Cold email agency vs in-house SDR
Updated
An agency gets you to a tested message faster; an in-house SDR gets you closer to the customer. The deciding question is whether you already know which message works.
Both options cost real money and both can work. This page sets out what each actually costs, how long each takes to produce a first meeting, and the conditions under which each is the wrong choice. Every external figure is sourced and dated at the bottom of the page.
At a glance
| Dimension | In-house SDR | Cold email agency |
|---|---|---|
| Median US base salary | $60,000 | n/a |
| Median US on-target earnings | $85,000 | n/a |
| Typical published engagement range | n/a | $3,000–$10,000 per month |
| Message angles tested at once | One person, in sequence | 10 to 15 in parallel |
| Who owns the sending infrastructure | You | You, at the end of the engagement |
| Takes live sales calls | Yes | No |
| Product depth over time | Builds continuously | Stays shallower |
| What happens if you stop | You keep the person | You keep the infrastructure and the data |
What does an in-house SDR actually cost?
The median base salary for a US sales development representative is $60,000 and median on-target earnings are $85,000, across 8,334 verified submissions on RepVue. Salary is not the whole cost. US Bureau of Labor Statistics data puts benefits at 30.0% of total employer compensation cost in private industry, so wages represent roughly 70% of what an employer actually spends.
Applying that ratio to on-target earnings gives a fully loaded cost of roughly $121,000 a year before tooling, recruiting or management time. That number is arithmetic, not a surveyed figure: $85,000 divided by 0.70. Substitute your own salary and the same ratio holds.
Two things that arithmetic leaves out. Recruiting costs money and time before anyone starts. And an SDR needs sending infrastructure, data and a sending platform whether they are employed or not, which is the same bill an agency engagement already includes. The inbox calculator will size that infrastructure on your own volume.
What does a cold email agency actually cost?
GrowthPal does not publish its own pricing, and the reason is on the FAQ: scope moves with market size and sending volume, so a number on a website would not mean much. What GrowthPal does publish is the range it considers honest for the market: $3,000 to $10,000 per month for a real system, and $200 to $800 per held meeting on pay-per-meeting deals.
That range covers the build and the ongoing work: sending domains, inbox warmup, list engineering, the message angles and the reply classification. A fuller breakdown sits on the cost page. Commission-only arrangements are a fourth model and GrowthPal does not take them, because the research, infrastructure and writing all happen before the first reply arrives.
How fast does each get to a first meeting?
A GrowthPal engagement takes two to four weeks from kickoff to campaigns sending, with first replies within days of the first sends and the winning message usually identified by week eight. Those are GrowthPal's own published figures and they assume the infrastructure is being built from scratch.
An in-house SDR has a ramp period before they reach full productivity, and it is longer than a build. The honest caveat on both sides: speed to a first meeting is not the same as speed to a repeatable channel, and the second one is what you are buying.
Where each one wins
Neither option is better across the board, and the honest answer depends on which of two things you are short of: knowledge of your market, or a tested message. An in-house SDR accumulates the first. A cold email agency produces the second faster.
Where an in-house SDR wins
An SDR is in your building. They pick up product detail, sit in on calls, hear objections first-hand and carry a relationship from first touch to close. They can take a live call the moment someone answers. Over a year they become a person who knows your market, which is an asset an agency engagement does not leave behind. If your product needs real discovery on the first call, that matters more than testing throughput.
Where an agency wins
An agency runs 10 to 15 message angles at once against your market, then scales the one that works. One SDR tests sequentially, which takes months to cover the same ground. The infrastructure already exists: domains, warmup, deliverability monitoring and reply classification. On a documented month, one GrowthPal client reached 67,934 contacts and got 449 interested replies, roughly 1 per 150 contacts reached. How these numbers are counted defines every term in that sentence.
Choose an in-house SDR if…
- You already know which message works and you need more conversations, not more tests.
- Your market is small or highly technical, and precision matters more than volume.
- Your product needs real discovery on the first call rather than a handoff.
- You want the market knowledge and the customer relationships held in-house.
- You have someone senior who can actually manage and coach an SDR. This is the condition most companies fail. A first SDR with no manager, no list engineering and no message testing is the usual reason the experiment gets written off.
Choose a cold email agency if…
- You do not yet know which message lands, and finding out quickly is the point.
- You have no sending infrastructure and do not want to build it.
- You have people who can take meetings but nothing filling the top of the funnel.
- Your market has roughly 30,000+ reachable contacts, which is the volume needed to test 10 to 15 angles and reach a real answer.
GrowthPal is also explicit about when outbound is the wrong call at all: if nobody is free to take the meetings, if you need results by Friday, or if commission-only is a requirement. Those are set out on the homepage and on the cold email agency page.
Can you do both?
Yes, and for most companies past the first year it is the end state. The system finds and proves the message and fills the calendar; the people take the meetings and carry the relationships. The order that usually works is agency first to find the message, then hire against a message you already know converts, rather than asking a new hire to discover it.
Sources
- SDR median base salary, median on-target earnings, top-performer earnings and quota attainment: RepVue, Sales Development Representative Salary, 8,334 verified submissions, United States. Read September 21, 2026.
- Benefits as a share of total employer compensation cost: US Bureau of Labor Statistics, Employer Costs for Employee Compensation, private industry, June 2026. Read September 21, 2026.
- Engagement ranges, angle counts, timelines and campaign outcomes: GrowthPal's own published figures, defined at how the numbers on this site are counted.
Salary figures are United States medians and will not hold in the United Kingdom, Canada or Europe. Figures last verified September 21, 2026. All prices and salaries are US dollars.
Frequently asked questions
GrowthPal gets asked four questions more than any others when a company is weighing an agency against a first hire: what the real cost gap is, who owns what at the end, whether an agency can be a stopgap, and what happens to a message that was found by someone else.
Is an agency cheaper than an SDR?
Not necessarily, and cost alone is the wrong test. A US SDR runs roughly $121,000 a year fully loaded on median on-target earnings, and the published market range of $3,000 to $10,000 a month annualises to $36,000 to $120,000. They overlap. The difference is what you get for it: one person testing sequentially, or a system testing 10 to 15 angles at once.
Who owns the sending infrastructure at the end?
With GrowthPal, you do. The domains, the inboxes and the campaign data are yours when the engagement ends, which is set out on the FAQ. That matters because the infrastructure is most of what takes two to four weeks to build, and rebuilding it is the expensive part of switching.
Can we use an agency and hire later?
Yes, and that sequence usually works better than the reverse. The engagement finds the message and builds the infrastructure; the hire then works against a message that is already proven rather than being asked to discover it while also learning your product.
What if our market is smaller than 30,000 contacts?
Then outbound at volume is probably the wrong tool, and GrowthPal will say so. Below roughly 30,000 reachable contacts there is not enough room to test 10 to 15 angles and reach a reliable answer. Precision account-based work, done by a person, tends to beat volume at that size.
Talk it through on a 30-minute call →
Or start with the free GTM plan, which maps your market and writes five messaging angles with no call required.