International e-commerce has entered an era of more focused, efficient operations. Whether selling on Amazon, TikTok Shop, eBay, or multiple independent stores, managing multiple accounts has become a key growth strategy. Platform anti-fraud algorithms are evolving just as quickly. Simply clearing the cache, using virtual machines, or sharing a VPN often leaves related stores vulnerable to being suspended together.

1. How platforms link e-commerce accounts

E-commerce risk systems do not link accounts using a single signal. They rely on a multidimensional Bayesian probability-weighted network. When two login environments share several of the following characteristics or anomalies, the system immediately raises their risk score:

  • Network identity (IP & ASN): The IP’s physical location, ISP type, frequent address changes, and listings on abuse blocklists (Spamhaus, UCEPROTECT).
  • Device and browser fingerprints: Hardware characteristics exposed by the operating system, rendering engine hashes, and installed font lists.
  • Behavior and timing patterns: Signs of automation in mouse movements, time spent on pages, request frequency, and interaction patterns.

2. Five common browser fingerprint signals

Modern browser monitoring scripts, such as FingerprintJS Pro and DataDome probes, can collect more than 30 browser characteristics within 20 milliseconds of a page loading:

Detection signal How it is detected Illustrative risk level Mitigation
Canvas rendering hash Render invisible graphics with 2D Canvas and calculate a CRC checksum of the pixel data Depends on the platform Noise injection in the anti-detect browser
WebGL GPU vendor Read the UNMASKED_VENDOR_WEBGL and RENDERER strings Depends on the platform Emulate the driver characteristics of commonly used dedicated GPUs
WebRTC local network exposure STUN requests bypass the proxy and expose the host’s local or real public IP Depends on the platform Force WebRTC to use the proxy’s public exit IP
Installed system fonts Measure text layout with CSS to infer the operating system’s font list Depends on the platform Use the operating system’s standard font set and hide local Chinese fonts
Time zone and location mismatch Compare JS Intl.DateTimeFormat and the time zone offset with the IP’s location Depends on the platform Synchronize the browser’s time zone and coordinates with the IP location

3. Datacenter proxies, ASN classification, and fraud scores

Many sellers buy inexpensive datacenter proxies to save money. To a risk-control system, however, the network classification associated with an ASN (Autonomous System Number) is an unambiguous signal.

When you visit Amazon or TikTok, the platform queries databases such as IPQualityScore, MaxMind GeoIP2, or Scamalytics in the background:

  • Datacenter IP attributes: The ASN type is hosting / data_center (for example, DigitalOcean, AWS, or OVH). This may be one signal in a platform’s risk assessment; it does not determine a universal score such as a fixed high-risk rating.
  • Static residential IP attributes: The ASN type is isp / residential (for example, AT&T, Comcast, Verizon, or Deutsche Telekom). The IP’s reputation must be checked individually; an ISP classification does not guarantee a clean risk score.
// Illustrative fields only; not a live lookup or a guarantee of a provider’s response
// Example residential IP record:
{
"ip": "73.168.21.94",
"connection_type": "Residential ISP",
"isp": "Comcast Cable Communications",
"asn": "AS7922",
"is_crawler": false,
"is_vpn": false,
"is_tor": false,
"fraud_score": 0 // Example value only
}

4. Integrating anti-detect browsers with IP123 static residential proxies

Using a mainstream anti-detect browser as an example, a standard isolated browsing environment is configured as follows:

  1. Get a dedicated static IP: Choose an available country and, where supported, a city in the IP123 console. Copy the issued IP address, port, username, and password for the store’s connection.
  2. Create a browser profile:
    • Operating system: choose Windows 11 or macOS Sonoma to match your target audience.
    • User-Agent: choose a widely used Chrome version that exactly matches the browser engine version.
  3. Configure the proxy and enable automatic synchronization:
    • Select SOCKS5 as the proxy protocol for efficient transport without modifying headers.
    • Enable “Match time zone to IP automatically” and “Match location coordinates to IP automatically”.
    • Set WebRTC mode to “Replace” rather than disabling it entirely. Some websites flag disabled WebRTC as a sign of automation.
  4. Check connectivity and fingerprint consistency: After launching the profile, visit browserleaks.com and whoer.net to inspect the public IP, location, and time zone. Test results help diagnose configuration issues; they do not guarantee that a platform will accept the account.

5. Four rules for consistent store access

1 Keep each store permanently assigned to one IP and one profile

Never reuse the same residential IP across multiple stores, or sign in to the same store from different physical machines or frequently changing IP addresses.

2 Keep payment details strictly separate

Even with 100% network fingerprint isolation, matching bank account holders or credit cardholder names can instantly establish a firm link between accounts in the platform’s payment system.

3 Keep each store on a fixed IP

Assign a dedicated static IP to each store and keep it for daily logins and operations. Frequent location changes may trigger extra verification, depending on the platform’s rules.

4 Match activity to real user habits and local time zones

Operate each store during working hours in its target country where possible. Avoid high-frequency bulk price changes or other sensitive operations at 3 a.m.

IP123 static residential proxies keep a fixed, dedicated IP throughout the rental period. Use separate browser profiles for your accounts and verify each platform’s requirements when configuring your workflow.