Why Did My Amazon Sales Drop? A Data-Driven Diagnosis via API

To diagnose why Amazon sales dropped, pull the product's history from /amazon-product-statistics and check four causes in order: did another seller take the Buy Box, did the landed price rise, did the sales rank worsen with price unchanged, or did the rating fall. One call returns all four series for the full tracked period, so the check runs in a single request instead of four separate guesses.

What are the most common causes of a sales drop?

Most drops trace back to one of four visible changes: the Buy Box moved to a competitor, the price rose (yours or a repricer's), the sales rank worsened even though price held (fewer actual purchases), or the rating fell after a run of negative reviews. Checking in that order catches the loudest cause first, since losing the Buy Box alone can zero out sales for a listing that previously converted well.

How do I check Buy Box and price history programmatically?

/amazon-product-statistics returns buyBoxSellerIdHistory, buyBoxPriceHistory, salesRankHistory and productRatingHistory as parallel timestamped series inside data.stats. None of them are cleaned: a rank of -1 means no reading was recorded at that time, not a rank of negative one, and price series simply have gaps rather than zeros when no offer existed. Compare the two most recent points of each series rather than averages, since a repricer change is a step, not a trend.

diagnose.py - check the four causes in order

"""Diagnose a sales drop from the product's own history, not guesswork.

Pulls /amazon-product-statistics (price, Buy Box holder and sales-rank history)
and checks the four most common causes in order: lost the Buy Box, a price rise,
a rating/review drop, or a rank spike (less visible) with no price change at all.
"""
import os

import requests

BASE = "https://sellermagnet-api.com/api"
SESSION = requests.Session()
SESSION.headers["X-Api-Key"] = os.environ["SELLERMAGNET_API_KEY"]


def diagnose(asin: str, marketplace_id: str, my_seller_id: str) -> list:
    resp = SESSION.get(f"{BASE}/amazon-product-statistics",
                        params={"asin": asin, "marketplaceId": marketplace_id}, timeout=120)
    body = resp.json()
    if resp.status_code != 200 or not body.get("success"):
        raise RuntimeError(f"{asin}: {resp.status_code} {body.get('message')}")
    stats = body["data"]["stats"]

    findings = []
    bb_history = stats.get("buyBoxSellerIdHistory") or []
    if len(bb_history) >= 2:
        before, now = bb_history[-2][1], bb_history[-1][1]
        if before == my_seller_id and now != my_seller_id:
            findings.append(f"lost the Buy Box to seller {now!r}")
        elif now == "-1":
            findings.append("nobody currently qualifies for the Buy Box")

    price_history = stats.get("buyBoxPriceHistory") or []
    if len(price_history) >= 2:
        before_price, now_price = price_history[-2][1], price_history[-1][1]
        if now_price > before_price * 1.05:
            findings.append(f"landed price rose {before_price/100:.2f} -> {now_price/100:.2f}")

    rank_history = stats.get("salesRankHistory") or []
    clean_ranks = [r for _, r in rank_history if r != -1]
    if len(clean_ranks) >= 2 and clean_ranks[-1] > clean_ranks[-2] * 1.5:
        findings.append(f"sales rank worsened {clean_ranks[-2]} -> {clean_ranks[-1]} (fewer sales, same price)")

    rating_history = stats.get("productRatingHistory") or []
    if len(rating_history) >= 2 and rating_history[-1][1] < rating_history[-2][1] - 0.2:
        findings.append(f"rating dropped {rating_history[-2][1]} -> {rating_history[-1][1]}")

    return findings or ["no cause visible in price/Buy Box/rank/rating history; check PPC spend and inventory next"]


if __name__ == "__main__":
    for finding in diagnose("B0CLTBHXWQ", "APJ6JRA9NG5V4", my_seller_id="A11IL2PNWYJU7H"):
        print("-", finding)

Output for a listing that lost the Buy Box and more

- lost the Buy Box to seller 'A999OTHER'
- landed price rose 418.00 -> 449.97
- sales rank worsened 1200 -> 5400 (fewer sales, same price)
- rating dropped 4.6 -> 4.3

The stats fields this check reads (trimmed)

{
  "success": true,
  "data": {
    "stats": {
      "buyBoxSellerIdHistory": [["2026-09-01 00:00:00", "A11IL2PNWYJU7H"], ["2026-10-13 00:00:00", "A999OTHER"]],
      "buyBoxPriceHistory": [["2026-09-01", 41800], ["2026-10-13", 44997]],
      "salesRankHistory": [["2026-09-01 00:00:00", 1200], ["2026-10-13 00:00:00", 5400]],
      "productRatingHistory": [["2026-09-01 00:00:00", 4.6], ["2026-10-13 00:00:00", 4.3]]
    }
  }
}

What if the history shows nothing unusual?

A flat Buy Box, price, rank and rating history rules out the four causes this endpoint can see, which means the drop sits outside product data entirely: advertising spend paused or exhausted, a stock-out the history has not caught yet, a suppressed listing from an account-health issue, or a seasonal dip shared across the whole category. Those need the advertising console, inventory dashboard and account health page directly, not an API call.

Bottom line: before assuming the worst, pull one /amazon-product-statistics call and check Buy Box, price, rank and rating in that order. For the full field list and history shape, see the product data without SP-API guide, and budget the check with the pricing guide.