Cold email agency vs in-house SDR: the real cost comparison
Updated July 2026 // by Mark Glazer
The short answer: an in-house SDR costs the salary plus the same tooling stack an agency runs, about $2,841 a month at 50,000 emails, plus months of ramp before meetings arrive. An agency spreads that infrastructure and skill across clients. In-house wins when the product needs deep discovery and the volume justifies a team.
Most build-or-buy comparisons stop at salary versus fee. Both halves of that framing are wrong: the salary is the start of the in-house bill, not the total, and the honest case for hiring in-house is real. Here is the full stack on both sides.
The salary line is real, and it is not the total
The Bridge Group's 2025 SDR research report puts median on-target earnings for a US SDR at $80,000, about $55,000 of it base salary. That is the visible line in the budget. What sits on top of it rarely makes the spreadsheet: employer taxes and benefits, a manager's time (SDR is a high-coaching role, and that time comes from someone senior), recruiting cost if the hire misses, and re-recruiting cost when a good rep is promoted or leaves. None of those items settles the build-or-buy question by itself, but they all point the same direction: the true people cost sits meaningfully above the OTE line.
The tooling stack does not disappear when you hire
The common assumption is that a salary replaces agency-style costs. It does not, because the machinery of cold email is identical whoever operates it: mailboxes and dedicated sending domains, a sequencer, data and enrichment tools, live email verification, and scheduling. Our published fee table on the ROI calculator prices that stack at $2,841 per month at a 50,000-emails-per-month volume, which is a little over $34,000 a year before your SDR sends a single message, on top of compensation. The line-by-line breakdown, and what agencies themselves charge, lives on the cold email agency cost page. The point here is narrower: compare salary plus stack against the agency fee. Never compare salary against fee.
Two ramp clocks, and only one of them is short
Cold email has a hardware clock and a human clock. The hardware clock is fixed either way: a new mailbox needs two to two and a half weeks of warmup before it carries campaign volume, whoever owns it. The human clock is where the paths split. A new SDR has to learn the product, the ICP, the objection patterns and the tooling before their output looks like output, and that is measured in months, not weeks, with salary running the whole time. An agency's ramp is mostly just the hardware clock, because the operators, playbooks and copy patterns already exist. If you need pipeline this quarter, ramp time, not cost, is usually the variable that decides.
Coverage: one person versus a system
Volume in cold email is an infrastructure property, not an effort property. Since 20 to 30 emails per mailbox per day is the defensible ceiling, real volume comes from a fleet of warmed mailboxes and the monitoring around it; the capacity maths is worked through in how many cold emails per day. An in-house program concentrates all of that on one person. When that person is on vacation, out sick, or gone to a new job, the sequencer may keep sending but the judgment stops: nobody is reading rejection strings, catching a bounce-rate slide, or pulling a list that started going stale. A system with several people watching it has no single point of failure. For a sense of what sustained system output looks like, one portal-tracked client program ran 1,080,000 emails over 12 months and produced 263 sales-qualified leads; that consistency is a property of the system, not of any one person's calendar.
When in-house wins
This is not a one-answer question, and the honest cases for building are strong ones:
- Complex products that need deep discovery. If the first conversation requires real technical depth, an SDR who lives inside your product will out-converse any outside team. Buy volume where the conversation is simple; build where the conversation is the product.
- Enough volume to justify a team. In-house economics improve with headcount: the tooling stack costs roughly the same whether one rep or four sit on top of it, management amortises, and the learning compounds inside your company instead of at a vendor.
- An existing RevOps function. If you already run the stack, own the data and have playbooks that work, adding a rep is a marginal cost. The build-versus-buy question is really a question about whether the machine already exists.
A verdict framework
Three questions decide most cases:
- Can you wait months for full output? If pipeline is needed this quarter, both ramp clocks favour buying.
- Is your sale discovery-heavy? If yes, that pushes toward building, or toward a hybrid where an agency fills the top of the funnel while your own people run the conversations.
- Does the machine already exist? If you have the stack, the data and someone who has run deliverability before, a hire is a marginal decision. If not, you are pricing a build project plus a hire, not a hire.
If two of the three point the same way, that is your answer. The options are also not exclusive: plenty of companies buy first, learn what converts from live campaign data, then build in-house on top of proven copy and a proven ICP.
The comparison at a glance
Agency pricing depends on the model. Ours is revenue share with a lean infrastructure amount, described in full on the pricing page.
Common questions
What does an in-house SDR really cost per year?
Compensation first: the Bridge Group's median $80,000 OTE is a reasonable anchor. Then the tooling stack, about $2,841 a month at a 50,000-email volume, then employer costs, management time and recruiting. Salary alone understates the bill by a wide margin.
How long until a new SDR produces meetings?
The infrastructure is ready after two to two and a half weeks of mailbox warmup. The rep is the longer clock: months of learning the ICP, the copy and the objections before output stabilises, with salary running throughout. Budget for both clocks, not just the start date.
When is hiring in-house clearly the right call?
When the product needs deep discovery a generalist cannot fake, when pipeline demand supports a team rather than a single rep, or when RevOps already owns the stack and playbooks so the hire is a marginal cost. Two of those three and the answer is build.
Do you still need salespeople if you hire an agency?
Yes. An agency covers the top of the funnel: lists, sending infrastructure, deliverability and first replies. Qualified conversations still land with your closers. Build-or-buy is a question about who generates conversations, not who finishes them.
What does a cold email agency cost?
Models range from retainers to pay-per-appointment to revenue share, and the numbers vary as much as the models. We keep our published fee table and the full market comparison on the cold email agency cost page rather than repeating it here.
Run both sides of the maths before you decide
The ROI calculator publishes our whole tooling table and funnel benchmarks, and the pricing page explains the revenue-share model in plain terms: a lean infrastructure amount, and we earn when you close.
Run the ROI calculator How pricing works