How Does Helium 10 Get Its Data, and How Accurate Is It?

Helium 10 gets its data two ways. For tools that read your own account (Profits, Refund Genie, Follow-Up and Alerts) it uses Amazon’s API, so the figures are Amazon’s. Everything else, including competitor sales estimates, keyword search volume and rank, is Helium 10’s own estimate, produced by AI and machine-learning models from data it parses daily, using methods it doesn’t disclose (help center, checked 10 October 2026).

That split decides how far you can trust a number. Helium 10’s help-center article, dated 30 September 2025, says the account-based tools represent “actual Amazon data”, although Refund Genie’s refund amounts are still estimates. For the rest, it calls the results its “best interpretation” of sales, search volume, relevance and rank, and says the methods are proprietary.

Correction (10 October 2026): an earlier version of this page listed Inventory Protector among the tools that use Amazon’s data, said Helium 10 draws on 450 million ASINs, quoted 98% search-volume accuracy against Brand Analytics, and described anonymized user activity as a third data source. Helium 10 retired Inventory Protector on 30 September 2025, and we found no Helium 10 source for the other three claims, so they are removed.

Which Helium 10 numbers are Amazon’s and which are estimates

The Walmart row comes from Helium 10’s Magnet guide (August 2025), which says Walmart search volume is a proprietary estimate “currently in beta” with no exact or historical data. More on that in our Helium 10 for Walmart guide.

How your account data reaches Helium 10

Your own figures arrive through what Helium 10 calls a token: the authorization you give it in Seller Central, Walmart Seller Center or TikTok Shop. Helium 10’s token FAQ (August 2026) adds the rules:

• each Amazon, Walmart or TikTok account can be connected to only one Helium 10 account;
• you must connect with the main Seller Central login, not a sub-user;
• disconnecting a token stops the sync and permanently deletes the history tied to it, even if you reconnect later;
• the number of tokens depends on the plan: 2 on Platinum (and on a free account), 10 on Diamond.

Walmart accounts now connect through an authorized app rather than API keys. Helium 10 says the app has read and write access only to generate reports, and read-only access to items, orders, inventory, shipping and returns.

Does Helium 10 have a public API?

No. Helium 10 doesn’t offer a public API for third-party developers, and its help center documents no REST API for customers: a search for “API” on 10 October 2026 returned 19 articles, none describing one. What Helium 10 does offer is the Helium 10 MCP connector, which lets AI tools such as Claude and ChatGPT query your Helium 10 data. It is included with Diamond and above, with 1,000 calls a month. Our Helium 10 API page covers what it can do, what extra calls cost, and which competitors do sell a documented API.

How accurate is Helium 10?

Helium 10’s help center says its tools “provide useful data for research purposes and can be used confidently”. The specific accuracy figures it publishes all come from its own studies:

• Sales estimates, November 2024. Helium 10’s study of 29,906 Amazon products in 32 categories, using July 2024 sales, reports 89.59% accuracy for Xray’s estimates against 60.00% for Jungle Scout on its chosen measure, with half of Helium 10’s errors between −9% and +13%. The post says 59% of the sampled ASINs had no Jungle Scout estimate and were left out, and it doesn’t publish the underlying data.
• Search volume, August 2024. A smaller study ranked 31 keywords from one brand the author sells, checking Helium 10’s order against Amazon Brand Analytics and Jungle Scout’s against Search Query Performance. Helium 10 had all but two in the same order (93.5%); Jungle Scout had 13 of 31 (41.9%). It tests ranking order, not the volumes themselves, and the two tools were checked against different Amazon reports.
• Homepage claim. Helium 10’s homepage FAQ says that on average it estimates 38.2% of product sales exactly right, against 29.5% for its “closest competitor”, and that its estimates fall within 22 units above or below actual sales. It gives no date, sample or method, and on 10 October 2026 the neighbouring FAQ answer still quoted plan prices from before Starter was retired.

Helium 10 cites no independent audit for any of these figures. Helium 10’s sales-data post itself notes that Jungle Scout has published a 2019 case study reaching the opposite conclusion. The most useful advice comes from Helium 10’s own 2020 post: rather than comparing tools with each other, compare a tool’s estimates with actual Amazon data you have access to. The same post notes that the estimate often needs 30 days of sales data to be accurate.

What to trust and what to verify

How to cross-check Helium 10 data

• Check estimates against products you sell. Look up your own ASIN in Xray and compare the monthly estimate with Seller Central. If the gap is small on your products, the estimates for similar products are more believable.
• Use Amazon’s own search data. Brand Analytics requires a Professional selling account and Brand Representative status for a brand in Brand Registry (Amazon). Its Search Query Performance dashboard shows query volume, impressions, clicks, cart adds and purchases for searches that led to your brand. Our guides to Amazon Brand Analytics and Search Query Performance for keyword research show how.
• Check price and rank history. Keepa’s charts show how a product’s price and sales rank moved; our Helium 10 vs Keepa comparison explains how the two fit together.
• Be more careful when: you are committing to a large inventory order on one estimate, the category is seasonal or volatile, the product is new or was recently out of stock, or you are working with Walmart data.

How this compares with Jungle Scout

Jungle Scout also sells estimates. Its help center says sales figures come from its AccuSales engine and reflect the last rolling 30 days, not a forecast (Jungle Scout, August 2025). Neither company publishes its method in full, and each publishes studies that favour itself. One concrete difference: Jungle Scout documents an API, and Helium 10 doesn’t. Our Helium 10 vs Jungle Scout comparison covers the rest.

For what each Helium 10 tool does with this data, see the Helium 10 features guide; for whether the suite is worth paying for, see our Helium 10 review.