What does B2B lead generation actually cost?
Five ways agencies charge for the same work. They look similar on a quote and behave nothing alike when a campaign underdelivers. This page prices one outcome - a qualified opportunity - under every model, using a funnel we measured rather than a range we invented, and it will tell you when the answer is not to hire anyone.
Price the same outcome five ways
Set the outcome you want, then read what each commercial model charges to deliver it. You supply the qualified-meeting volume and the sending volume; this page does not derive one from the other. Industry-specific write-ups are available for manufacturing companies and recruiting and staffing agencies.
Cost per qualified opportunity
Everything below updates as you move the controls. With JavaScript unavailable, the figures shown are for 5 opportunities a month at a $25,000 contract value and a 20 percent close rate.
| Model | Cost this month | Per opportunity | If it delivers nothing | Who carries the risk |
|---|---|---|---|---|
| Monthly retainer | $5,000 | $1,000 | $5,000 | You |
| Pay per appointment | $1,665 | $333 | $0 | Agency |
| Infrastructure plus revenue share | $5,773 | $1,155 | $2,023 | Shared |
| In-house SDR plus tooling | $8,690 | $1,738 | $8,690 | You |
Adjust the quotes you have actually been given
Why these defaults and not a price list. Only one figure here is a published market price. The others are our own arithmetic, or the number you were quoted - and the tool is more useful when you enter the real one. We are not going to invent a market survey we have not run.
The arithmetic, in full. Volume =
opportunities x 4,100 x (2.12 / your reply
rate) = 20,500 emails. That
volume sits in our published $2,023
infrastructure tier. Revenue = opportunities x close rate x contract value =
$25,000. Revenue share =
infrastructure + share x revenue = $5,773. In-house =
$80,000 / 12 + infrastructure =
$8,690. The tier table is
itemised on what a cold email agency costs, and the
measured constants are on cold email benchmarks.
What this does not model. Sales cycle. Every figure here is a monthly cost set against first-year revenue that may take two quarters to arrive. Revenue share looks cheap on a long cycle precisely because nobody is paid until it closes - that is the model working as intended, not an arbitrage. It also ignores your own time, which is the largest hidden cost in the in-house row.
Can your market even supply it?
This is the question nobody selling lead generation asks first, and it outranks price. Our 12-month programme contacted 540,000 leads to produce 263 qualified sales meetings. The coverage check below asks how much of your reachable market your own contact volume consumes in a year, using figures you supply rather than a conversion rate carried over from another company. If the number below is larger than the universe of companies you can credibly sell to, the constraint is arithmetic, not effort.
Market feasibility check
Reachable companies means the ones that plausibly have the problem you solve and are the size you sell to - not a market-size figure in dollars.
All four verdicts are listed below. With JavaScript enabled, only the one matching your inputs is shown.
Workable
Your reachable market is large enough to be contacted once a year and still hit this target. Volume is not your constraint; targeting and copy are.
Tight but possible
You would have to contact every company in your reachable market about twice a year. That is survivable, but it means your list quality and your copy have to be right the first time - there is no room to burn the market on a weak test.
Strained
Hitting this target means contacting your entire reachable market three or four times a year. Reply rates fall on repeat contact, so the later touches will not perform like the measured median. Expect to need more volume than this figure, not less.
Not reachable by outbound
This target requires contacting your entire reachable market more than four times a year. No pricing model fixes that. Either the opportunity target is too high for the market, the definition of the market is too narrow, or the channel is wrong. An agency that accepts this brief without saying so is selling you volume it cannot source.
Why the ratio is conservative. The 2,053 figure comes from one programme in one market over twelve months, and reply rates fall on repeat contact. A market you have to touch three times a year will not perform on the third pass the way this ratio assumes, so treat a strained verdict as optimistic rather than borderline.
The five models side by side
Pay-per-lead appears here but not in the calculator above, deliberately. A lead is not an opportunity, and no honest arithmetic converts one into the other without the vendor's own qualification definition - which is the exact thing that varies, and the exact thing you should get in writing.
