Build or Buy: What Outbound Actually Costs

Outbound costs more than the tools and less than a headcount, and almost nobody budgets for the part in the middle. The infrastructure is cheap and predictable. The data is more expensive than expected. The labor is the whole decision, and it is the line item most build-or-buy math leaves out entirely. This is the full breakdown, and the framework for deciding who should run it.

Why Does This Question Come Up at All?

For most companies between $500K and $3M ARR, the same person is responsible for both prospecting and closing. That arrangement contains a structural contradiction. When deals heat up, prospecting stops. Sixty to ninety days later the pipeline is empty, so prospecting restarts, and by the time it produces anything the closing work has dried up. The cycle repeats.

This is not a discipline problem. No amount of willpower fixes a role that requires two mutually exclusive modes of work to happen at the same time. Prospecting is repetitive, low-status, and pays off in 60 days. Closing is urgent, high-status, and pays off this week. A human being with both jobs will always choose the second one, and should.

The fix is structural too: someone prospects every day regardless of what else is happening. How you staff that function is the subject of the rest of this page. Whether it exists at all matters more than which option you pick.

What Does the Infrastructure Actually Cost?

This is the floor, before a single email is written. Domains, mailboxes, a sending platform, warmup, and verification. It scales close to linearly with volume.

Notice how small these numbers are. Infrastructure is not why outbound is expensive. It is, however, the line most often cut, and cutting it is what produces the deliverability failures that make everything downstream worthless.

Why Is Data the Most Underestimated Expense?

Good contact data costs $0.10-0.75 per contact. Cheap data costs $0.01-0.05 and carries 30-40% stale rates. That gap does not just waste sends. Stale contacts bounce, bounces damage sender reputation, and a damaged domain costs you the warmup period plus the pipeline you would have generated during it. Cheap data is the most expensive line item on this page once you price in what it breaks.

What Are the Four Ways to Staff It?

Every company solves this one of four ways. The costs are not close, and neither are the ramp times.

What Does DIY Actually Require?

The tool bill is the part people quote. The labor is the part that decides whether it works. Running outbound well is not one job, it is a dozen small technical disciplines: DNS and domain configuration, warmup sequencing, list building, enrichment, copy iteration, reply handling, bounce management, and deliverability monitoring. Each is learnable. Together they consume most of a working week.

Does Hiring a VA to Run It Work?

Rarely. The assumption behind it is that outbound is repetitive work that can be delegated cheaply. Some of it is. The parts that determine whether it produces meetings are not: judging which accounts are worth targeting, writing copy that sounds like a person who understands the buyer, diagnosing why a campaign stalled. Those are judgment tasks. Handing them to someone hired for throughput produces high volume and near-zero replies, and the domain damage from unverified sending outlasts the experiment.

When Does Each Option Make Sense?

When Should You Hire an In-House SDR?

When you already have a playbook that works. An SDR is an execution hire. Given a proven ICP, a message that gets replies, and working infrastructure, a good SDR ramps in about six weeks and scales what exists. Given none of those, the same person spends six months inventing them, usually badly, because inventing a playbook is a different skill from running one.

What Does the Decision Look Like by Stage?

What Unit Economics Actually Matter?

Monthly spend tells you nothing on its own. Cost per meeting is the number that lets you compare options that look nothing alike. Here is how reply rate moves it, holding send volume constant.

Those are marginal costs on sends alone. Loaded with labor and data, real cost per meeting lands far higher. The point of the table is the slope: a reply rate moving from 1% to 3% cuts cost per meeting by two thirds without changing a line of the budget. Targeting and copy move that number. Spending more rarely does.

Where Do Companies Waste the Most Money?

• Cheap data. The savings are real and the downstream cost of bounces and burned domains is larger.
• Skipping verification to save $50-150/month, then spending weeks recovering a domain.
• Sending before warmup completes, which converts an infrastructure investment into a spam-folder problem.
• Buying volume before fixing reply rate. Ten thousand sends at 0.5% is worse than two thousand at 3%, and costs five times more.
• Hiring for execution before a playbook exists.
• Measuring open rates. Between 30% and 50% of reported opens are artificial, so the metric mostly measures Apple Mail and security scanners.

What Sequence Should Most Companies Follow?

1. 1Validate that the channel works for your ICP at all, at the smallest spend that produces a real signal. Ninety days is the minimum honest test.
2. 2Buy the playbook rather than invent it, if nobody in-house has built one before. What you are paying for is the ICP definition, the message that gets replies, and working infrastructure.
3. 3Measure cost per meeting, not spend or activity. Hold it against a 5:1 pipeline ratio.
4. 4Hire in-house once the playbook is proven and the volume justifies a dedicated person. Ramp drops from months to weeks because they inherit something that works.
5. 5Keep the infrastructure and the systems in your own accounts throughout, so that step four is a transition rather than a rebuild.