The target sees the exit, not the gateway
A common support pattern: the user sends the proxy connection string and says the target blocked it. The target never saw that string. It saw the public IP where the request left the proxy network, then scored it on ASN ownership, country, subnet history, and the request itself. The gateway address stays inside the tunnel.
Test one route before scaling. If the exit is producing bad results, adding volume only obscures which part failed and makes the bandwidth meter harder to read.
The gap shows up clearly in response bodies. Same script, same URL, same cookie jar, two different exits: one returns a 200 with 200 KB of usable HTML; the next returns a 200 with 12 KB — a block page or a challenge stub. Status-only monitoring misses that entirely. The exit IP is what changed between those two requests.
Residential means consumer ASN, not a guarantee
Residential typically means the exit belongs to a consumer ISP or ISP-style network rather than a hosting network. Comcast, AT&T, Vodafone, and Deutsche Telekom look different from AWS, Hetzner, OVH, or DigitalOcean when a target checks ASN ownership. That difference matters a lot on some targets and barely at all on others.
A residential exit can fix a route-class problem when the target refuses datacenter traffic. It does not fix a bad browser fingerprint, a country and timezone mismatch, cookies that do not fit the session, or request timing no real user would produce. Those are separate variables.
Rotation window determines session identity
With rotating residential, the exit can change per request or per session. Per-request rotation works for broad public collection where each request stands alone. It breaks multi-step flows, because a normal visitor does not jump between ISPs or cities on every click. Session rotation holds the exit longer. On a Proxynade pool, the lifetime-<minutes> token in the username controls the window — for example, rt97db6958d9-plan-volume-country-us-lifetime-30 keeps one exit for up to 30 minutes.
Static ISP is the right shape for account-like workflows, monitoring jobs, and anything that needs to return from the same network across multiple sessions. A stable exit does not prevent the target from remembering it, and a rotating pool can still repeat a tired subnet too often. The question is whether the exit behavior matches the workflow, not which plan label is attached.
Metered bytes include the requests that failed
Volume Residential is $0.89/GB. Premium Residential is $5.00/GB. Static ISP is per IP with unlimited bandwidth. Those numbers are only useful after accounting for waste.
App-side counters measure saved rows or accepted bodies. Provider-metered bytes include redirects, retries, block pages, challenge pages, and responses the parser discarded. A cheaper plan that burns retries on a hostile target can cost more than a pricier plan that delivers clean responses. The dashboard network logs show host, outcome, latency, and byte totals per request — that is where the actual cost-per-result calculation lives, not in the per-GB rate.
What a useful debug row looks like
Exit IP, ASN, owner, subnet, country, session id, status code, body size, row count, retries, and billed bytes. If the country is correct but the timezone is wrong, the client configuration is the suspect. If the same subnet repeats across too many requests, the route sample is narrow. If body size collapses while status stays 200, the request is being answered by a block layer, not the real page — subnet history and listing status can be checked against Spamhaus.
The short definition: a residential exit IP is the address the target sees. The longer operational definition is that the target scores that address together with the client fingerprint, session history, and traffic pattern. Debugging stops when all four align.