Proxies for Farfetch read a global luxury marketplace from behind Akamai. On 24 August 2026 we sent farfetch.com a plain request from a home connection and from a datacenter server, and both returned an Akamai "Access Denied" 403. The wall is Akamai, judged on the request from either address, so a real browser is required. Behind it sits a marketplace whose defining feature, for a data project, is that price is a function of where you are and which boutique the item comes from.
Farfetch, owned by Coupang since 2024, does not hold inventory; it connects several hundred boutiques, brands and department stores to shoppers in more than 190 countries, coordinating shipping and taking a commission, with localized currency, tax and duties by destination (Latterly, Farfetch business model). That makes cross-country landed-price intelligence the main reason to point proxies at it, close to SSENSE but with a boutique-network twist.
What proxies work best for Farfetch?
Residential proxies exiting in the country whose prices you read, behind a real browser, because the landed price follows the destination and Akamai refuses bare clients. A static ISP proxy in the account's country for checkout. Datacenter proxies get the 403 on contact.
Akamai at the door, localization behind it
Akamai decided on the request, the TLS handshake and the header order a real browser sends with client hints (what is JA3/JA4 fingerprinting, how to scrape past Akamai). Past the door, the localization is the data. Farfetch sets language, currency, tax and duty treatment by destination, so the landed price of a piece depends on where it ships, and a dataset needs one row per country per item with the currency and duty context recorded. Geo-testing with proxies confirms the exit lands where you intend.
One item, many boutiques
The marketplace structure is the second axis. Because Farfetch lists stock from many boutiques rather than owning it, the same item can be offered by more than one seller in more than one source country at once, each with its own price and shipping math. The useful data is per-seller and per-source-country, not one blended number, so a price model captures the boutique spread the way it would any marketplace (proxies for Grailed covers the resale side of the same market).
Item viewed
from an exit country
Localized price
currency, tax, duties
Boutique source
several may list it
Landed price
per country, per seller
Which proxy type fits which job
| Farfetch job | Proxy type | Why |
|---|---|---|
| Cross-country landed-price comparison | Residential, one pool per country | The price localizes by destination |
| Per-boutique and per-source capture | Rotating residential | Many sellers per item |
| Seasonal sale and markdown tracking | Residential in that country | Prices vary by market |
| Brand and catalog collection | Rotating residential | Volume across a pool |
| Account and checkout | Static ISP in the account's country | Consistent country signals |
| Parser tests against saved HTML | Datacenter or free | The live site refuses bare clients |
Rotating residential IPs pass Akamai and land in the country you choose, and ISP proxies hold a country-correct address for an account. The two are compared in rotating vs static residential proxies.
Setup
Drive a real browser so the sensor is satisfied, and keep the cookies (proxies for Playwright). Run one residential pool per country you price, verify the exit country, and record the currency, duty context and boutique source with every price. Read the boutique spread, not one number. For accounts, pin a static ISP address in the account's country.
Sizing
Country is the first multiplier: one pool per market. Within each, find how fast one residential IP reads behind a browser before Akamai challenges it, and add exits. Budget browser-level bandwidth for an image-heavy luxury catalog (headless browser proxy bandwidth cost). Our pricing is pay-as-you-go with a balance that does not expire, so pricing several markets in a burst costs only what it uses.
Free versus paid for Farfetch
The test settles it: a datacenter client with no browser got the 403 on both IPs. Free proxies confirm which country a page serves and that an address is alive, which our free proxy list and proxy checker cover. Reading landed prices needs paid residential in each country, from $0.44/GB pay-as-you-go with no KYC, behind a real browser. The line is in datacenter vs residential proxies.
Staying unblocked
- Do not retry the 403 without a browser. Akamai will keep refusing; the fix is the sensor.
- Match the exit to the market. A mismatched exit reads the wrong landed price.
- Capture the boutique source and duty context. A landed price without them is incomplete.
- Pace like a shopper. A steady flood from one IP is what Akamai catches.
- Watch the challenge rate per country pool. A rising 403 share means slow that pool. The full list is in avoiding IP bans while scraping.
What a proxy does not do here
A proxy lands a browser in the right country and keeps accounts apart. It does not run Akamai's sensor, and it does not equalize the localized landed prices, which are Farfetch's by design. Scraping farfetch.com runs against its terms of use, a risk that stays with you regardless of the IPs. What a country-correct residential exit does is read the local landed price a datacenter client cannot reach, across a boutique network where the source matters as much as the destination. Confirm the destinations with the free proxy list, then price each market with residential from $0.44/GB.
Sources
- Latterly, Farfetch business model: the boutique-marketplace model, 190-plus countries, localization and commission.
- Business of Fashion, Coupang completes Farfetch acquisition: the 2024 ownership change.
- Akamai, Bot Manager: the vendor behind our 403.
- HProxy test on 24 August 2026: plain GET requests to farfetch.com from a residential connection and a datacenter server, both returning an Akamai Access Denied 403.