B2B Lead Generation Services: The Five Delivery Models and What Each Costs

CIENCE publishes a first month at $7,499, made up of a $5,000 one-time GTM setup, a $2,000 a month strategic team and a $499 a month platform licence, with SDR capacity quoted separately. SalesRoads publishes engagements starting at $9,950 for four weeks, and $16,750 for a two-SDR engagement. Sopro says pricing "starts at around £3K per month" and explains on the page why it lists no tiers.

Three vendors, three genuinely different commercial shapes. The category is usually described by what firms do, which is roughly the same everywhere. The useful distinction is what the fee is attached to, because that determines who carries the risk when the programme underperforms.

The five models

RetainerFee per month

• You pay for activity and capacity
• Vendor is paid whether or not it works
• Easiest to compare between vendors
• The default across the category

Per seat or per SDRFee per head

• Priced by dedicated headcount
• Scales in lumpy, expensive increments
• Cost is legible; output still is not guaranteed
• Common in outsourced SDR

Per outcomeFee per meeting or lead

• Vendor carries delivery risk
• Requires a written definition of the outcome
• Aligns incentives on volume, not always on quality
• Rare, and worth understanding why

Those three plus two more cover almost every proposal you will receive.

1. Retainer agency. A monthly fee for a defined scope: list building, copy, sending infrastructure, campaign management, reporting. Sopro's model, and Belkins', where the retainer includes the tooling. The fee buys capacity and expertise, and it is indifferent to results within any given month.

2. Outsourced SDR. You rent people. SalesRoads publishes this shape clearly: a dedicated SDR plus a sales operations team, a director of client success and a talent development manager, from $9,950 per four weeks, with two SDRs at $16,750 and a stated cost per rep of $8,375.

3. Setup plus managed growth. A one-time build followed by a smaller recurring fee, which is CIENCE's published structure: $5,000 setup, then $2,000 a month for the team and $499 a month for the platform, with SDR capacity optional and quoted.

4. Pay per lead. A price per contact meeting agreed criteria. Cheap-looking, and the model where definitional disputes are most common, because a "lead" can mean a verified contact record or a person who expressed interest, and those differ by an order of magnitude in value.

5. Pay per qualified meeting. A price per attended meeting meeting a written definition. This is the model we run, and the definition is the whole product.

CIENCE's published pricing structure, captured August 2026.

What each model actually costs

Published anchors, all from vendors' own pages in August 2026.

Two observations from assembling that table.

The published end of the market clusters between roughly $2,500 and $10,000 a month. Below that you are usually buying a tool with light service attached; above it you are usually buying dedicated headcount.

The outcome-based models are the ones nobody publishes, which is not evasion so much as arithmetic. A price per meeting depends on your ICP, deal size and how hard your buyers are to reach, and a vendor quoting it blind is either guessing or protecting themselves with a definition you will not like.

SalesRoads' published engagement pricing, captured August 2026.

Who carries the risk in each

This is the question the fee structure answers, and it is the one worth deciding before you shortlist.

On a retainer, you carry it. If the campaign underperforms, you have paid for the months regardless, and your recourse is to leave. That is not unreasonable: the vendor genuinely incurs the cost of building and running the programme whether or not your market responds.

On per-seat, you carry it too, in a more legible form. You know exactly what a rep costs and you still do not know what they will produce.

On per-outcome, the vendor carries it, and the consequence is that the definition of the outcome becomes the contract. A vendor paid per meeting has an incentive to book meetings, so the qualification criteria are doing all the work of keeping those meetings worth attending.

That is why we publish ours rather than negotiating it per deal. A meeting is billable when the company is in the pre-approved audience, the participant has reasonable responsibility for or influence over the relevant area, they agree to a relevant business conversation, they attend and participate, and they were not disclosed as an existing customer or active opportunity before outreach. Budget, timing and decision authority are deliberately not conditions, because they are not knowable before the conversation and making them billing conditions turns every invoice into a dispute.

