Most Amazon pricing decisions are made in a hurry: a competitor drops, the Buy Box moves, the price follows. That keeps you visible, but it can also leave you winning sales that make nothing once referral fees, FBA fees, storage, returns and ad spend come out.
The 14 prompts below build a pricing strategy from the bottom up: a price floor from your own unit economics, a map of where you sit in the market, price tests that give you an answer, guardrails for Automate Pricing or a repricer, discount checks, and a policy check before a change goes live.
They work in ChatGPT or Claude. Paste your own reports where a prompt says [paste ...] and replace the bracketed numbers. A price change touches margin, Buy Box eligibility and sometimes account health, so review every recommendation and make the change yourself.
Know your price floor before you look at competitors
Your floor is the lowest price at which a unit still pays for itself, ads included. Seller Central gives you the inputs: the Fee Preview report (under Reports, Fulfillment) estimates referral and FBA fees per SKU at your current price, and the Revenue Calculator shows fees at any price you type. Referral fees, FBA fees and the price bands they apply to change, sometimes more than once a year, so take current numbers from Seller Central Help, not an old spreadsheet.
Prompt 1 · Know your price floor before you look at competitorsChatGPT Claude
Act as an Amazon marketplace finance analyst. For each SKU below, build a unit economics table at the current price. Inputs from my Fee Preview report: sku, asin, your-price, estimated-referral-fee-per-unit, expected-fulfillment-fee-per-unit, estimated-fee-total [paste rows]. From me: landed cost per unit [COGS by SKU], inbound shipping per unit [amount], monthly storage per unit [amount], return rate [percent] and the share of returns that cannot be resold [percent], and ad spend per unit sold over the last 30 days [amount by SKU]. Return per SKU: price, total Amazon fees, landed cost, return cost per unit, ad cost per unit, contribution per unit and contribution margin percent. Sort by contribution margin, lowest first, and flag every SKU below [target margin].
Why: The SKUs that hurt are the ones profitable only until their ad cost is counted.
Prompt 2 · Know your price floor before you look at competitorsChatGPT Claude
Act as an Amazon pricing analyst. Using the unit economics from the previous step [paste table], calculate three prices per SKU: the floor (contribution zero after ads), the minimum acceptable price (contribution equals [minimum margin] percent), and the target price (contribution equals [target margin] percent). Remember that the referral fee is a percentage of the sale price, so recalculate fees at each new price instead of holding them fixed. Return a table with the three prices and the current price, and mark SKUs where the current price is already below the minimum acceptable price.
Prompt 3 · Know your price floor before you look at competitorsChatGPT Claude
Act as an Amazon fee specialist. Here are the current fee rules for my category and size tiers, copied from Seller Central Help: [paste referral fee rate and any price thresholds for my category, and the FBA fee bands by price and size tier]. Here are my SKUs with price, category and size tier: [paste]. For each SKU, tell me whether a price change of up to [plus or minus 10] percent would cross a threshold where the referral fee rate or FBA fee changes, and what the contribution per unit would be just below and just above that threshold. Use only the fee rules I pasted; if something is missing, say so instead of assuming.
Why: A few cents across a fee threshold can cost more than the price change brings in. Making the model use only the rules you paste stops it from inventing fee numbers.
Map where your price sits in the market
Prompt 4 · Map where your price sits in the marketChatGPT Claude
Act as an Amazon competitive pricing analyst. Below are the top [15] organic results for my main keyword [keyword], with title, price, shipping cost if any, pack size or quantity, star rating, review count and whether it ships with Prime: [paste]. My product: [title, price, pack size, rating, reviews]. Build a table with landed price (price plus shipping), price per unit of [ounce, count or other unit], rating and reviews. Then group the market into price tiers, show where my product sits in each measure, and point out any tier with strong ratings and few offers.
Prompt 5 · Map where your price sits in the marketChatGPT Claude
Act as an Amazon brand strategist. Using the competitive table [paste output of Prompt 4] and my listing [paste title, bullets and main differences from competitors], recommend one position: premium (above the main tier), parity (inside it) or value (below it). For a premium position, list what the listing must prove in the main image, the first two bullets and A+ content to justify the gap. For value, check the price against my minimum acceptable price of [amount]. Give the recommended price range and the evidence for it.
