Amazon Product Selection and Operations
Black Friday Lightning Deal Price Keeps Changing? Work Out Your Listing's Reference Price
The Black Friday and Cyber Monday (BFCM) event starts in a little over half a month. If you sell on Amazon, take a moment to check your backend: has the Lightning Deal price you submitted been flagged as invalid, and is the price the system recalculated much lower than what you submitted? Many people have run into this lately and still can't figure out why. The reason is not a system problem — it's that your Amazon listing's reference price is changing.
To push unit sales or ranking before the big event, many sellers rotate through sale prices, coupons and member discounts, so the actual transaction price keeps sliding down — and their BFCM reference price gets dragged down by their own everyday pricing.
The reference price for a big-event Lightning Deal is officially called the Typical Price — what we call the regular price. The calculation isn't complicated. Amazon pulls the listing's sales records from the past 90 days, updates them on a rolling basis every day, and takes the median of what buyers actually paid over those 90 days — note, what they actually paid, not the Your Price you set yourself. So this number reflects how much most people actually spent to buy it.
Which prices get counted takes some care. Your normal everyday selling price definitely counts. An occasional few days of discount or a single Lightning Deal generally does not count, and the reference price still follows your everyday price.
But after May 18 this year, Amazon has a new reference price calculation rule: if you sold at a discount on more than half of the days in those 90 days, Amazon will consider the discounted price to be your real price, count all those discounted prices in, and the reference price drops right away.
However, prices during big events like Prime Day and BFCM do not count — no matter how low you sell on those days, it doesn't affect your reference price.
So how is the BFCM Lightning Deal price calculated? First, the selling price must be 15% below the reference price — for example, a reference price of 39.99 USD times 85% is 33.99 USD. Second, it must be another 5% below the lowest price you sold at recently — if your recent low was 32.99 USD, another 5% off gives 31.34 USD. You take the lower of the two, and that is the highest price you can submit for a BFCM Lightning Deal.
This rule leads to the situation we mentioned at the start: you keep cutting your price for unit sales, the discounted days pass the halfway mark, your reference price gets dragged down, and the Lightning Deal price the system recalculates is naturally much lower than what you originally submitted.
So these days we need to check the backend from time to time to see whether the price we submitted earlier is still valid. Then we need to revise the Lightning Deal price, otherwise there is no Lightning Deal badge on the event day.
Going forward, we also need to change our habit of changing prices casually. Be especially careful with price adjustments, and don't cut your selling price at will for short-term ranking, otherwise it easily leads to real pain during the big event: submit a Lightning Deal and there is no profit; skip the Lightning Deal and you miss the golden window of the event.
Amazon is now placing tighter and tighter constraints on listing prices, and the old spiral playbook is harder and harder to make work. For sellers, what needs more thought going forward is how to sell products through product competitiveness and operational capability, without needing low prices to get products selling.
BSRBOT Amazon Seller Advisor (bsrbot.com) can generate multiple reports for a category in one click — buyer demand analysis, category & competitor analysis, and keyword analysis — to help you judge whether any category or product is worth pursuing.
Generate product selection and operations research reports in one clickData note: category structure, demand signals and competition levels are aggregated from public Amazon category pages and our own research snapshots. Figures are ranges, not individual seller data.