Outsource lead generation or bring it in-house: what you keep

The proposal says $6,000 a month, a 90-day minimum, and 10 to 15 qualified meetings. It reads well, the references check out, and the deck has a chart. Three months in you have had nine meetings, two proposals out and one signed client, which is roughly break-even. Then comes the question that decides whether any of it was a good deal: if you stopped paying tomorrow, what would still be yours? The domains the emails went out on. The list they were sent to. The replies, the copy that worked, the record of which 40 companies opened three times and never answered. On a standard outsourced engagement the honest answer is almost none of it, and that is the part the pricing conversation never covers.

What are you actually choosing between?

Three routes, not two, and the one most comparisons leave out is the one that settles the argument.

Route one is renting. A lead generation agency runs the campaigns on its own infrastructure: its domains, its sending accounts, its data credits, its inbox. You get meetings on your calendar and a monthly report. Route two is building it yourself, which means buying the same software the vendor buys and putting a person on it. Route three is having the system built inside your business and then run for you: your domains, your accounts, your lists, your inbox, someone else's hands on the daily work.

The reason the question keeps getting answered badly is that routes one and two are compared on price, and price is the least interesting variable. A rented programme and an owned one can cost the same in month six and leave you in completely different positions in month eighteen. One of them has produced an asset. The other has produced a receipt.

The split that matters is between the mechanical work and the judgement. Research, list building, verification, sequencing and follow-up discipline are mechanical. Deciding who belongs on the list, reading a reply that is neither a yes nor a no, and running the call are judgement. Every route has to answer both, and most proposals only answer one.

Why does almost every answer to this question say outsource?

Because lead generation agencies wrote most of them.

Search the question and the first page is mostly vendor content: Callbox, SalesHive, Cleverly, a dozen smaller shops, each with a cost-comparison table that ends where their service begins. The tables are not fabricated. They are just built from one side of the ledger, and they stop counting at the point where the interesting cost appears, which is what the arrangement leaves behind.

The independent numbers tell a less tidy story. SparkToro's February 2026 agency survey found that 59 percent of agencies have attempted outbound sales and 9 percent describe the results as very effective. In the same survey, 70 percent have no full-time sales staff and only 14 percent call their pipeline healthy. So the majority have tried this, through one route or another, and roughly one in eleven is happy with how it went.

Promethean Research's 2026 industry report, built from 119 agency leaders with an average headcount of 31, puts sales, profit margins and lead generation as the three biggest concerns going into the year, against an average net margin of 13 percent. Those two facts sit together. An agency at 13 percent net cannot absorb a $6,000 monthly experiment twice, which is exactly why the decision deserves more than a price comparison.

What does outsourced lead generation cost in 2026?

Between $2,000 and $15,000 a month for the work, plus whatever the contract does to your exit.

Cleverly's March 2026 pricing breakdown is the most specific public rate card in the category, and it matches what lands in founders' inboxes.

Pricing model	Published 2026 range	What the invoice is for	Where it gets tested
Cold email retainer	$2,000 to $8,000 a month	Effort: lists, sends, replies worked	Month four, when the first list is exhausted
LinkedIn retainer	$3,000 to $10,000 a month	Connection and message volume on your profile	Account restrictions and who presses send
Multichannel retainer	$5,000 to $15,000 and up	Both channels plus calling	Attribution between channels
Enterprise programme	$15,000 to $40,000 and up	Named accounts and a dedicated team	Cycle length against contract length
Pay per appointment	$150 to $500 SMB, $300 to $900 mid-market	A meeting held	The definition of qualified

Two things about the shape of that market are worth reading as signals. Belkins publishes appointment volumes rather than prices: its tiers are described as 30 or more yearly appointments for small business, 100 or more for Growth and 200 or more for Growth Plus, with the number you pay available only on a call. SalesHive publishes terms rather than prices too, and its terms are the better ones in the category: one flat monthly fee covering the team, the platform, the data and the tools, month to month, no setup fees, nothing billed until you approve.

Neither page says who owns the domains, the lists or the reply history at the end. That silence is not an oversight. It is the standard arrangement.

What does building it in-house actually cost?

The software runs $120 to $1,270 a month at list price. The cost is the person.

