As of October 7, 2026, Shifter has the largest pool of live US IPs, more than Oxylabs and Bright Data, at a fraction of the cost.Shifter now has the largest pool of live US IPs

See benchmarks

Knowledge

Winning the Amazon Buy Box: Using Competitor Price Data

How the Amazon Buy Box (the Featured Offer) is decided, which competitor data actually helps you win it, and how to collect that data by marketplace and delivery location.

Priya Nair

October 17, 2026 · 8 min read

When several sellers offer the same product on Amazon, only one offer sits in the box with the “Add to Cart” and “Buy Now” buttons. Sellers call it the Buy Box; Seller Central calls it the Featured Offer. Shoppers who click those buttons buy from whoever holds it, usually without opening the list of other sellers, so on shared listings the Buy Box decides who gets the sale.

Price is not the only factor, but it is the one that moves fastest and the one you control hour by hour. Winning it consistently means knowing what competing offers cost, as shoppers see them, in each marketplace you sell in. That is a price intelligence problem, and this guide covers how to approach it.

Amazon does not publish its formula, but its seller guidance is consistent about the inputs. To be eligible at all, a seller generally needs a Professional selling plan, healthy account metrics and the item in stock. Among eligible offers, the factors Amazon points to are:

  • Total price. The item price plus delivery cost, often called the landed price. A lower item price with a high shipping charge can lose to a slightly higher price with free delivery.
  • Delivery speed and reliability. Faster, dependable delivery helps. Offers fulfilled by Amazon tend to score well here because the delivery promise is Amazon’s own.
  • Seller performance. Order defect rate, late shipments, cancellations and valid tracking all count.
  • Availability. An offer that keeps running out of stock is a weaker candidate.

Two consequences follow. First, matching the lowest price does not guarantee the box if your delivery or metrics are weaker; you may be able to win it at a price above the cheapest offer. Second, Amazon can show no Featured Offer at all when every offer looks overpriced, for example against recent prices on or off Amazon. In that case shoppers see a “See all buying options” button, and conversion on the listing drops for everyone.

The competitor data that matters

For each product (ASIN) you care about, the useful data is narrower than a full scrape of the page:

Data pointWhy it matters
Current Featured Offer holder and its landed priceThe price you are competing against right now
Every competing offer: item price, delivery cost, fulfilment type, conditionWho could take the box next, and on what terms
Delivery promise shown for each offerDelivery speed is part of the decision
Whether a Featured Offer is shown at allSuppression means the problem is price level, not the competitor
Timestamp and marketplace of each checkPrices and holders change during the day and differ by storefront

Collected over time, this shows patterns a single snapshot hides: which competitors reprice automatically and how fast, what time of day the box changes hands, and the price floor below which a competitor stops following. Our online price volatility study shows how quickly prices can move on large marketplaces, which is why the timestamp matters as much as the price.

Two ways to get it: the API and the storefront

For your own listings, start with Amazon’s Selling Partner API. The Product Pricing API returns the competing offers and the Featured Offer price for items you sell, and the notifications service can push an event when any offer on your listings changes. It is official, structured and free to use within its rate limits. Amazon also offers its own automated pricing tool in Seller Central.

The API does not show everything a shopper sees. It covers the products you sell, in the marketplaces where you are registered, from Amazon’s side. It does not show:

  • products you are considering but do not yet sell;
  • how a listing looks in marketplaces where you have no account;
  • what the page shows to shoppers in a particular delivery location, which affects delivery promises and sometimes which offer wins.

For that, teams check the public product pages and offer listings the way a shopper would. Our guide to scraping Amazon product data with proxies covers the mechanics of collection; the points specific to Buy Box monitoring are below.

Collecting by marketplace and delivery location

Each marketplace is separate. amazon.com, amazon.co.uk, amazon.de and the rest have their own sellers, prices and Featured Offer holders. A competitor can hold the box in Germany while you hold it in the UK.