Outcome definition

• ✓Yes: The qualifying criteria are written down and agreed by both sides
• ✓Yes: There is a rejection window with a stated number of business days
• ✓Yes: Valid rejection reasons are enumerated
• ✓Yes: Suppression covers customers, live deals and partners before launch
• ✕No: Budget, timing or authority appear as billing conditions
• ✕No: Subjective call quality can void a meeting that met the criteria
• ~Depends: No-shows and reschedules, which need their own explicit rule

Choosing between the models

The honest selector is not budget. It is what you are actually short of.

Short of expertise and infrastructure points at a retainer or a setup-plus-managed deal. You are buying a working system and the knowledge of how to run it, and the monthly fee is the price of not building it.

Short of hours points at outsourced SDR. You know what to do and there is nobody to do it, so renting a trained person with management around them is the direct fix.

Short of certainty points at per-outcome. If you cannot justify spend against an unknown return, moving the risk to the vendor is exactly what the model is for, and you pay a premium per unit for it.

Short of nothing except volume points at doing it in house, which is a real option that the category rarely mentions. Our guide to building an outbound engine from scratch covers what that involves, and the lead generation agency cost guide covers the comparison in more depth.

How each model behaves in a bad month

Every engagement has one. What the models do differently is what happens next, and that is worth thinking through before you pick one.

On a retainer, a bad month is a conversation. You raise it, the vendor proposes changes, and you both continue paying and working. The risk is drift: three bad months can pass as optimisation because nothing forces a decision.

On per-seat, a bad month is a utilisation question. The rep exists and costs the same, so the discussion becomes whether they are working the right list rather than whether the programme works. That is sometimes the right question and sometimes a distraction from a targeting problem.

On per-outcome, a bad month costs the vendor. That produces fast reaction, which is the model's real advantage. It also produces pressure on the edges of the definition, which is why the criteria and the rejection window have to be settled before the first invoice rather than during the first dispute.

The practical consequence: on a retainer, set a review point with a written success condition in advance, because the model will not force one. On per-outcome, spend the diligence effort on the definition rather than on the price.

Hybrids, and the one that usually works

Most real contracts are hybrids, and the common shapes are worth naming.

Setup fee plus performance. A build fee covering infrastructure and list work, then a per-meeting rate. This is usually the fairest structure available, because the vendor's genuine upfront costs get covered while the ongoing risk stays with them.

Reduced retainer plus per-meeting bonus. Splits the risk. Workable, and the bonus has to be large enough to matter or it changes nothing.

Retainer with a performance floor. A minimum number of meetings or the next month is credited. Simple, and only as good as the definition of the thing being counted, which puts you back in the same place.

The shape to avoid is a full retainer with a guarantee attached that has no teeth, because it reads as risk-sharing while being a retainer with marketing on top. Ask what specifically happens if the floor is missed, and whether the remedy is credit against future months, which only has value if you intend to stay.

What the fee usually excludes

Three costs sit outside most quotes and appear later.

Sending infrastructure. Domains, inboxes and warmup. Some retainers include tooling, and Belkins advertises the tools as part of the retainer, but many do not, and the line item is real.

Data. Contact data and verification, either as a pass-through or as your own contract. Ask which, because a vendor absorbing data cost has an incentive to use cheaper data.

Your own time. Copy approval, meeting attendance, feedback on quality. A programme where the client never responds underperforms regardless of the model, and it is the most commonly underestimated input. The vendor-selection side of this, including the six diligence questions worth asking before any of it, is covered in B2B lead generation companies.

The short version

Compare vendors by what the fee attaches to rather than by service description, since the descriptions are nearly identical and the risk allocation is not. Published entry points cluster between roughly £3,000 and $10,000 a month, dedicated SDR capacity runs materially higher, and outcome-based pricing is rarely published because it depends on your market. If you go outcome-based, the written definition of the outcome is the product, so settle it before the first invoice rather than after.

If a per-qualified-meeting arrangement with the definition agreed up front is the shape you want, you can see what a campaign would look like for your market.

Vendor pricing verified against each company's own pricing page as of August 2026. Verify current terms with the vendor before relying on them.

Sources: CIENCE pricing, SalesRoads pricing, Sopro pricing