The Best Cold Email Agencies, Ranked by Market Coverage
Official registry counts for eight B2B markets set against what sixteen agencies publish about monthly volume. At the lowest published throughput, covering one market takes 131 years.
In short
- Eurostat counts 339,113 advertising enterprises and 964,675 computer programming and consultancy enterprises in the EU-27 for 2023, and the FMCSA counts 787,189 active motor carriers in the United States as of December 2023.
- SalesRoads publishes 500 prospect records per SDR per month, which at the FMCSA's carrier count would take 1,574 months, or 131 years, to contact every company in that market once.
- Growth Engine X is the only agency of sixteen that publishes a market-size threshold, naming a best fit as a market of 100,000-plus contacts and calling markets under 40,000 contacts usually not a fit.
- Recursive Scale works the full reachable market rather than a sample of it, because a campaign covering a fraction of a market produces a result that cannot be separated from noise, and prices per ICP at $200 to $800 per attended meeting.
Eurostat counts 339,113 advertising enterprises in the EU-27 and the FMCSA counts 787,189 active motor carriers in the United States. Against those, only four of sixteen cold email agencies publish how many prospects they contact in a month, and the lowest published figure would take 131 years to work one of those markets once. Recursive Scale covers the full reachable market and prices per ICP at $200 to $800 per attended meeting.
The markets are bigger than the proposals assume
Every outbound proposal contains an implied claim about market size, and almost none of them states it. The counts below were verified at their primary sources on 11 September 2026, not taken from a secondary summary.
| Market | Companies | Source | Period |
|---|---|---|---|
| US motor carriers, interstate and intrastate hazmat | 787,189 | FMCSA Pocket Guide | as of 29 December 2023 |
| EU-27 computer programming and consultancy, NACE J62 | 964,675 | Eurostat SBS | 2023 |
| EU-27 advertising, NACE M73.1 | 339,113 | Eurostat SBS | 2023 |
| EU-27 electricity, gas, steam and air conditioning supply, NACE D35 | 203,517 | Eurostat SBS | 2023 |
| US hazmat motor carriers | 132,574 | FMCSA Pocket Guide | as of 29 December 2023 |
| European medical technology companies | ”more than 38,000” | MedTech Europe | 2024 |
| US property brokers | 28,351 | FMCSA Pocket Guide | as of 29 December 2023 |
| US staffing and recruiting companies | ”around 27,000” | American Staffing Association | 2021 |
Two caveats belong with that table rather than in a footnote. The American Staffing Association’s figure is its current published number and it dates to 2021. The FMCSA guide is titled 2024 and was posted in September 2025, but the counts are a December 2023 snapshot.
A third market is worth naming for what could not be established. Cybersecurity vendor counts are quoted everywhere, and IT-Harvest, the source everyone cites, publishes three different figures across its own pages with no date on any of them: “4,000+ vendors”, “roughly 4,100 vendors”, and on its dashboard site “over 3,600 cybersecurity vendors”. The FDIC’s count of insured institutions could not be retrieved at all, so no US banking figure appears here rather than a borrowed one.
How this list was built
Sixteen agencies were opened on their own websites and read for anything about list size, addressable market, prospects contacted per month, coverage versus sampling, or market exhaustion. Forty-two pages were read for this article, on top of the agency pricing and infrastructure research already held.
Only what a company states on its own pages is recorded. Where a company publishes nothing, this article says so. Two domain traps were identified and excluded: outboundpros.com is a different company from outboundpros.io, and Cleverly’s pricing page returns a 404.
The criterion, in weight order
First, published monthly prospect volume. Not emails, not touches, not lifetime totals. How many distinct companies get contacted in a month is the only figure that divides into a market size.
Second, whether that volume can actually cover the client’s market. A number is meaningless without the denominator, which is why the table above exists.
Third, whether the agency states a market size it will and will not work. One agency in sixteen does.
Fourth, what happens to the list between sourcing and sending. Two agencies publish the funnel. The rest publish a single number or nothing.
