How many proxies do you need? A sizing guide for drops, restocks and tickets

Every proxy sizing question collapses into one rule: one dedicated IP per task or per account. Everything else in this guide is working out what counts as a task, and which traffic should not be on dedicated IPs at all.

The rule, and why it exists

Bot protection on release-day sites profiles identity: how many carts, entries or logins came from one address, and whether that address behaves like one person. Two tasks sharing an IP look like one visitor doing twice too much — which is precisely the pattern the systems exist to catch. Give every checkout task and every account its own static ISP proxy and each one carries a clean, singular history.

Sneaker drops

Count your checkout tasks, buy that many ISP IPs. A 10-task setup wants 10 IPs; heavy runners scale to 25–100. Account-bound platforms like Nike SNKRS tighten the rule further: one IP per account, permanently — the account should always appear from the same address, release after release. The full per-site breakdown lives on the sneaker proxies page.

Shopify releases

Same math for carting — one ISP IP per task — with one addition: never let monitors touch those IPs. Hyped Shopify stores soft-ban by subnet score, and an IP that polled a product page all night walks into checkout already suspicious. Monitors belong on rotating residential, where volume is the point and billing is per GB.

Retail restocks

Retail flips the ratio. Big-box protection cares about account and session consistency, so the sizing unit is accounts, not tasks: 5–10 dedicated ISP IPs holding one quiet session per retailer account covers most setups. The monitor fleet — however many SKUs you watch — runs entirely on rotating residential and costs GB, not IPs. More on the split on the retail proxies page.

Ticket on-sales

Ticketing is not sized in IPs at all. Entries run on rotating residential and mobile pools, so the budget is data: each queue session is mostly idle waiting, light on transfer, and the real spend scales with how many entries you push. Start with a modest GB balance — it never expires, so an oversized buy is never wasted — and hold sticky sessions for the queues themselves. Details on the ticketing proxies page.

What most people get wrong

• Buying IPs for monitor traffic. Monitors need volume and anonymity, not identity. Rotating residential does it cheaper and safer than any static list.
• Sharing one list across accounts. The moment two accounts appear from one IP, both inherit each other’s risk.
• Sizing for the biggest imaginable day. Delivery is instant — you can top up during the queue. Start at your real task count and scale when your results demand it.
• Running everything on one proxy type. Static and rotating solve different problems; most winning setups run both. The ISP vs residential comparison settles which goes where.

Term length: the second sizing decision

Daily plans exist for the runner who bots one release a month — pay for the drop day, let it lapse. Weekly and monthly terms cost less per day as soon as you run most weekends, and renewals are optional either way: cancel, and the proxies simply run out their term.