B2B Lead Generation Companies: How to Compare Them on Pricing and Risk

Sopro's pricing page explains why it does not list tiers: "We offer a tailored service. Your business is unique. Your audience is unique." It then gives a number anyway, "around £3K per month", and adds "no minimum contract".

That combination, an honest floor plus an honest reason for not publishing more, tells you more about a vendor than any capabilities deck. Most of this category will not give you either. Here is how to run the selection so you find out what you need to know before signing rather than in month three.

Sopro's published pricing page, captured August 2026.

Shortlist by shape, not by capability list

Every vendor in this category describes the same activities: research, list building, copy, sending, qualification, reporting. Comparing those descriptions is close to useless because they converge.

What actually differs is commercial shape. Some publish real numbers: CIENCE lists a $5,000 one-time setup plus $2,000 a month for the team and $499 a month for the platform, $7,499 for the first month all in. SalesRoads lists engagements from $9,950 per four weeks and $16,750 for two SDRs. Others quote everything.

Sort your shortlist into retainer, per-seat and per-outcome before you take a single call, because those three are not comparable to each other and a call comparing them will not make them so. The models and their economics are covered in B2B lead generation services.

The six diligence questions

These are ordered by how often the answer changes a decision.

1. Who actually does the work, and where do they sit? Some firms deliver with in-house teams; many subcontract research, copy or sending. Subcontracting is not disqualifying, and not knowing about it is. Ask directly which parts are theirs.

2. Whose sending infrastructure is used, and who else is on it? If your campaigns share domains and IPs with other clients, another client's bad list can damage your deliverability. Ask whether infrastructure is dedicated, and if it is shared, what happens when a neighbour causes a problem.

3. What is the ramp to first send, week by week? Domain warmup alone takes weeks if new domains are involved. A vendor promising sends in week one is either using aged infrastructure, which is fine and worth confirming, or is about to burn something.

4. What does the data cost and who owns it? Whether contact data is a pass-through, included, or your own contract determines the vendor's incentive on data quality, and it determines whether you keep the list when you leave.

5. What is the actual definition of the thing you are buying? For a lead, is it a verified contact or a person who replied. For a meeting, what makes it qualified. Get it in writing before pricing, not after.

6. What happens when it does not work? Not the guarantee language, the process. Who reviews it, on what evidence, at what point, and what changes. A vendor with no answer beyond "we optimise" has no diagnostic process.

1. Step 1Sort by commercial shape

   Retainer, per seat or per outcome. Only compare within a group.

2. Step 2Run the six questions on a first call

   Subcontracting, infrastructure, ramp, data, definitions, failure process.

3. Step 3Ask for account-level reporting from a live client

   Names removed. A document, not a dashboard tour.

4. Step 4Run a paid pilot with a written success condition

   Small, time-boxed, and defined before it starts rather than judged afterwards.

5. Step 5Check the exit terms before you sign

   Notice period, data ownership, and what you keep.

Designing a pilot that tells you something

A pilot exists to reduce uncertainty, and most are designed so they cannot.

Make it long enough to include a full cycle. Outbound has a lag: build, warm, send, wait for replies, book, hold. A four-week pilot on a market with a three-week reply-to-meeting lag measures the build and nothing else. Six to eight weeks is usually the minimum that includes held meetings.

Define the success condition in writing before it starts. Not "see how it goes". A number of held meetings meeting the agreed definition, or a reply rate threshold, decided in advance. A pilot judged afterwards is judged by whoever argues best.

Keep the variables down. One ICP, one offer, one channel. A pilot testing three segments produces three underpowered samples and no conclusion.

Agree what happens to the assets. Who owns the list, the copy and the domains when the pilot ends is a cheap question before and an expensive one after.

Reading the reference call

Reference calls are usually theatre because the vendor picks the reference. Two questions make them useful anyway.

Ask the reference what the first two months were like, specifically. Every engagement has a rough start; a reference who cannot describe theirs is either unusually lucky or not being candid.