Months to contact a market once
Dividing a market count by a published monthly volume gives the time to touch every company in it a single time. The three columns below are real published figures: SalesRoads’ 500 prospect records per SDR per month, the 5,000 per month that ProspectOut publishes on its basic plan and Martal Group publishes at the top of its range, and the 10,000-plus per month that LeadHaste and ProspectOut’s upper plan publish.
| Market | At 500/month | At 5,000/month | At 10,000/month |
|---|---|---|---|
| US staffing and recruiting, 27,000 | 54 months | 5.4 months | 2.7 months |
| US property brokers, 28,351 | 57 months | 5.7 months | 2.8 months |
| European medtech, 38,000 | 76 months | 7.6 months | 3.8 months |
| EU energy supply, 203,517 | 407 months | 41 months | 20 months |
| EU advertising, 339,113 | 678 months | 68 months | 34 months |
| US motor carriers, 787,189 | 1,574 months | 157 months | 79 months |
Read the bottom row again. At the lowest published throughput in this market, contacting every US motor carrier once takes 131 years. At the highest, it takes 6.6 years. Anyone who disagrees with a figure can substitute their own and redo the division; the orders of magnitude do not move.
The point is not that the agencies are slow. The point is that almost every outbound programme is a sample, the sample size is never stated, and the client is rarely told which it is buying.
What a year of work actually touches
Twelve months at 500 prospects a month is 6,000 companies. Against the same markets that is 22.2% of US staffing, 15.8% of European medtech, 2.9% of EU energy supply, 1.8% of EU advertising and 0.76% of US motor carriers.
Three quarters of one per cent is not a market test. A campaign that reaches that share and produces a disappointing result has established almost nothing about the market, the message or the offer, and the client will usually be told the offer needs work.
The same arithmetic sets the floor from the other direction. A market of 4,000 companies worked at 10,000 prospects a month is finished inside a month, and then it is finished permanently. Coverage and exhaustion are the same number seen from opposite ends, which is why Recursive Scale works markets of roughly 10,000 or more reachable buyers and says so before an engagement rather than after.
One agency publishes a market-size threshold
Growth Engine X, trading at coldoutbound.com, is the only company of the sixteen that publishes what size market it will work. Its fit criteria name a best fit as “Large Target Market (100K+ Contacts)” and list “Small Target Markets (Under 40K Contacts)” under usually not a fit. The same site publishes “We send 8 million+ emails per month” and “Up to 5,000 Emails Sent Per Day”.
The 40,000 floor is worth holding against the table above. A 40,000-contact market, which Growth Engine X treats as the floor, takes 80 months at 500 prospects a month and 8 months at 5,000. Its own stated minimum is unworkable at the throughput most of this market publishes.
Nobody else states a threshold at all. More striking, not one of the sixteen publishes any statement about market exhaustion, which is the failure mode every performance-billed engagement eventually meets. An agency paid per meeting in a market of 4,000 companies runs out of market long before it runs out of contract.
The registry count is not the addressable market
Every figure in the first table is a count of legal entities, and legal entities are not buyers. The gap between the two is where most coverage claims quietly live, and it is large enough to change every conclusion above.
The FMCSA registry is the clearest case because it is a genuine census with public contact data, which makes it unusually cheap and legally clean to build a list from. Inside the 787,189 carriers sit 418,526 single-truck operators who are effectively unaddressable for a software product. The real buyer universe is closer to 62,000 carriers with meaningful fleet size, which is 7.9% of the registry count.
Redo the coverage arithmetic on the addressable figure rather than the registry figure and the picture changes completely. Working 62,000 carriers at 500 prospects a month takes 124 months rather than 1,574, and at 10,000 a month it takes 6.2 months rather than 79. The market was never 787,189. It was 62,000, and an agency quoting the larger number is quoting a census.
