Proxy windows
What the 1d to 90d windows of InternetData's proxy databases mean, how their columns count an address, and how to choose the window for a job.
New to InternetData? Make your first lookup
On this page
Each proxy family (residential, mobile, datacenter and public proxies) comes in six windows: 1, 7, 14, 30, 60 and 90 days. A window holds every address seen working as a proxy exit within that many days.
Choosing one
| Window | Holds | Suits |
|---|---|---|
| 1d, 7d | Exits active right now | Blocking at a login or a checkout, where a false positive costs a customer |
| 14d, 30d | Exits active this month | Scoring and review queues |
| 60d, 90d | Every exit seen this quarter | Investigations, analytics and enrichment, where coverage matters more than freshness |
Home and mobile addresses change hands, so a residential exit seen once two months ago may be an ordinary household today. The shorter the window, the fewer such addresses it holds.
Reading the columns
first_seenandlast_seen: the first and last day the address worked as an exit within the window.hits: how many times it was seen. An address seen many times is a dependable exit; one seen once may be a passing one.hits_days_pct: the share of the window's days it was seen on.providers_num: how many providers it was seen with in the window.
A rule can weigh these, requiring a minimum hits so that repeated sightings count for more than a single one.
Each window's page under Database Reference has its own columns and a sample. The API answers from these same datasets in its resproxy, mobproxy, dcproxy and pubproxy details.