Prompt 6 · Map where your price sits in the marketChatGPT Claude
Act as an Amazon catalog pricing specialist. Here are all the variations and pack sizes of my product with price and units per pack: [paste]. Check the price ladder: the price per unit must fall as pack size rises, and no multipack may cost more per unit than the single unit, because Amazon's fair pricing policy treats that as a violation. Flag every break in the ladder and propose corrected prices that keep each SKU above its floor [paste floors from Prompt 2].
Situation
Pricing approach
What to watch
Private label, you are the only seller
Price for margin and conversion; test rather than chase the featured offer
Featured offer eligibility can still drop if your price is high against prices elsewhere
Shared listing, reseller or wholesale
Compete for the featured offer inside a firm floor
Repricers racing each other to the floor
Launch
Introductory price with a planned date to step up
A long low price lowers the reference for later discounts
Aged or excess stock
Step-downs timed before storage surcharges hit
Compare the price cut with the cost of holding the stock
Rising costs
Small staged increases
Unit Session Percentage after each step
Test price changes instead of guessing
Prompt 7 · Test price changes instead of guessingChatGPT Claude
Act as an Amazon experimentation analyst. Design a price test for [SKU] at a current price of [price], floor [floor] and target [target]. Propose up to three test prices, the order to run them, and how long each must run (at least two full weeks, avoiding [list known events such as Prime Day or holidays]). List what must stay constant during the test: ad budgets and bids, coupons, images and stock levels. Define the decision metric as contribution per day, not revenue, and write the rule for choosing the winner before the test starts.
Why: Amazon's Manage Your Experiments tests content, not price, so a price test is sequential. Fixing the rules in advance keeps you from reading noise as a result.
Prompt 8 · Test price changes instead of guessingChatGPT Claude
Act as an Amazon analytics specialist. Below is the Detail Page Sales and Traffic by Child Item report from Business Reports for each test window, with Sessions - Total, Units Ordered, Unit Session Percentage, Ordered Product Sales and Featured Offer (Buy Box) Percentage [paste one block per window, labelled with the price], plus ad spend and ad orders per window [paste]. For each window, calculate units per day, conversion rate, featured offer percentage, contribution per day using my unit economics [paste], and organic units (total minus ad orders). Tell me which price won on contribution per day, whether sessions or featured offer percentage moved enough to distort the result, and how confident the comparison is given the sample size.
Prompt 9 · Test price changes instead of guessingChatGPT Claude
Act as a pricing economist. From the test results below [paste table from Prompt 8], estimate how sensitive my demand is to price between [lowest] and [highest]: the percent change in units for each one percent change in price. Using that estimate and my unit economics, find the price that maximises contribution per day in the tested range. Say that the estimate holds only inside the tested range, and how much more data would make it reliable.
Set Automate Pricing or repricer rules with guardrails
Automate Pricing in Seller Central is free. Its rules can follow the featured offer, the lowest price, prices Amazon finds outside its store, or your own sales, and every SKU in a rule needs a minimum price. Separately, you can set minimum and maximum allowed prices on each offer in Manage Pricing, which helps catch pricing errors. Rule names and options get updated, so check the labels in your own account.
Prompt 10 · Set Automate Pricing or repricer rules with guardrailsChatGPT Claude
Act as an Amazon repricing strategist. My SKUs with floor, target and current price: [paste]. For each group (shared listings where I compete for the featured offer, and listings where I am the only seller), recommend an Automate Pricing rule type and settings: what to compare against, how far above or below to price, the minimum price (never below my floor), the maximum price, and what the rule should do when no competing offer exists. Return a table I can enter rule by rule, with a one-line reason for each.
Prompt 11 · Set Automate Pricing or repricer rules with guardrailsChatGPT Claude
Act as a skeptical Amazon pricing auditor. Stress-test these repricing rules [paste rules from Prompt 10] against four scenarios: a competitor repricer that undercuts by one cent every hour, the main competitor going out of stock, a lower price for my product appearing at another major retailer, and my own stock falling below [days of cover] days. For each scenario, describe what my rule does, where the price ends up after a few rounds, and contribution per unit at that price. Recommend changes to any rule that ends at the floor or leaves money on the table.
