What did standing up a hundred accounts actually cost you, all in?

My friend, i have doubt and i think nobody write the real number anywhere.

Everybody say build a farm, buy proxy, warm up, is easy. Nobody say what is the cost. I want to make one hundred accounts, not for resell, is for my own offers. Actually i have the time, what i dont have is idea if this is 200 dollar or 2000 dollar.

So please, who already did this - what you pay in total for one hundred? Numbers, proxy, phone, the software, everything. Even if is old price is better than nothing.

I'll give you ranges rather than exact figures because mine are two years old and prices move.

Per hundred, roughly: numbers are the cheap part, somewhere between $15 and $60 depending on how bad the platform is that month. Proxies are where it actually goes - if you're putting more than about five accounts behind one IP you're building a graveyard, so twenty subnets minimum, and residential at that quantity is $150-400 a month, recurring, forever. Software is $0 if you do it by hand and $50-150 a month if you don't.

The caveat that matters more than any of those numbers: the hundred accounts are not the cost. Replacing thirty of them every month is the cost, and nobody budgets for it because the first month always looks cheap.

The cost is measured in months, not dollars. You have the time. That was the expensive part.

whatever number you land on, double it and assume half are dead by christmas

Your threat model is missing from every figure in this thread.

A hundred accounts that only need to survive a campaign is one budget. A hundred accounts where any one of them getting traced back to you matters is a completely different budget, and the difference is not proxies, it's whether the payment for those proxies and those numbers points at a person.

Most people asking this question are in the first case and are being sold the second one. Work out which you are before you spend anything.

Is first case for me. Thank you, this save me money already.

Since nobody has laid out the actual arithmetic, and the ranges above are right but hard to plan against, here is how I'd build the budget for a hundred.

Split the cost into three buckets

The mistake is treating this as one number. It's a setup cost, a monthly cost, and a replacement cost, and they behave completely differently.

Setup, one-off. Numbers, and whatever you pay for the identity material. At a hundred accounts you are looking at roughly $0.20-0.60 per number, so $20-60, plus retries — budget 1.4 numbers per account because a meaningful fraction fail on first send. Call it $30-90. This is the bucket everyone quotes and it is the smallest one.

Monthly, recurring. Proxies dominate and there is no way around it. The relevant question is accounts per IP. Somewhere between three and eight is where people land; below three you are wasting money and above ten you are building a cluster that dies together. At five per IP you need twenty IPs. Residential at that count runs $8-20 each per month depending on how you buy. That is $160-400 monthly, and it does not stop when you stop.

Add automation if you use it. Add a phone farm if you're doing anything the browser can't.

Replacement. This is the bucket that decides whether the whole thing works. Take your monthly attrition rate — mine ran 20-35% in the first sixty days and settled around 8-12% after that — and multiply by your setup cost per account. It sounds small. It is not small, because attrition is not evenly distributed: you lose them in clusters, and a cluster loss takes the proxy's reputation with it, so you're often replacing the IP too.

A realistic first-year figure

Bucket	Month 1	Steady state
Numbers	$60	$15/mo
Proxies	$280	$280/mo
Software	$0-120	$0-120/mo
Replacement	—	$40-90/mo

So somewhere around $350 to start and $350-500 a month to hold position. Over a year that is four to six thousand, which is a different conversation from the two hundred dollars people imagine when they start.

I have not run a farm above about 250 accounts, so I don't know where the economics change. They do change — the people running thousands are not paying per-IP retail — but I can't tell you at what point.

Summary: the setup is cheap, the holding cost is the business, and if the monthly number doesn't work then the setup number never mattered.

The replacement bucket is the one I wish someone had shown me in 2021. Good post.