Proxies for price monitoring
A price is only meaningful alongside where it was observed. Get the location wrong and you are monitoring a number nobody is actually offered.
The problem
Retailers vary price, currency, availability and shipping by geography. Checked from the wrong country, a competitor’s price is not wrong so much as irrelevant.
Some sites go finer than country — regional pricing, store-level stock, city-level delivery estimates. Others do not vary below national level at all.
Knowing which of those you are dealing with determines whether city targeting is essential or a waste.
What we would configure
- Proxy type
- Residential — retail sites commonly filter datacenter traffic
- Rotation
- Sticky for a full product flow; per-request for a price list
- Targeting
- Country always; city where the retailer varies below national level
- Cadence
- Match the target’s real update frequency, not your dashboard refresh
Residential is the default here because retail platforms are among the most aggressive at classifying hosting ranges, and because an ad or price served to a datacenter IP may not be the one a customer sees.
A real connection
Built from the same parameter grammar the network accepts — copy it and change the country.
user-USERNAME-type-residential-country-de-session-sku_4471-rotation-0One session per product keeps the flow coherent from listing to cart, so the price you record is the price that flow actually produced.
What goes wrong
- Comparing prices collected from different countries as if they were the same market
- Ignoring currency and tax display rules, which change the number before it reaches the page
- Polling hourly against a catalogue that updates daily — you pay for the bytes either way
- Letting a rotation change mid-checkout, which can reset the cart and the price you were measuring
Questions
- Do I need city-level targeting?
- Only if the retailer varies below national level. Check the same product from two cities in one country first — if the price is identical, country targeting is enough and city targeting is precision you are not using.
- Why does the price differ from what my colleague sees?
- Currency, tax display rules and delivery estimates are all applied before the number reaches the page. Two observations are only comparable if the country, and often the session, match.
- How often should I check?
- At the rate the catalogue actually changes. Polling hourly against a daily-updated feed multiplies your bytes without adding a single data point.