How is IP intelligence different from a plain geo filter?

A geo filter only reads the country behind an IP; IP intelligence also reads connection type — home, mobile, datacenter, VPN — plus an accumulated fraud score. They're two different, complementary filters, not two names for the same thing.

Can IP scoring accidentally cut real users?

Yes, if the threshold is set too strictly without accounting for IP type — mobile ranges in particular sometimes score high just because of how a carrier's address pool is shared. Mobile and residential traffic should generally get a softer threshold than datacenter.

Is a separate service like IPQS still needed if the PWA builder has antibot built in?

If IP type and fraud score are already checked at the builder's cloaking layer, a separate subscription is usually redundant — unless a campaign needs narrow data the built-in filter doesn't cover.

How often does IP and fraud-score data get updated?

Datacenter ranges and IP reputation shift constantly, which is why a built-in antibot solution updates on the provider's side rather than as a one-time setup a buyer has to maintain.

IP-Intelligence Antibot Setup

IP-intelligence antibot setup starts by looking past geography, at where a request physically comes from — a home ISP, a mobile network, a datacenter, or a VPN/proxy. This is the layer that catches most review checks and scanners, because they almost never connect from an ordinary consumer IP. Here's how IP scoring works and how to set it up correctly.

What IP intelligence actually is

IP intelligence is data about a specific address beyond plain geolocation: connection type (home ISP, mobile network, datacenter, hosting, VPN/proxy), the autonomous system (ASN) it belongs to, and an aggregated fraud score built from that address's usage history. Where a geo filter only answers 'which country,' IP intelligence answers 'does this request look like a real person on an ordinary home or mobile connection' — two different filters that work best together, not one substitute for the other.

How fraud score catches review checks

Platform reviewers and automated scanners connect almost exclusively from datacenter or cloud IPs — hosting addresses, not residential ones — which reliably score high on fraud or carry an explicit 'datacenter/hosting' type. A rule like 'datacenter and hosting IPs see White, not Offer' closes off most of these checks before they ever reach the live offer page. VPNs and public proxies deserve a separate rule — plenty of real users run through a VPN too, so the raw fact of a VPN matters less here than the overall fraud score.

Setting thresholds and rules

A strict fraud-score threshold that ignores IP type risks cutting real mobile traffic — some carriers' mobile ranges carry an inflated score simply because many subscribers share one small IP pool. A more reasonable setup combines a hard cut for datacenter/hosting IPs with a softer threshold for mobile and residential ranges, so ambiguous cases don't cost real users. Thresholds are worth revisiting per geo — what's normal for one mobile market can be unnecessarily strict for another.

IP intelligence alongside other filters

IP scoring alone doesn't catch advanced checks run from clean residential proxies — those requests formally look like an ordinary user IP. That's why IP intelligence almost always runs alongside a UA filter, geo rules, and a JS challenge: one layer handles the cheap, high-volume checks, and the rest catch the more careful ones. Relying on IP type as the only filter is a common shortcut, and it's usually what breaks the first time a more persistent check comes through.

IP intelligence in APEX

In APEX, IP scoring is part of the built-in antibot cloaking: IP type and fraud score are checked automatically alongside geo, UA, and datacenter filters, with no separate service like IPQS to wire up and no API keys to manage. Rules apply across every domain in the funnel at once and update on APEX's side, so a buyer never has to manually refresh datacenter-range lists or IP reputation data.

FAQ

How is IP intelligence different from a plain geo filter?: A geo filter only reads the country behind an IP; IP intelligence also reads connection type — home, mobile, datacenter, VPN — plus an accumulated fraud score. They're two different, complementary filters, not two names for the same thing.
Can IP scoring accidentally cut real users?: Yes, if the threshold is set too strictly without accounting for IP type — mobile ranges in particular sometimes score high just because of how a carrier's address pool is shared. Mobile and residential traffic should generally get a softer threshold than datacenter.
Is a separate service like IPQS still needed if the PWA builder has antibot built in?: If IP type and fraud score are already checked at the builder's cloaking layer, a separate subscription is usually redundant — unless a campaign needs narrow data the built-in filter doesn't cover.
How often does IP and fraud-score data get updated?: Datacenter ranges and IP reputation shift constantly, which is why a built-in antibot solution updates on the provider's side rather than as a one-time setup a buyer has to maintain.

Build your PWA in APEX in minutes

Cloaking, anti-bot, push, split tests and your own domains — in one service.

Get started free