Then ask what the vendor does badly. A reference who says nothing has told you the call is scripted. A reference who says something small and specific, and then explains how it was handled, has told you what working with the vendor is genuinely like.

Vendor selection signals

• ✓Yes: They publish a starting price, or state one on the first call
• ✓Yes: They can describe their diagnostic process when results are flat
• ✓Yes: They ask about your suppression list before you raise it
• ✓Yes: They tell you which parts they subcontract without being pushed
• ✕No: The proposal guarantees meetings without defining a qualified meeting
• ✕No: Reporting is a dashboard tour rather than a document they send
• ✕No: They recommend follow-up sequences as the main lever

The last row is our own bias made explicit, and it is worth stating for that reason. We run one message per campaign and no bumps, so a vendor whose plan for a flat campaign is more follow-ups is proposing something we would not do. Judge that as our position rather than as a neutral standard, but ask what else they would change, because "send more" is a volume answer to a relevance problem.

The contract terms that matter

Three clauses decide how much a bad decision costs.

Notice period. Monthly rolling versus a twelve-month commitment is the difference between a cheap mistake and an expensive one. Sopro advertises no minimum contract, and CIENCE lists its recurring fees as month to month, so short commitments do exist in this category.

Data ownership. Whether you keep the enriched list, the suppression list and the campaign history on exit. This is frequently silent in the contract, which means it is negotiable.

Domain and inbox ownership. If the vendor bought sending domains for you, whether they transfer matters, because the warmup history has real value.

Red flags in a proposal

Five patterns that predict trouble, in rough order of how reliably they do.

A guaranteed number of meetings with no definition of a meeting. The guarantee is doing marketing work rather than commercial work. Without criteria, the vendor decides what counts, and they decide it while being paid per unit.

A projection with no assumptions attached. Any meetings forecast rests on assumed list size, reply rate and booking rate. A proposal showing the output without the inputs cannot be evaluated, and cannot be held to anything later.

No questions about your existing relationships. A vendor who does not ask for a suppression list before proposing has not thought about the most obvious way to damage a client, which is contacting their current customers.

Volume as the answer to every problem. More sends, more sequences, more channels. Sometimes correct, and as a first instinct it indicates a vendor without a diagnostic process.

Reluctance to name the team. Who specifically writes the copy and manages the campaign, and how many other accounts they carry. Vague answers usually mean a shared pool at a ratio you would not accept if stated.

What you have to do yourself regardless

Three inputs no vendor can supply, and the engagement fails without them.

The suppression list. Current customers, live opportunities, partners, competitors, and anyone a rep is protecting. Only you have this, it must exist before launch, and it is the single cheapest way to prevent the most damaging kind of mistake.

The offer. What you are actually asking a prospect to do and what they get. A vendor can sharpen the wording; they cannot decide your commercial proposition, and an engagement that stalls on this stalls entirely.

Fast reply handling. Whether the vendor books meetings or hands you replies, someone on your side has to answer quickly. A reply answered four days later is frequently a meeting lost, and no amount of vendor performance compensates.

Agree who owns each of these in writing at kickoff. The most common cause of a disappointing engagement is not vendor incompetence; it is one of these three sitting unowned while both sides assume the other has it.

Where an agency genuinely beats in house

Worth being clear about, because the comparison rarely gets made properly.

A vendor has run this motion across many markets, holds infrastructure already warm, and has production capacity idle between clients. That is genuine leverage on execution speed and on avoiding the specific mistakes that cost weeks.

Where they have least leverage is knowing which companies are worth contacting and what will persuade them, because that depends on your market and your won deals. An agency can structure and challenge that knowledge; it cannot originate it. Our guides to defining an ideal customer profile and lead generation agency cost cover both halves.

The short version

Sort the shortlist by commercial shape before comparing anything, then use the six questions to establish who does the work, whose infrastructure it runs on, and what the thing you are buying is actually defined as. Run a paid pilot long enough to include held meetings, with a written success condition set in advance. Check notice period and data ownership before signing, because both are usually negotiable and neither is negotiable afterwards.

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

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

Sources: Sopro pricing, CIENCE pricing, SalesRoads pricing