There are three ways to get B2B leads, and they cost money in three different shapes. Most teams pick one without doing the arithmetic on the other two.
This is the arithmetic. Plug in your own numbers as you go.
The three models
In-house SDR. You hire someone. They research, build lists, write outreach, and follow up. You pay salary plus tooling plus management time.
Agency retainer. You pay a monthly fee for a certain amount of attention. What that attention produces varies month to month.
Per-lead. You pay a fixed amount for each qualified lead delivered. Volume scales with what you buy, not with who you hire.
Each one has a different failure mode, and the failure mode matters more than the headline price.
What in-house actually costs
Start with the salary, then add what's underneath it.
A B2B SDR in a US or UK market costs a salary plus payroll taxes and benefits, typically another 20-30% on top. Then tooling: a data provider, an email verification service, a sequencer, a CRM seat. Then the manager's time, which is real even though it never appears as a line item.
Now the part most teams skip: ramp. A new SDR is not productive on day one. Depending on how complex your product is, expect one to three months before output is steady. That ramp period is a full cost with a fraction of the output.
Then divide by how many qualified leads they actually produce in a month. Be honest here, and use a number you can verify from your CRM rather than the number from the interview.
The result is usually higher than people expect, because the salary is the smallest part of the calculation once ramp and tooling are included.
The failure mode: the person leaves, and you start the ramp over. Everything they learned about your ICP goes with them.
What agency retainers actually cost
The invoice is simple. The economics are not.
A retainer buys hours of attention, not units of output. In a good month the agency ships a lot. In a slow month you pay the same and get a report explaining the slow month.
Some agencies price per qualified meeting, which is closer to output pricing and worth looking for. Most price per month.
Two questions worth asking before signing a retainer:
1. What happens to the fee in a month with half the usual output? 2. Who owns the data and the sequences if we stop working together?
The answers tell you whether you're buying a service or renting a headcount.
The failure mode: incentives point in different directions. The agency's incentive is to keep the retainer. Yours is to get leads. Those overlap most of the time, but not all of it.
What per-lead pricing actually costs
Per-lead is the easiest model to compare because there's one number. It's also the model where the definition of "lead" matters most.
A lead can mean a name and a company. It can mean a verified email address. It can mean a person who fits a defined ICP, with a written reason for the fit and a deliverable contact. These are not the same product and they don't cost the same.
Before comparing per-lead prices between vendors, get all of them to answer the same four questions:
- Is the contact verified as deliverable, or just found? - What's the qualification criteria, and who sets it? - Can I reject a lead that doesn't meet the criteria, and do I get the credit back? - Where did the data come from?
A cheap lead that bounces costs more than an expensive one that converts. The price per unit only means something once the unit is defined.
The failure mode: loose definitions. If "lead" isn't specified tightly, per-lead pricing becomes per-name pricing, and you're paying for a spreadsheet.
The costs nobody puts on the invoice
Three of them, and they apply across all three models.
Bounce cost. Every undeliverable email hurts your sending domain's reputation, which affects deliverability for the messages that would have landed. The cost isn't the wasted send. It's the sends after it.
Rework cost. Time spent cleaning, deduplicating, or re-researching a list you already paid for.
Opportunity cost. The deals your team didn't work because they were building lists instead.
None of these show up in a cost-per-lead comparison spreadsheet, and all three are real.
How to compare properly
Take one number to the comparison: cost per qualified lead that a rep actually contacts.
Not cost per name. Not cost per hour. Not monthly spend. Cost per lead that survives verification, matches your ICP, and gets a real conversation attempted.
For in-house, that's total loaded cost divided by monthly output. For a retainer, it's the monthly fee divided by qualified leads delivered that month. For per-lead, it's the price, adjusted for whatever percentage you reject.
Run that number for all three. The answer is often different from what the headline pricing suggests, and it's the only number that compares like with like.
Where Agenclaw sits
We price per verified lead delivered. If a lead doesn't meet the criteria you set, you reject it and the credit goes back. You're not buying hours, and you're not paying for the ones that get spiked at the quality gate.
Current pricing is on the pricing page, including how it compares to typical market rates. No card is required to start, and there's no contract to exit.
If you want to see the unit before you buy it, start a run and look at what comes out.