The price depends on where you appear to be
Geo pricing is standard practice in retail, travel and software. A page fetched from Frankfurt and the same page fetched from Chicago frequently differ, and not only by currency conversion. If you are monitoring prices for a market, you have to appear to be in it.
Residential returns the retail price, datacenter often does not
Sites that price by geography usually also detect datacenter traffic, and what they serve it is either a block or a default price that no customer actually pays. A residential IP in the target country returns the page a shopper sees, which is the only figure worth recording.
City targeting matters more than you would expect
Anything with local stock or delivery, groceries, large items, same day services, prices at the postcode. Country level targeting is not fine grained enough for those categories, and city targeting is available on the residential pools.
Consistency beats volume
For tracking a price over time, the same location checked on a schedule is far more useful than many locations checked once. Sticky sessions hold one exit for the length of a check, so a multi step flow, search, product page, cart, stays coherent.
What to buy
Residential, per gigabyte, with country targeting at minimum and city targeting for anything with local stock. Sticky sessions if your check spans more than one page load.
Common questions
Do I need residential proxies to track competitor prices?
Usually not. Most retail sites serve public prices to anyone, so datacenter addresses collect them at a fraction of the cost. Move up to residential only for the specific sites that block you, and keep the cheap pool for the rest. Paying residential rates across a whole catalogue when nine tenths of it answers a datacenter request is the most common way to overspend here.
Why do I see different prices than my customers do?
Because the price you are shown depends on where the request appears to come from. A retailer can vary price by country, and almost all of them vary currency and tax presentation that way. If you want the price a shopper in Berlin sees, the request has to come from Germany, so country targeting is the feature that decides whether the numbers mean anything.
How often can I check without getting blocked?
There is no universal number, because the limit is set by each site and is usually a rate per address rather than a rate overall. The practical approach is to spread requests across more addresses rather than to slow down, and to watch for the response changing shape before it starts failing outright. A page that quietly starts returning a cached or default price is more dangerous than one that returns an error.