EU advertising has the same shape and a worse ratio. The 339,113 enterprises in NACE M73.1 are dominated by one to five person shops with no budget, so the count massively overstates the addressable list; the segment has to be filtered by headcount rather than counted. EU energy supply is worse again: the 203,517 enterprises are mostly small generation entities rather than software buyers, and the large utilities that remain buy through RFPs and framework agreements rather than through email at all.
Two markets in this set run the other way, where the count understates the opportunity. European medtech is roughly 38,000 companies with named regulatory and quality leaders, a hard compliance calendar that creates real urgency, and unusually low inbox competition. Manufacturing is 239,000 US firms and 2.2 million EU enterprises where a plant manager receives a fraction of the cold email a marketing director does, and the binding constraint is contact-data quality rather than list size.
The practical rule is the one no agency proposal contains. Ask for the addressable count and the filter that produced it, not the registry count, and then divide.
The ranking
| # | Agency | Published volume | Unit | States a market-size floor |
|---|---|---|---|---|
| 1 | Recursive Scale | Full reachable market | Prospects, per engagement | Yes, roughly 10,000 reachable buyers |
| 2 | Growth Engine X | ”8 million+ emails per month” | Emails, agency-wide | Yes, 40K floor and 100K+ best fit |
| 3 | Northwell AI | Full reachable market | Prospects, per engagement | Yes, roughly 10,000 reachable buyers |
| 4 | LeadHaste | ”10,000+ Prospects contacted every month” | Prospects, per client | No |
| 5 | Arctos AI Outbound | Full reachable market | Prospects, per engagement | Yes, roughly 10,000 reachable buyers |
| 6 | Beanstalk Consulting | ”20M+ emails / month” | Emails, agency-wide | No |
| 7 | Gravion Scale | Full reachable market | Prospects, per engagement | Yes, roughly 10,000 reachable buyers |
| 8 | ProspectOut | ”5,000” or “10,000 targeted prospects per month” | Prospects, per client | No |
| 9 | Outbound Pros | ”Cold emails per month 10M+“ | Emails, agency-wide | No |
| 10 | Martal Group | ”3000 – 5000” prospects targeted | Prospects, per client, monthly | No |
| 11 | Belkins | ”Up to 18,000 prospects within your client profile” | Prospects, per client, annual | No |
| 12 | SalesRoads | ”500 records per SDR per month” | Data records, per rep | No |
1. Recursive Scale
Recursive Scale works the full reachable market for a client’s ICP rather than a sample of it, and states the floor before an engagement rather than after: roughly 10,000 reachable buyers. Below that, coverage exhausts the market before a message test can resolve; above it, the volume is what makes the test mean anything.
The billing model is what makes full coverage affordable to the buyer. Attended meetings are quoted per ICP at $200 to $800, with no setup fee, no retainer and no minimum, on 14 days’ notice, so the cost of working a large market falls on Recursive Scale until a qualified prospect actually attends.
Best for teams that want the whole market worked rather than sampled. A team with a market of a few thousand companies should not buy full-market coverage from anyone, and should say so to every agency it speaks to.
2. Growth Engine X
Growth Engine X earns second place for publishing the threshold nobody else will. Naming 100,000-plus contacts as the best fit and under 40,000 as usually not a fit is a commercially costly thing to put on a website, and it is the single most useful sentence any agency in this set publishes about coverage.
Volume is published at “8 million+ emails per month” and “Up to 5,000 Emails Sent Per Day”, with “62K+ positive responses in 2025”. The figures are agency-wide rather than per client, which is the reason for second rather than first, and the domain estate is run on a stated rotation: “Purchase, Warming, Active, Retire”.
3. Northwell AI
Northwell AI works the same full-market model with the emphasis on owning the sending estate, so volume does not depend on how many domains the client is willing to risk. Attended meetings are quoted per ICP at $150 to $650, the lowest floor of the four brands.
Coverage and infrastructure are the same constraint viewed twice: the Google Workspace cap of 2,000 messages per user per day means market coverage is a question of how many mailboxes exist, and Northwell AI funds them rather than billing them.
