Measure real consumer purchase intent. Amazon search volume reveals what people are actively looking to buy - not just research. Historical demand curves and growth signals, queryable by your AI.
Updated 2026-08-19
Trends MCP returns Amazon product and category search volume as a weekly series on a 0-100 index, with about 5 years of history. The keyword is the phrase a shopper would type into Amazon, not an ASIN and not a Google-style research query. The same MCP connection covers 25+ sources, so Amazon demand can sit next to Google Search and TikTok without a Product Advertising API key. The free plan includes 100 requests per month. Starter is $19/month.
Amazon search is a buying-adjacent action. A person who types "air fryer" into Amazon is closer to a cart than a person who types the same words into Google. The Trends MCP series records how that on-site search interest moved over time. Peaks often line up with Prime events, winter gifting, and viral kitchen gadgets. Troughs show categories that were a one-season fad.
The index is not units sold, Buy Box share, or advertising impression share. Sellers who need order curves still use Seller Central. The trend series answers a prior question: is shopper search for this phrase expanding, seasonal, or dead.
get_trends returns the history. get_growth returns period percentages so an assistant can score a list of SKU ideas without plotting every chart. get_top_trends with type Amazon Best Sellers Top Rated skips the keyword and reads a live bestseller-style feed when the task is discovery rather than tracking a known phrase.
Short shopper language wins. air fryer, standing desk, and protein powder match how people search. Model numbers help when the market is SKU-specific (dyson v15) and hurt when shoppers search the generic category. ASINs, seller names, and Amazon URL paths are the wrong type of keyword for this source. Catalog metadata belongs in Amazon product data workflows, not in the search-volume series.
Category phrases aggregate demand across competing listings. That is useful for "is this aisle growing" and weak for "is my exact listing growing." Two brands in the same category will share much of the category curve. Brand-versus-brand reads need the brand as the keyword, then a second call on the category as a baseline.
Spelling should follow US shopper defaults unless the catalog is clearly localized. Hyphenation (e-bike vs ebike) can split series. When two spellings both appear in Amazon autocomplete, query both and compare growth rather than assuming they merge.
Amazon's Product Advertising API returns listings, prices, images, and review summaries for affiliate and catalog apps. It does not expose a normalized, multi-year search-demand series for an arbitrary phrase. Selling Partner and Ads APIs expose account-scoped search-term reports for advertisers who already spend. Those reports are not a public research index and are not available to an assistant that only holds a Trends MCP key.
Trends MCP does not replace Buy Box logic, inventory feeds, or Sponsored Products reporting. It supplies the missing demand curve: how often the phrase is searched, scaled 0-100, with optional volume, through one MCP session that already includes Google Trends and Google Search data.
Scraping Amazon search-result pages for rank tracking is a different product category. Rank trackers watch listing position. This source watches query demand. A listing can lose rank while the query still grows, which is why the two should not be treated as substitutes.
Private-label and retail-arbitrage teams use the series to reject keywords that only spike in December. Brand managers compare Amazon versus Google to see whether a campaign created shoppers or only browsers. Investors treating consumer names as demand proxies watch category keywords around earnings, then confirm with web traffic and app downloads. Content teams writing buying guides check whether Amazon search already exists before investing in a roundup.
The shared requirement is a phrase that exists in Amazon's search box. Abstract jobs-to-be-done language ("healthier weeknight cooking") will under-read compared with the appliance name shoppers actually type.
Sparse keywords and tiny niches can return thin history or a miss. The 0-100 index is relative to that keyword's window, so two products at 80 are not equal in dollars. Volume estimates are directional, suitable for momentum and not for buying ads to a precise CPC budget.
Amazon search also includes non-purchase looks: parts, manuals, and "is this a scam" queries. A spike after a recall can look like demand. Pairing Amazon with news sentiment and Google Search data separates shopping from damage control.
Quota: 100 requests per month on the free plan, Starter at $19/month. Pulling daily series for hundreds of SKUs every morning will exhaust the month. Weekly category checks plus a short daily list of launch SKUs is the usual pattern.
Trends MCP is one MCP connection covering 25+ sources. Amazon search sits on the same 0-100 index as Google and TikTok, with about 5 years of weekly history. No Amazon developer account is required for the series.
Product Advertising API remains the path for live price and listing cards. It does not replace search-demand history.
Keepa, Jungle Scout, Helium 10, and similar seller suites add rank, fees, and product research UI. Those are operator tools. Teams should check current vendor pricing for seller suites. Trends MCP is the assistant-native demand series, not a replacement for reconciling FBA fees.
Tools for this workflow
get_trendsChart about 5 years of Amazon product search volume for any keyword and see whether purchase intent is steady, seasonal, or in decline.
get_trends(keyword='air fryer', source='amazon', data_mode='weekly')get_growthCompare Amazon purchase intent growth against Google Search and TikTok. Amazon rising after TikTok often means social buzz is converting into active buying.
get_growth(keyword='air fryer', source='amazon, google search, tiktok', percent_growth=['1M', '3M', '6M'])get_ranked_trendsGet a ranked list of the fastest-growing Amazon product search categories this quarter - ideal for finding white-space product opportunities before they become crowded.
get_ranked_trends(source='amazon', sort='qoq_pct_change', limit=30)get_top_trendsSee Amazon's best-selling and top-rated products right now - no keyword needed. Use this to spot consumer demand shifts and product category breakouts in real time.
get_top_trends(type='Amazon Best Sellers Top Rated', limit=25)FAQ