Platform Operations · 2026-04-13
What Is Amazon Account Association? How Multi-Account Detection Works
Amazon tracks far more than your login credentials. Understanding how account association works — and what triggers it — is essential before you ever create a second seller account.
What Is Amazon Account Association? How Multi-Account Detection Works
The Question Nobody Asks Until It Is Too Late
A seller opens a second Amazon account. Different email, different bank account, different business name. Everything looks separated. Six weeks later, both accounts are suspended with the same message: "related accounts detected."
The seller is confused. Nothing was shared between the accounts. Or so they thought.
This is account association in action — and it operates on data layers that most sellers never consider. Understanding how it works is not about finding loopholes. It is about understanding what platforms actually see when they look at your accounts.
What Is Account Association?
Account association is the process by which a platform determines that two or more accounts are controlled by the same person or entity. Amazon, Walmart, eBay, and other marketplaces all use variants of this system.
The purpose is straightforward: marketplaces enforce a one-account-per-seller policy to maintain a fair marketplace. If a seller is suspended for policy violations, they should not be able to simply open a new account and continue the same behavior. If a seller manipulates reviews or pricing, they should not be able to do so across multiple storefronts.
Association detection is not about punishing legitimate businesses. It is about preventing abuse at scale. But the system does not distinguish intent — it only sees data patterns.
The 5-Layer Detection Model
Amazon's association detection is not a single check. It is a multi-layered system where each layer provides independent signals. When multiple layers align, confidence in association increases.
Layer 1: Address and Identity Data
The most obvious layer. Amazon cross-references:
**Business registration address** — the address on your seller account
**Personal address** — home address provided during verification
**Bank account holder address** — the address associated with your deposit account
**Credit card billing address** — the billing address on your charge card
**Tax identification** — EIN, SSN, or equivalent national tax ID
If two accounts share any of these data points, they are immediately flagged for review. This includes partial matches — same street address with different suite numbers at a known commercial mail receiving location will trigger the same flag as an exact match.
Layer 2: Network Identity (IP and ASN)
Every time you log into Seller Central, Amazon records your IP address. But it goes deeper than the IP itself:
**ASN (Autonomous System Number)** — identifies which network operator owns the IP range
**IP geolocation** — physical location estimated from the IP
**Connection type** — residential, commercial, datacenter, or mobile
**Login time patterns** — when you access each account relative to each other
Two accounts accessed from the same residential IP address are strongly associated. Even if you use different devices, the network identity is the same. Logging into Account A, then switching to Account B from the same wifi network minutes later creates a clear association signal.
Layer 3: Device Fingerprinting
Your browser and device reveal dozens of technical signals that combine into a near-unique fingerprint:
**Canvas rendering hash** — how your GPU renders specific graphics
**WebGL parameters** — GPU vendor and renderer strings
**Installed fonts** — the set of fonts available on your system
**Screen resolution and color depth**
**Browser plugins and extensions**
**Hardware concurrency** — CPU core count
**Timezone and language settings**
These signals are collected passively every time you load a page. Two accounts accessed from a device with the same fingerprint are linked regardless of IP address or login credentials.
Layer 4: Payment Instruments
Financial data provides some of the strongest association signals:
**Bank account numbers** — even partial matches across accounts
**Credit card numbers** — the charge card used for seller fees
**Payment processor identity** — the routing information behind your bank account
**Billing name variations** — "John Smith" on one account and "J. Smith" on another
Payment data is particularly difficult to separate because financial institutions perform their own identity verification. You cannot easily open multiple unrelated bank accounts without leaving identity traces.
Layer 5: Behavioral Patterns
The subtlest and most sophisticated layer. Amazon analyzes:
**Product listing patterns** — similar product categories, titles, or descriptions across accounts
**Pricing behavior** — coordinated price changes
**Inventory sourcing** — shared suppliers or fulfillment patterns
**Customer service response patterns** — similar language, response times, templates
**Seller Central navigation patterns** — how you click through the interface
Behavioral analysis is probabilistic rather than deterministic. No single behavioral signal triggers association. But when behavioral patterns align with signals from other layers, they increase overall confidence.
Association Does Not Mean Instant Suspension
A common misconception is that any association signal triggers immediate account suspension. In reality, Amazon uses a tiered response system:
Signal Detection — the system flags a potential association. This happens constantly and most flags are reviewed without seller impact.
Review Queue — flagged accounts enter a review process. Depending on the strength and number of association signals, this may be automated or manual.
Warning or Verification — in some cases, Amazon contacts the seller to verify their identity or explain the relationship between accounts. Legitimate businesses with multiple accounts (approved by Amazon) resolve at this stage.
Restriction or Suspension — when association signals are strong and no legitimate explanation exists, Amazon restricts or suspends the accounts. The severity depends on the seller's history and the nature of the association.
The key insight: association is a confidence score, not a binary flag. A single shared data point might raise the score slightly. Multiple shared data points across different layers push the score past the threshold for action.
The Most Common Triggers
Based on patterns reported across seller communities and published Amazon enforcement data, these are the most frequent association triggers:
Same wifi network — Two accounts accessed from the same home or office network. This is the most common accidental trigger. Even if you use separate devices, the IP address is identical.
Same physical address on two accounts — Using the same home address or business address for two seller accounts. This includes addresses at known CMRA locations where Amazon can see multiple sellers registered at the same facility.
Same credit card or bank account — Reusing a payment instrument across accounts, even briefly.
Shared device — Logging into two accounts from the same computer or phone, even in different browsers. Device fingerprinting persists across browsers on the same hardware.
Family member accounts — A household member opens a seller account from the same network and address. Amazon sees the same environmental signals and flags association.
What Amazon Cross-References
Amazon does not check each layer in isolation. The system cross-references signals across layers to build a comprehensive association profile:
IP address + device fingerprint + login timing = **access pattern**
Business address + bank address + tax ID = **identity cluster**
Product categories + pricing patterns + listing language = **behavioral profile**
When elements from multiple clusters align between two accounts, the association confidence reaches actionable levels.
Amazon also cross-references against historical data. If you used an address three years ago on a suspended account, and a new account appears with that same address, the historical association is detected.
What This Means for Multi-Account Sellers
The detection system is designed to find connections that sellers think they have hidden. The response is not to find more sophisticated hiding methods — that approach leads to an escalating arms race where the platform always has more data.
Instead, the practical approach is to ensure that each account has genuinely independent infrastructure:
**Separate physical addresses** — not virtual mailbox addresses at the same facility, but actual distinct locations
**Separate network connections** — dedicated internet service, not VPN endpoints that share ASN ranges with known proxy services
**Separate devices** — physically distinct hardware, not virtual machines or anti-detect browser instances on the same machine
**Separate financial instruments** — bank accounts at different institutions, tied to different business entities
**Separate behavioral patterns** — different product categories, different operational workflows
The question is whether the separation is genuine or cosmetic. Platforms are increasingly able to distinguish between the two.
For a deeper look at how Amazon's fingerprinting model works at the technical level, see How Amazon Detects Linked Accounts: The Fingerprint Model.
For building seller infrastructure with real address separation, see Bulletproof Seller Infrastructure: Real Address Networks.