Cold Email Agency: You Would Be Paying for the Cheap Half

A cold email agency sells the half of cold email that tooling made cheap: extra domains, warmed mailboxes, a sending tool, a list pulled from a database and a sequence written from a template. The half that earns replies, a list you understand and one sentence only the recipient could receive, stays on your desk whether you pay a retainer or not. For a solo founder, that makes most agencies the wrong purchase.

When I would hire one anyway

Start with the exception, because it is the decision you came for. An agency is the right purchase when three things are already true. Your offer has a measured reply rate from your own sends, so you are buying more of a working thing. The market is big enough that volume is the constraint: thousands of accounts, not two hundred. And one closed deal pays for the retainer several times over, so a pay-per-meeting price is a rounding error.

If you hire, make the contract answer four questions before it answers anything about meetings on autopilot. Who builds the list and from what signal, not just which database. Are the domains and mailboxes registered in your name, so the reputation you paid for leaves with you. How is a qualified meeting defined, and who defined it. And can you read the exact sequence, every step, before it goes out under your name.

The second question is the one people skip, and it is the same test I apply to development tools in Replit alternatives: pick the one you can leave. If leaving the agency means starting from a cold domain, you did not buy a service. You rented a dependency. If none of the three conditions holds yet, what follows is why the retainer would buy you the wrong half, and what to run instead.

What a cold email agency actually sells

Read the agencies' own pages and the offer is spelled out in numbers. Belkins lists an average starter price "from $5,000", 1,500 leads a month, and says its SDRs can send over 10,000 emails a month on some projects. Cleverly promises 10 to 30 qualified B2B meetings every month, done for you. SalesHive sells one flat monthly fee, month to month. LevelUpLeads leads with 1,000 or more new contacts sourced per month.

Put the numbers side by side and the product is clear. It is not writing. It is volume: contacts sourced by the thousand, emails sent by the ten thousand, meetings counted by the month. Every agency page describes the same five-part stack: secondary domains and mailboxes, warm-up, a sending tool that rotates them, a list pulled from a data provider, and a sequence of three to five templated steps with a merge field for your first name and company.

Each of those parts used to be a skill. In 2026 each one is a product with a monthly plan.

“

The retainer buys four rentable parts, assembled. What it cannot buy is the part that decides whether anyone replies.

”

Every part of the stack is rentable except one

Take the five parts in order and price each one for a solo founder.

Domains and mailboxes: a secondary domain costs about the price of a lunch per year, and a mailbox on Google Workspace or Microsoft 365 costs under ten dollars a month. Authentication (SPF, DKIM, DMARC) is three DNS records, and since February 2024 Google and Yahoo require them from anyone sending 5,000 or more messages a day to their users.

Warm-up: a service. Mailivery does email warm-up by exchanging real conversations between real inboxes until the domain has a reputation. Start it three weeks before you need it.

Sending: Instantly, Smartlead, lemlist and a dozen others cost tens of dollars a month and do the rotation, throttling and timing that used to be the agency's craft.

The list: Apollo, Clay and their competitors will give you a thousand contacts matching a title and an industry in an afternoon. So will the agency: that is where it gets them too.

The sentence: nothing. There is no plan for the line that says why this email exists for this person this week. I explained why in the piece on why the Tim Ferriss cold email template still gets replies: the line that works is the one that breaks when you send it to someone else. An agency copywriter who has never used your product or read your support inbox cannot write it, and at 10,000 emails a month nobody is trying to. The merge field is what stands in for it.

So the retainer buys four rentable parts, assembled, plus a person to read replies. What it cannot buy is the part that decides whether anyone replies.

Volume is the agency's edge and your weakest position

The mechanism is arithmetic. A retainer only makes money for the agency if a small number of people can serve many accounts. That forces the same shape onto every client: big lists, a shared sequence with the nouns swapped, and enough domains that no single one carries the reputation cost. The agency is built for the case where the offer is already proven and the only missing input is more of the same emails.

A solo founder is in the opposite case. You do not yet know which segment answers, and your offer changes every month. Your unfair advantage is that you can write the one true sentence, because you built the thing and talked to the last ten customers. That advantage vanishes at volume. The piece on email subject lines for cold emails and the tested-list myth makes the same point from the subject line down: a line that worked on someone else's list tells you about their list.

The threshold I use: below a few dozen emails a day from two or three mailboxes, you do not have the problem an agency solves. Above it, the thing that got you replies stops working, and you are competing with agencies on their terms, minus their volume.

What I would run instead, this week

The solo version costs less than one agency meeting and takes an afternoon to set up.

Buy one secondary domain close to your main one. Create two mailboxes on it and set the three DNS records. Start warm-up the same day and do not send a cold email from it for three weeks. Pick one sending tool and connect the mailboxes.

Then build the list by hand, and keep it at fifty. Not fifty titles from a database: fifty companies where you can point to a signal that says now, a job post, a pricing change, a launch, a complaint in public. Write each first email around that signal, in the shape Ferriss wrote down in 2008: what you noticed, one clause on who you are, one question they can answer from their phone. Send ten a day. Follow up the way I argued in follow up cold email: stop sending reminders, with one new fact per message, never a bump.

Fifty emails will tell you whether the offer answers, and which sentence did it. That sentence is what you would have needed to hand an agency on day one. Now you have it, and you can decide from a position of knowing.

FAQ

How much does a cold email agency cost?

On the agencies' own pages, the visible prices are a starter "from $5,000" at Belkins, a flat monthly fee with no long-term contract at SalesHive, and quotes based on scope everywhere else. Pay-per-meeting models also exist. Assume a monthly retainer in the thousands.

Do cold email agencies still work in 2026?

For proven offers in large markets, yes. Since the 2024 Google and Yahoo bulk sender rules, volume costs more to sustain, which favors agencies with infrastructure over anyone running one domain hard. Whether it works for you depends on whether your offer already answers at small volume.

Should I start a cold email agency?

Only if you can sell the sentence and not just the stack. Your prospective clients can rent the stack for tens of dollars a month, so the agencies that last know a vertical well enough to write lines a merge field cannot. Pick one industry you have worked in and stay there.