Tip: Running low on stock is a reason to raise price, not hold it. To time that with reorders, see our inventory planning prompts.
Price discounts, coupons and reference prices carefully
Prompt 12 · Price discounts, coupons and reference prices carefullyChatGPT Claude
Act as an Amazon promotions analyst. I am planning [promotion: sale price, coupon, Prime exclusive discount or deal] on [SKU] from [date] to [date]. Current price [price], discount [amount or percent], and current promotion fees copied from Seller Central Help [paste fees]. Using my unit economics [paste], calculate contribution per unit and the extra units needed per day to match my current contribution per day [amount]. If several promotions could apply at once, show the worst-case combined discount and tell me to confirm in my Promotions settings which ones can stack.
Prompt 13 · Price discounts, coupons and reference prices carefullyChatGPT Claude
Act as an Amazon reference price specialist. Amazon's Was Price is based on the median price customers paid on Amazon over a recent period (currently described as 90 days, excluding some promotional prices), and a List Price is shown only when it is supported by recent sales on or off Amazon. Here is my price history for [SKU] for the last [90] days with daily price and units sold [paste], and the List Price I have set [amount]. Estimate whether a sale at [planned price] is likely to show a strikethrough saving, and whether my List Price looks supported. Flag anything I should verify against the current reference price policy in Seller Central Help.
Stay inside the fair pricing policy
Amazon's fair pricing policy lets it act on pricing that harms customer trust: a price significantly higher than recent prices on or off Amazon, a misleading reference price, a multipack priced higher per unit than the single unit, and excessive shipping fees. Penalties range from losing the featured offer to removed offers and, for repeated cases, losing selling privileges. If you have lost the Buy Box to a competitor rather than to a price flag, start with our Buy Box prompts.
Prompt 14 · Stay inside the fair pricing policyChatGPT Claude
Act as an Amazon account health advisor reviewing a price change before it goes live. Planned changes with SKU, current price, new price and shipping charge if I fulfil orders myself: [paste]. Price history for the last [90] days: [paste]. Any Pricing Health alerts or emails about featured offer eligibility, with the reference or competitive price Amazon quoted: [paste]. Check each change against the fair pricing policy: large increases against recent prices, multipack per-unit prices, shipping charges against normal carrier rates, and List Price claims. For each alert, tell me whether to match the quoted price, hold and accept losing featured offer eligibility, or check whether the outside price is a like-for-like offer, using my floor [paste floors]. Return a table: SKU, risk, reason, recommended action. Do not soften the risks.
Why: An increase that is fine on margin can still look like gouging if it is sudden and steep. Staged increases are easier to defend.
FAQ
What is the best pricing strategy for Amazon sellers?
There is no single best one. Start from a price floor that covers fees, landed cost and ad spend per unit, then choose premium, parity or value positioning based on how your listing compares, and test changes one at a time.
How does Amazon Automate Pricing work?
Automate Pricing is a free Seller Central tool that adjusts your price by rules you choose, such as following the featured offer, the lowest price, prices found outside Amazon, or your own sales. You set a minimum price for every SKU, and the rule never goes below it.
What is Amazon's fair pricing policy?
It lets Amazon act against pricing that harms customer trust, such as a price far above recent prices on or off Amazon, a misleading reference price, a multipack priced higher per unit than the single unit, or excessive shipping fees. Penalties range from losing the featured offer to losing selling privileges.
Why does Amazon say my price is not competitive?
Usually because Amazon found a lower price for the same item at another major retailer, compared against your price plus shipping. Check Pricing Health in Seller Central to see which offers are affected and what reference price triggered it.
How often should I change my price on Amazon?
Resellers on shared listings often reprice automatically within a floor and a ceiling. Private label sellers usually do better holding a price for at least two weeks at a time so the effect on conversion can be measured.
SB
Sophie Brooks
Prompt engineer and AI strategist helping marketplace sellers save time and grow. Specializes in building high-impact prompts for listings, ads, and support workflows.
Free download
27 prompts top sellers actually use.
Listings, ads, reviews, analytics and support: the Amazon Seller Prompt Pack as one PDF.