lightning Deals, Best Deals and vouchers starting in 2025. The overhaul introduces open‑ended scheduling, dramatically lower daily fees and a variable, performance‑linked charge. Sellers should pay attention because the new cost structure removes most of the financial risk that previously limited promotional experiments.

Key Points

• Scheduling freedom — Promotions can now start on any weekday and run from a single day up to two weeks, replacing the former fixed‑date calendar.
• Daily fee cut — The charge for each day of a Lightning or Best Deal falls from £50/£25 to a flat £2.50, a reduction of roughly ninety‑five percent.
• Variable sales fee — Instead of a flat fee, Amazon will take 0.75 % of the revenue generated by the deal, with a ceiling of £400 per promotion.
• Voucher pricing shift — The per‑unit cost of £0.45 is replaced by a £2 upfront fee plus 1.50 % of voucher sales, applied both on event and regular days.
• Cap on high‑performing deals — Even when sales are large, the variable fee will never exceed £400, protecting sellers from runaway expenses.

How Deals and Vouchers Work

1. Select promotion length — A seller launching a new kitchen gadget can choose a 10‑day Lightning Deal that begins on a Wednesday, matching a mid‑week traffic spike.
2. Pay the reduced daily rate — For each of the ten days, the seller is billed £2.50, totaling £25, instead of the previous £500‑plus upfront cost.
3. Amazon applies the performance fee — If the deal generates £8,000 in sales, Amazon adds 0.75 % of that amount (£60) to the invoice, but the charge stops at the £400 cap.
4. Create a voucher — To promote a seasonal sweater, the seller opts for a voucher, paying a £2 setup fee plus 1.50 % of the voucher‑driven revenue. If the voucher sells 200 units at £30 each (£6,000 total), the fee equals £2 + (£6,000 × 1.50 %) = £92.
5. Track results in Seller Central — All fees, sales figures and performance metrics appear in the Promotions dashboard, allowing the seller to compare the new model against historic data.