The same lookup answers very different questions depending on where you place it. These are the deployments we see most, and the signals that matter in each.
The problem. Card testing and stolen-card orders arrive from rotating residential and datacenter proxies so that velocity rules keyed on a single address never fire.
How to use the data. Score the address at checkout, weight hosting and proxy classifications heavily, and route elevated-risk orders to 3-D Secure or manual review rather than an outright decline. Country mismatch between billing address and connection is a strong secondary signal.
The problem. One person farms hundreds of free trials, referral bonuses or promo codes by cycling through proxy exits and disposable identities.
How to use the data. Check the address at registration. Treat datacenter and known proxy ranges as a reason to require verification, and track how many accounts you have already seen on the same ASN or prefix — the range is the fingerprint, not the individual address.
The problem. Paid campaigns and affiliate programs quietly bleed budget on clicks that originate from server infrastructure rather than from people.
How to use the data. Log a verdict for every click or impression and reconcile it against conversions. Hosting-originated traffic with zero downstream conversion is the clearest evidence you need when disputing an invoice with a traffic partner.
The problem. Banned players return through VPNs, and multi-accounting or boosting rings hide behind proxy exits.
How to use the data. Evaluate the address at login and at match join. Many studios allow VPNs for latency or privacy reasons but restrict what a fresh VPN-connected account can do — trading, ranked play, voice — until it has established history.
The problem. Content, betting and financial products are licensed per territory, and regulators expect a documented effort to enforce those boundaries.
How to use the data. Combine country geolocation with proxy, VPN and TOR detection so a deliberate location bypass is visible, and keep an auditable record of the verdict at the moment access was granted or refused.
The problem. Automated buyers sweep limited stock in seconds from distributed proxy pools, and real customers see a sold-out page.
How to use the data. Gate the queue rather than the purchase: raise friction for datacenter and proxy connections before they reach the cart, and reserve the fast path for connections that look like ordinary consumer ISPs.
Pick the single step where bad traffic costs you the most — signup, checkout or login — score it in log-only mode for a week, then turn on enforcement once you can see the distribution of your own traffic.
