Think you can outsmart Amazon or eBay by creating a fresh account after getting banned? Good luck with that. These platforms have spent years (and serious money) building detection systems that catch linked accounts with scary accuracy.
The days of just using a new email are long gone. Modern marketplaces track way more than most sellers realize.
Your Browser Is Telling On You
Here’s what most people don’t know: your device broadcasts a ton of information every time you visit a website. Screen size, installed fonts, your graphics card specs, timezone settings, browser plugins. Individually, none of this seems like a big deal.
Together, it creates something unique. Security researchers found that 89.4% of these browser profiles are one of a kind. That means platforms can probably identify your specific device out of millions.
And it gets worse. Platforms stack multiple detection methods on top of each other. IP tracking is just the obvious layer. Cookie analysis, hardware fingerprints, and behavioral patterns fill in the gaps. Anyone running multiple seller accounts should check out a more than one ebay account guide to understand what they’re actually up against.
The Detection Stack Goes Deep
When a suspended seller tries coming back with a new account, platforms often catch them within hours. Not days. Hours. The systems pull together payment details, shipping addresses, tax IDs, and phone numbers automatically.
But the really interesting stuff happens under the hood. Machine learning models analyze login patterns and transaction behaviors. This technology creates what’s called a device fingerprint, and it sticks around even when you think you’ve covered your tracks.
Browser storage is another trap. Local storage, IndexedDB, cached images. All of it can hold identifying data that survives a simple cookie wipe. Most people have no idea this stuff exists on their machine.
Timing gives away more than you’d expect too. Two accounts logging in from the same general area at similar times? That’s a red flag the system picks up on.
Getting Technical About It
Canvas fingerprinting sounds complicated, but the concept is simple. Platforms render invisible graphics through your browser and check how your specific hardware handles it. Different setups produce slightly different results. Your computer basically signs its name without you knowing.
WebGL does something similar with your graphics card. Audio fingerprinting analyzes how your device processes sound. Stack these together and platforms have a pretty solid way to identify devices even across different browsers.
TCP/IP fingerprinting goes even deeper, examining how your operating system handles network traffic. Research from IEEE shows these combined methods hit over 95% accuracy when detecting fraudulent account connections. That’s not a number sellers can afford to ignore.
Then there’s behavioral biometrics. How fast you type, how you move your mouse, your scrolling habits. Studies on e-commerce fraud detection show these patterns stay consistent even when people actively try to change them. Your physical habits are harder to fake than you’d think.
Why Platforms Go This Hard
The money involved is staggering. Account fraud drains billions from online marketplaces every year. Banned sellers who sneak back in tend to repeat whatever got them kicked off: review manipulation, counterfeit goods, orders that never ship.
Amazon’s policy is clear. One seller account per region unless you can prove a legitimate business reason for more. Get caught breaking this rule and they’ll suspend every connected account, hold your funds, and potentially take legal action.
Trust keeps these marketplaces running. Buyers expect real products from honest sellers. Every fraudster that slips through chips away at that trust.
Common Triggers
Some signals raise alarms immediately. Same bank account across multiple seller profiles? That’s an obvious one. Matching phone numbers or shipping addresses get flagged too.
Less obvious: shared tools can create connections you didn’t intend. Inventory software, fulfillment services, even certain browser extensions might leave traces that link accounts together.
Network data tells stories too. Logging into different accounts from the same WiFi (even days apart) creates associations. Platform systems maintain huge databases tracking these network relationships over time.
Where This Is Heading
Detection keeps getting better. Platforms process millions of transactions daily, so they spot new fraud patterns fast. Evasion techniques that worked last year might already be obsolete.
Behavioral analysis will only get sharper. Hardware fingerprinting will expand. The systems learn constantly.
For sellers playing by the rules, none of this matters much. Keep one account healthy, follow the guidelines, and you’re fine. Anyone with real business reasons for multiple accounts should just ask for permission upfront. Trying to game the system rarely ends well anymore.