| Model | What you pay | Who carries delivery risk | If it delivers nothing | Right for you when |
|---|---|---|---|---|
| Monthly retainer | A fixed fee every month | You carry all of it | The full fee | You want control of targeting and messaging and you have someone in-house to direct it. Not for you if: You are buying because you do not have that person. A retainer buys effort, not outcomes. revenue share vs retainer |
| Pay per lead | A fee per delivered contact | Agency carries volume risk, you carry quality risk | Nothing | The definition of a qualifying lead is written down, narrow, and enforced by you. Not for you if: The definition is loose. A lead is whatever the contract says it is, and the incentive is volume. This is the one model where getting exactly what you paid for can still be worthless. how pay per lead works |
| Pay per appointment | A fee per meeting that happens | Agency, until the meeting is booked | Nothing for meetings, but the infrastructure underneath is usually still billed | You can close, your calendar is the bottleneck, and show-rate is defined in writing. Not for you if: Nobody has defined what counts as qualified, or who eats a no-show. pay per appointment vs retainer |
| Infrastructure plus revenue share | A published monthly amount plus an agreed share of revenue that actually closes | Shared. The monthly amount is owed either way; the share is not | The infrastructure amount only | Your ACV is high enough that a share is worth more to the agency than a retainer, and you are willing to report closed revenue honestly. Not for you if: You will not share revenue data, or your sales cycle is so long that nobody can be paid for a year. This is our model, and it is not right for everyone. lead generation with no retainer |
| In-house SDR | Salary, tooling and management time | You carry all of it, plus the hiring risk | The full cost, and the ramp starts again with the next hire | You have enough volume to keep a person busy and a manager who has done the job before. Not for you if: This is your first outbound motion. You are paying to learn on your own payroll. agency vs in-house SDR |
When not to hire a lead generation agency
We sell this service. These are the cases where we think you should not buy it from us or from anybody else, and each one is checkable against your own numbers rather than a feeling.
Your ACV is under about $5,000 and you sell one seat at a time
At the measured constant of 263 qualified sales meetings over 12 months, the cheapest published infrastructure tier - $2,023 a month at 25,000 emails - buys roughly six qualified meetings a month. If closing all six does not clear the infrastructure cost, the machine cannot pay for itself before anyone's margin is added.
Your reachable market is under a few thousand companies
One qualified meeting took 2,053 contacted leads on our 12-month programme. Run your own number in the feasibility check above. If it says the market has to be contacted more than about twice a year, outbound is not the constraint you should be spending on.
You cannot describe a problem your buyer already knows they have
Cold email works when it names a problem the reader recognises in the first two lines. If the value only makes sense after a demo, the channel is wrong and no agency can fix that with better subject lines.
You need pipeline this month
Domains have to warm before they send at volume. Anyone who promises meaningful volume in week one is either sending from your primary domain or from infrastructure somebody else has already burned.
Nobody on your side will answer replies within a day
Every model on this page is priced per opportunity. An opportunity that sits unanswered for a week is not an opportunity, and you will still be billed for it under three of the five models.
If none of these describes you, the question is which model, not whether. If two or more do, the honest answer is that outbound is not your constraint yet.
How we measured this
Two separate datasets sit underneath this page, with different denominators. They are not mixed, and the distinction matters enough to state plainly.
The funnel constants
One portal-tracked client programme, 12 months. 1,080,000 emails to 540,000 leads produced 5,028 replies and 263 qualified sales meetings. That gives 263 qualified sales meetings over 12 months, across that campaign and in that 12-month campaign. Limitation: this is one programme in one market. It is a real measured funnel, not a benchmark, and your own ratios will differ.
The reply-rate distribution
115 campaigns across 7 client programmes, 429,763 emails and 242,669 leads contacted, 6,249 unique replies, counters pulled 12 August 2026. Of those, the 81 campaigns that contacted 500 or more leads replied at between 0.36 and 8.00 percent, with quartiles at 1.38, 2.12 and 2.97 percent. Denominator is unique replies over leads contacted, per campaign. Full workings are on cold email benchmarks.
How the two are combined, and how they are not
The programme above ran close to the median campaign reply rate, so the performance control
scales volume by median / your rate - a relative multiplier only. We never
substitute one dataset's reply rate for the other's, because the first is measured on leads
over twelve months and the second on leads contacted per campaign. Treating those as the same
number would overstate the required volume by roughly a factor of three.
What is not first-party
The $80,000 SDR figure is on-target earnings from the Bridge Group's 2025 SDR metrics report, not our payroll. The $5,000 retainer default is the starting figure Belkins publishes on its own appointment-setting page, read 12 August 2026. Everything else on this page is ours.
Our full standards are on methodology and editorial standards.
How much does lead generation cost?
The honest answer is a shape, not a number: lead generation pricing follows the five models on this page, and each converts to a different cost per outcome. Retainers price capacity, so lead generation cost per meeting falls as volume rises and punishes quiet months. Pay-per-lead prices the unit, so your cost per lead is fixed and your cost per qualified opportunity depends entirely on the definition of a lead. Per-appointment prices meetings; revenue share prices outcomes. Comparing two quotes therefore means converting both to the same unit - cost per held, qualified meeting - which the calculator does, with the infrastructure line included via the cost breakdown. Average cost-per-lead figures quoted around the industry are near-useless without that conversion: a cheap lead that never becomes a meeting is the most expensive thing on this page.
Go deeper on one model
Each of these takes a single question from this page and answers it properly.
If the feasibility check said your market can supply it and you want the machine run for you, we work on published infrastructure tiers plus an agreed share of what closes. Tell us the numbers you entered above and we will tell you whether we think it works.
Apply to work with us