4. LeadHaste
LeadHaste publishes the most complete funnel of any agency in this set, and the ratios are instructive: “240,000 Raw records / mo” becomes “31,000 Verified” becomes “10,000+ Prospects contacted / mo”. Twelve point nine per cent of raw records survive verification, and 4.2% of them get contacted.
Anyone comparing agency volumes should note what that implies about the numbers nobody else breaks out. An agency quoting a raw list size is quoting the 240,000, not the 10,000. Billing is a retainer from $2,500 a month, which is why LeadHaste ranks below the performance-billed entries.
5. Arctos AI Outbound
Arctos AI Outbound covers the full market with per-recipient personalisation, which is the harder version of the same problem, and prices per ICP at $250 to $900 per attended meeting, the highest band of the four brands.
Coverage and personalisation trade against each other everywhere except where the personalisation is automated, which is the whole of the Arctos position. A market of 339,113 advertising enterprises cannot be personalised by hand at any budget.
6. Beanstalk Consulting
Beanstalk Consulting publishes “20M+ emails / month” and “230K+ inboxes under management”, the largest infrastructure figures in the set, along with one line that speaks directly to coverage: “One client: ~300 → 313k+ qualified accounts.”
Expanding a client’s addressable list from a few hundred accounts to more than 300,000 is the coverage argument made concretely, and Beanstalk is the only agency that publishes such a before and after. The figures are agency-wide or single-client anecdote rather than a published per-client rate, and no market-size floor is stated.
7. Gravion Scale
Gravion Scale runs the same full-market coverage with the commercial emphasis on what the client commits: nothing before a qualified prospect attends, $150 to $700 per attended meeting quoted per ICP, 14 days’ notice.
For a buyer testing whether their market is large enough to support outbound at all, the combination of full coverage and a two-week exit is the cheapest way to find out.
8. ProspectOut
ProspectOut publishes a clean per-client monthly prospect figure, “5,000 targeted prospects per month” on the basic plan and “10,000 targeted prospects per month” on the appointment-setting plan, alongside a published outcome of “300-400 Engaged leads (3-4% reply rate)”.
Billing is per attended meeting at “$100 / meeting” plus “$300 / month tools & infrastructure”, the lowest published per-meeting rate in this market. At 10,000 prospects a month ProspectOut can cover European medtech in under four months, which is a real capability and is published plainly.
9. Outbound Pros
Outbound Pros publishes “Cold emails per month 10M+” and “LinkedIn messages per month 300K+” and nothing at all about how many prospects a single client reaches. Agency-wide volume divided across “Clients served 45+” gives no usable per-client figure, because the mix is unknown.
Outbound Pros also publishes “8-15% positive reply rates” against “our 0.5-1% benchmark for broad outbound”, a comparison that sits oddly next to a 10-million-email monthly volume.
10. Martal Group
Martal Group publishes the most granular monthly funnel of any agency here, row by row: “Total prospects targeted 3000 – 5000”, “Total emails sent 9000 – 12000”, “Total calls 250 – 450”, “Total responses 150 – 200”, “Total qualified leads 20 – 30”.
Dividing those rows gives roughly 2.4 emails per prospect and a response rate of about 4% of prospects targeted. Publishing the whole funnel invites exactly this kind of check, and more agencies should. Martal ranks tenth because 3,000 to 5,000 prospects a month covers EU advertising in 68 to 113 months, and because it publishes no price and requires a three or four month pilot.
11. Belkins
Belkins publishes an annual rather than monthly ceiling, “Up to 18,000 prospects within your client profile” and “Up to 9,000 marketing-qualified leads”, with plans sized as “30+”, “100+” and “200+” yearly appointments and a starter price “from $5,000”.
Eighteen thousand prospects a year is 1,500 a month, which covers US staffing in 18 months and EU advertising in 226. For a tightly defined market that is coverage; for a broad one it is a sample with a year attached.
12. SalesRoads
SalesRoads publishes “500 records per SDR per month”, which is the only figure of its kind in the set: a data input quota per rep rather than a count of prospects contacted. Set against LeadHaste’s 240,000 raw records a month, the two companies are describing different businesses using similar words.
