What is a residential proxy — the short answer to this question is clear: IPs that exit through the internet connections of real home users. Behind the phrase “a real home connection”, however, lies a highly layered architecture that makes it possible to reach millions of exits from a single gateway address.
In this article we unpack that architecture step by step: how the pool is formed, which nodes a request passes through, how a session is preserved, and why it is structurally different from a datacenter proxy.
Where Does the Pool Come From?
Residential IPs are not rented in a data center — they belong to real subscribers. Providers usually obtain this access in one of three ways:
| Source | How it works | Transparency |
|---|---|---|
| Consent-based SDK | An app offers the user “share bandwidth in return for an ad-free version” | The user is informed and gives consent |
| Rewarded network (peer program) | The user signs up directly and is paid for the traffic they share | The most transparent model |
| ISP partnership | The provider rents an IP block directly from an operator | Offered separately as ISP proxy |
Pools that route traffic through a device without the user's knowledge and consent are a serious problem in the industry. When choosing a provider, asking how the IP source was obtained matters both ethically and legally. We covered this topic in our article on the ethical and legal framework in detail.
The Journey of a Request
The target site sees no signal whatsoever that the request passed through a data center; the source ASN is a real internet service provider.
The critical detail is this: the exit node does not read your traffic — it only carries it. On HTTPS connections the content is encrypted end to end, so the node owner cannot see what you visit; they only know that a certain amount of bandwidth has been used.
The Duties of the Gateway Layer
All of this complexity is hidden from you. On the client side there is only a single address and one set of credentials.
We looked at the details of the gateway architecture in a separate article .
How Is a Node Selected?
The gateway does not scan the whole pool on every request; it selects from a pre-filtered subset ranked by health score. The criteria that affect selection:
- Geographic filter: The country, region or city you requested.
- Online status: Is the node on the network at that moment? Home devices are not online all the time.
- Bandwidth capacity: How much load the node can carry at that moment.
- Last used time: A cool-down period that prevents the same node from being selected repeatedly.
- Target history: Whether that node has recently received errors on the same target.
The exit point is a real home connection: upload speed is limited, the link is asymmetric and the user may be watching a video at the same time. The reason a residential proxy is slower than a datacenter one is not the provider's infrastructure but the nature of the exit point.
Rotating and Sticky Behavior
The same pool can be used in two different ways. The difference lies in the session parameter you send to the gateway:
If you do not send a session key, every request is selected independently from the pool. When you do send a key, the gateway tries to route you to the same node.
The practical details of session management are covered in sticky session article, and rotation strategies in our rotation settings article .
Structural Differences from Datacenter
The two architectures are strong on opposite axes. That is why the question “which is better” has no answer; only “for which job” is meaningful.
For a detailed comparison, see residential vs datacenter article.
Why Is Bandwidth Billed by the GB?
Because the provider's cost is not per IP but per byte carried . Every node owner in the pool is paid, or compensated, in proportion to the traffic they share. That cost is directly proportional to the data transferred, and so is the pricing.
This structure explains why residential traffic is many times more expensive than datacenter. To reduce your consumption you can apply the methods in our bandwidth article .
Reading Pool Size Claims
Figures such as “50 million IPs” usually mean the total unique addresses seen in the pool over the course of a year . What matters to you is how many nodes are concurrently online in a given country right now. There can be a large difference between these two numbers.
A practical way to measure it: send 200 requests without a session key and count how many different IPs you see. Our trial test article has a ready-made code example.
For Which Jobs Is It the Right Choice?
Residential is a good fit
- Collecting data from heavily protected marketplaces.
- Viewing search results from different countries.
- Ad verification and localization testing.
- Regional price and stock comparison.
- Simulating geographically dispersed users.
Another type is a better fit
- High-volume scraping of unprotected pages → datacenter.
- Panel access that requires a static IP → ISP.
- Account operations on the strictest platforms → mobile.
- Gaming scenarios where low latency is critical → static solutions.
- Large file downloads → the per-GB cost becomes exorbitant.
Summary
A residential proxy is a layered architecture that gathers a pool of real subscriber connections behind a single gateway. Its strength comes from the ASN source: the target site cannot distinguish the request from traffic coming out of an ordinary home connection. In return you accept variable speed, GB-based cost and a “best effort” guarantee on session stability. For product details see our residential proxy page, and for a setup test see our proxy checker tool you can use.