The delivery location changes the page. Amazon decides delivery promises, and therefore part of the Buy Box decision, from the shopper’s delivery address. Without a set address it falls back on location signals, including the visitor’s IP. For a consistent view, set the delivery location in the session and send the request from an IP in the same country, and where it matters the same city. A residential proxy gives you that: a household connection in the market you are checking, so the page you collect is the page local shoppers get.

Keep sessions consistent. Delivery location, currency, language and IP should agree for the whole check. A UK delivery address with a US IP and a German interface is not what any shopper sees. Use a sticky session for each check of an offer listing, then rotate for the next product.

Tier your checking frequency. Hourly checks on your best sellers and on listings where the box changes hands often; daily for the long tail. Checking a quiet listing every few minutes adds cost and load without adding information.

Turning price data into Buy Box wins

Collection only pays off if it changes decisions. A simple rule set:

  1. Compare landed prices, not item prices. Add delivery cost to every offer before comparing.
  2. Price against the eligible competitor, not the cheapest one. An offer with slow delivery or poor metrics may not be the one you need to beat.
  3. Set a floor from your margin, not from the competitor. Following a competitor that prices below cost is a race you do not need to win.
  4. Watch for suppression. If no Featured Offer is shown, cutting a little below a competitor will not help; the whole price level needs to come down or the listing needs attention.
  5. Measure share of time in the box. Track the share of checks where you hold the Featured Offer, per marketplace. It is the clearest single measure of whether your pricing works.

A short helper makes the comparison concrete. Each offer is a dictionary from your collected data:

def landed(offer):
    return round(offer["price"] + offer.get("shipping", 0.0), 2)

def price_to_beat(offers, my_seller_id, min_rating=0.9):
    """Lowest landed price among competing offers that are likely
    Featured Offer candidates: in stock and with a decent rating."""
    rivals = [
        o for o in offers
        if o["seller_id"] != my_seller_id
        and o.get("in_stock", True)
        and o.get("rating", 1.0) >= min_rating
    ]
    if not rivals:
        return None
    return min(landed(o) for o in rivals)

offers = [
    {"seller_id": "ME", "price": 24.99, "shipping": 0.0, "rating": 0.97},
    {"seller_id": "A1", "price": 22.49, "shipping": 3.99, "rating": 0.95},
    {"seller_id": "B2", "price": 23.99, "shipping": 0.0, "rating": 0.82},
]
print(price_to_beat(offers, "ME"))  # 26.48: A1's landed price, B2 filtered out

The rating threshold is a stand-in for whatever eligibility signals you can observe; tune it from your own history of which offers actually win.

When the problem is not price

Sometimes you lose the box to an offer you should never be competing with: a seller using your listing for counterfeit or grey-market stock, or a reseller breaking your pricing policy. Price changes will not fix that. Our guides to detecting third-party sellers hijacking your listings and monitoring unauthorised sellers and MAP violations cover what to collect and how to act on it.

FAQ

What is the Amazon Buy Box?

The box on a product page with the “Add to Cart” and “Buy Now” buttons. When several sellers offer the same product, Amazon picks one offer to feature there. In Seller Central it is called the Featured Offer.

Does the lowest price always win the Buy Box?

No. Amazon also weighs delivery cost and speed, seller performance and stock availability. A slightly higher landed price with fast, reliable delivery can win over the cheapest offer.

Why does a listing sometimes have no Buy Box?

Amazon can suppress the Featured Offer, for example when all offers look overpriced compared with recent prices. Shoppers then see “See all buying options” instead of “Add to Cart”.

Do I need proxies if I use the Selling Partner API?

For your own listings in your own marketplaces, the API covers most of what you need. Proxies matter when you need the shopper’s view: other marketplaces, products you do not sell yet, or what the page shows in a specific delivery location.

The bottom line

The Buy Box goes to the best overall offer, and price is the lever you can pull fastest. Use the Selling Partner API for your own listings, collect the shopper’s view by marketplace and location for everything the API does not show, compare landed prices against eligible competitors, and track your share of time in the box. For collection across many storefronts, see our ecommerce use case.

Ready to get started?

Try Shifter's residential proxies, 205M+ IPs, 195+ countries, from $0.10/GB.

Get Started