The real cost of outbound: in-house vs agency

I speak with a lot of founders, and many of them default to the same plan when they decide to test outbound: build an in-house team from day one.

On paper it looks cheaper. In reality, it rarely is, once you factor in salaries, infrastructure, and the number of messaging iterations you need before anything converts.

Before we go further, here’s the quick checkpoint: you should be confident outbound is actually the right motion for you.

If outbound does make sense for your economics, then the next step is simple: in-house or agency?

And since I run an agency myself, you might expect bias :) Fair. I’ll stick to numbers and show the exact cost model we use with US/EU teams when they’re running outbound for the first time.

Btw, I’m Rinat, I run Sally, B2B sales agency, where we build outbound systems for B2B companies in SaaS, fintech, and services, and Crona, an AI-powered lead engine that turns web data into structured ICP lists with verified contacts.

1. What an in-house outbound setup actually costs

Founders often underestimate how much machinery sits behind “a few emails a day”. A functioning outbound system requires three layers: people, infra, and oversight. So you’re gonna need:

People. Full-time SDR with international experience costs minimum $2K–3.5K/month
Infrastructure. Healthy outbound volume typically needs 20–30 domains, 3–4 LinkedIn accounts, enrichment tools, data sources, email infrastructure
Total: $800–1,000/month.
Oversight. Someone senior must review quality, coach the SDR, and keep the system from collapsing. Usually Head of Sales or a senior AE. Minimum ¼ of their time ≈ ~$1K/month in cost.

~$4.5K/month just to get the engine running. No hiring time, onboarding, and operational overhead included.

And outbound doesn’t validate in a month. You need 3–4 months of experiments before you see real signal.

$4.5K × 3–4 months = ~$15K
This is the realistic price of checking whether outbound can work for your team.

2. Why in-house feels cheaper — and why it isn’t

Most budgeting mistakes come from optimistic assumptions:

“Our marketer or AE can own outbound in their spare time.”
Never works. Outbound is a separate motion, and without ownership, it dies.
“We’ll get leads within a few weeks.”
Not true. You’re testing hypotheses, not running a mature funnel.
“We don’t need that much infra.”
You do, otherwise you burn domains, inboxes, and data.

When you stack the real costs, the in-house setup and an agency end up costing roughly the same on small to medium volumes.

The key difference is speed and error rate.

Agencies amortize infrastructure, buy data in bulk, use internal tooling, and avoid the typical traps every new team falls into.

With the same budget you get more experiments, faster iteration, and higher accuracy.

3. When starting with an agency makes more sense

If you’re still validating outbound as a channel, agencies are usually the more efficient entry point:

faster setup
quicker access to good data
tested messaging frameworks
fewer costly mistakes
earlier signal on what’s actually working

Meanwhile, as the founder/manager, you start developing “operator literacy”:
what metrics matter, where conversion breaks, which segments produce results, and what qualification should look like.

This literacy is what enables you to eventually build an internal team without repeating the expensive mistakes.

4. A hybrid model often beats both extremes

Once the channel proves itself, scaling in-house becomes logical.
And this is where most teams miss a key nuance: in-house and agency pair well together, as long as you divide ownership intelligently.

For example:

in-house team works one region; agency covers another
or you split by ICP, product line, or deal size

Around half of our enterprise clients run this hybrid model.
It gives them stable, predictable meeting flow without quality drops.

TL;DR

Outbound only works when treated as a system.

A real in-house system costs about $15K and 3–4 months just to validate the channel.

Agencies cost roughly the same, but give you:
• faster iteration
• fewer mistakes
• earlier signal

Once outbound starts working, scale in-house and split ownership intelligently.