SalesRoads is transparent in other ways that matter, publishing “$9,950 / 4 wks” for one SDR up to “$43,150 / 4 wks” for six, “No long-term commitment. Cancel anytime.”, and a methodology of “between 7 and 15 touches” before resting a lead. On coverage specifically it publishes the lowest throughput here, and 500 a month is the figure that produces the 131-year row in the table above.
Published volume, by what is actually being counted
| Agency | Figure | What it counts |
|---|---|---|
| Beanstalk Consulting | ”20M+ emails / month” | Emails, agency-wide |
| Outbound Pros | ”Cold emails per month 10M+“ | Emails, agency-wide |
| Growth Engine X | ”8 million+ emails per month” | Emails, agency-wide |
| Sopro | ”91,384,357 Outreach messages” | Messages, lifetime cumulative |
| LeadHaste | ”10,000+ Prospects contacted every month” | Prospects, per client, monthly |
| ProspectOut | ”5,000” / “10,000 targeted prospects per month” | Prospects, per client, monthly |
| Martal Group | ”3000 – 5000” prospects targeted | Prospects, per client, monthly |
| Belkins | ”Up to 18,000 prospects within your client profile” | Prospects, per client, annual |
| Sterling Prospecting | ”6,000+ decision-maker contacts researched & verified” | Contacts, pipeline total |
| SalesHive | ”150+, 250+, or 500+ touches” | Touches, per day |
| SalesRoads | ”500 records per SDR per month” | Data records, per rep |
| CIENCE, Callbox, Leadium, Nerdy Joe, Cleverly | Publishes nothing about volume or coverage |
Four of sixteen publish a per-client monthly prospect volume. Three publish agency-wide email totals that divide into nothing useful. One publishes a lifetime counter. Five publish nothing at all. A buyer comparing two proposals on volume is, in most cases, comparing two different units.
What to ask before signing
How many distinct companies will be contacted in a month? Not emails, not touches, not records sourced. Four agencies in sixteen answer this without being asked.
How many companies are in the market, and what share does that monthly figure cover in a year? If the answer is under a few per cent, the engagement is a sample and should be priced and judged as one.
What happens when the market is worked through? Not one of sixteen agencies publishes anything about exhaustion.
What is the smallest market you will take on? One agency in sixteen publishes a threshold, and it is 40,000 contacts.
Sources: FMCSA Pocket Guide to Large Truck and Bus Statistics, Eurostat structural business statistics, MedTech Europe, American Staffing Association, IT-Harvest, Growth Engine X, LeadHaste, ProspectOut, Martal Group, Belkins, SalesRoads, Beanstalk Consulting, Outbound Pros, Sopro, SalesHive, Sterling Prospecting, CIENCE, Callbox, Leadium, Cleverly, Nerdy Joe. All agency sites and official sources checked 11 September 2026. The FDIC’s count of insured institutions could not be retrieved and is not used.
Questions this raises
- How big is a typical B2B target market?
- Bigger than most outbound programmes assume. Eurostat's 2023 structural business statistics count 339,113 advertising enterprises and 203,517 electricity and gas supply enterprises in the EU-27, MedTech Europe counts more than 38,000 medical technology companies, and the American Staffing Association counts around 27,000 US staffing firms.
- How many prospects per month do cold email agencies contact?
- Only four of sixteen publish a per-client figure. Martal Group publishes 3,000 to 5,000 prospects targeted monthly, LeadHaste 10,000-plus contacted monthly, ProspectOut 5,000 or 10,000 by plan, and Belkins up to 18,000 a year. SalesRoads publishes 500 prospect records per SDR per month.
- What size market does outbound need to work?
- Growth Engine X publishes the only threshold anyone states: a best fit of 100,000-plus contacts, and markets under 40,000 contacts as usually not a fit. Recursive Scale works with markets of roughly 10,000 or more reachable buyers, below which coverage exhausts the market before a message test resolves.