Loading page content.
Loading page content.
Research
Desk research that only ever sees one country's version of the internet is not research, it is a sample of one. Country-targeted exits let a small team observe forty markets the way a customer in each of them does.
The problem
Almost every commercially interesting site now varies by visitor geography. Catalogues differ, launch dates differ, promotional messaging differs, pricing pages show different plan mixes, and some content is not served outside a region at all. A researcher in London studying a US category is reading the export version of it, and the differences are precisely the thing being researched.
The workarounds are worse than the problem. Consumer VPNs concentrate thousands of users behind a handful of well-known addresses, so the sites you most want to study are the ones most likely to block or degrade them. Asking a colleague in-market to send screenshots does not scale past a handful of pages and produces evidence nobody can reproduce.
There is also a cadence problem. Market intelligence is most valuable as a time series — when a competitor changed a price, added a plan tier, entered a country — and a time series needs unattended collection on a schedule, which puts the work back into the same territory as any other automated fetch.
Why proxies
The value of a proxy here is not evasion, it is fidelity. Requesting a page from an exit inside the market returns the artefact the market actually receives, and pinning the country makes the observation reproducible: the same URL, the same exit country, six months apart, is a comparison that means something.
Breadth is the second half. Studying twenty markets requires exits in twenty markets, which is a procurement problem for an in-house team and a checkout parameter here. Because the pool spans consumer ISP allocations rather than a handful of well-known ranges, the pages come back intact rather than degraded.
Recommended product
A fresh residential IP on every request, or one held for 30 minutes
from $1.20/GB
Bulk collection of public filings, standards documents, job postings and press archives is unaffected by ASN reputation and costs roughly half as much per gigabyte.
from $3.50/IP/wk
Longitudinal studies that need the same vantage point every week — tracking how one site treats one returning visitor over time — want a fixed address rather than rotation.
Worked example
Research evidence has to survive the question 'where did this come from'. The sweep below records the exit country and the fetch time alongside the content hash, so a later diff can say what changed and when, in which market.
import hashlib
import json
from datetime import datetime, timezone
import requests
USER = "fp_8s2k4d19"
PASSWORD = "Xk7mQ2pTz9vRn4Ls"
MARKETS = ["us", "gb", "de", "fr", "jp", "br", "au", "in"]
SOURCES = [
("competitor-pricing", "https://competitor.example/pricing"),
("competitor-launch", "https://competitor.example/whats-new"),
]
def fetch(url: str, country: str) -> requests.Response:
proxy = f"http://{USER}-country-{country}:{PASSWORD}@gate.fleetproxy.com:8080"
return requests.get(
url,
proxies={"http": proxy, "https": proxy},
timeout=(10, 30),
headers={"Accept-Language": "en;q=0.8,*;q=0.5"},
)
observations = []
for country in MARKETS:
for name, url in SOURCES:
response = fetch(url, country)
body = response.text
observations.append(
{
"source": name,
"market": country,
# The final URL matters: a market redirect is itself a finding.
"resolved_url": str(response.url),
"fetched_at": datetime.now(timezone.utc).isoformat(),
"status": response.status_code,
"content_sha256": hashlib.sha256(body.encode()).hexdigest(),
"bytes": len(response.content),
}
)
print(json.dumps(observations, indent=2))Replace the credential with the one in your dashboard. The gateway host, port and username format are the same across every product.
Pitfalls
Each of these is common, cheap to fix, and expensive to leave in place. They are listed in roughly the order teams hit them.
Serving English does not mean serving the US, and a German-language page may be the Austrian or Swiss variant. Treat exit country and content language as separate fields, and record both, or a comparison will attribute a market difference to the wrong country.
A single fetch cannot distinguish a permanent regional difference from an A/B test, a cache miss or a deploy in progress. Repeat the observation on a different day from a different exit in the same country before writing it down as a fact about the market.
Many sites bounce a foreign visitor to a regional domain, and the interesting information is in the bounce: which markets have a dedicated domain, which fall back to a global page, and which block outright. Log the full chain, not just the final response.
Images cannot be diffed, searched or re-parsed when the research question changes. Store the HTML and the response metadata; render screenshots from them when a slide needs one.
Questions
Start on 50MB of free residential bandwidth, measure your own targets, and scale into volume tiers that step the rate down as the job grows. Unused bandwidth never expires.
No card required for the trial. Cancel or downgrade at any time.