Here are three real stack shapes at published rates. Pick by channels and volume, not by ambition.

Job	Lean, email only	Standard, email and LinkedIn	Multichannel at volume
Mailboxes	Mailforge, 10 at $3 a month billed yearly	Instantly AirMail, 20 at $4	AirMail, 50 at $4
Email sending	Smartlead Base, $39	Instantly Growth, $47	Instantly Hypergrowth, $358
Lists and signals	Manual search, no extra cost	Sales Navigator Core $119.99 plus Apify Starter $19	Sales Navigator Advanced $159.99 plus Apify Scale $199
Enrichment	Anymail Finder Popular, EUR 39	Anymail Finder Business, EUR 129	Clay Launch, $167
LinkedIn sending	None	HeyReach Growth, $79	HeyReach Growth, $79
CRM	A spreadsheet	Close Essentials, $49	Close Growth, $109
Monthly total	about $120	about $530	about $1,270

Verification sits outside the table because it is bought in packs rather than by the month: MillionVerifier sells 50,000 credits for $89, which at one verification per contact covers a year of lean sending and about four months of the multichannel column. Domains are the other one-off, at roughly $14 a year each.

So the thing an agency charges $6,000 a month to run costs about $1,270 a month in software at full volume. The rest of that gap is the operator, and the operator is the part founders underestimate.

The Bridge Group's 2025 study of 351 B2B companies measures the day the operator actually has. The median rep runs 112 activities and produces 4.1 quality conversations. Most of that is research, typing, looking things up and keeping track. If you build in-house, somebody absorbs those 112, and the realistic ask on a founder-led team is ten hours a week from one person who already has a job.

Hiring for it changes the arithmetic rather than removing the problem. The same study puts median on-target earnings at $80,000, split $55,000 base and $25,000 variable, with a 3.0 month ramp and 1.9 years average tenure. The Bureau of Labor Statistics has benefits at 30.0 percent of private-sector employer compensation cost, which adds roughly $34,000. Call it $115,000 for year one with nine productive months inside it, which is covered in more detail in the guide on hiring an SDR versus building the system first.

What do you keep when the engagement ends?

Whatever the contract names, and on a standard outsourced engagement that list is short.

This is the column that should decide the choice, because it is the only one that compounds. A rented programme pays for itself inside the month or it does not. An owned one keeps paying after you stop spending, because the warmed domains, the cleaned list and the reply history are inputs to the next campaign rather than someone else's property.

Asset	Rented from an agency	Built and run in-house	Built inside your business
Sending domains and mailboxes	The vendor's, warmed on their reputation	Yours	Yours
Verified list and enrichment	Usually the vendor's, exported on request	Yours	Yours
Sequences, copy and test results	The vendor's working files	Yours	Yours
Reply history	In the vendor's inbox	Yours	Yours
LinkedIn connections	Yours, since they sit on your profile	Yours	Yours
Pipeline records	Only what gets pushed to your CRM	Yours	Yours

Contract language is the whole game here, and it is boring enough that founders skim it. Sprintlaw's guidance on lead generation agreements puts the test plainly: confirm who owns prospect data, reports, accounts and any CRM records created during the campaign, because "if the agency builds everything under its own accounts, switching suppliers can become expensive and disruptive." The same guidance prefers a three-month arrangement with a clear break right over a long lock-in for smaller businesses, which is the opposite of the three to six month minimums most providers open with.

Which option costs less per booked meeting?

Run the number yourself, because on identical volume the three routes land at roughly $750, $1,600 and $310 a meeting.

Start from the rate rather than a promised count. lemlist's 2026 benchmark compilation puts the average cold email reply rate at 3.43 percent and the meeting-booked rate at 0.5 to 2.5 percent of contacts reached, with 58 percent of all replies arriving from the first email in a sequence. Mailshake's 2026 report is more specific for our side of the market: agencies and consultancies reply at 2.5 to 4.5 percent, with elite campaigns above 7 percent.

Say you run a $120,000 a month agency and you want eight booked meetings a month, which at a 25 percent close rate and a $4,000 average deal is $8,000 of new monthly revenue. At a 1 percent meeting-booked rate, eight meetings means about 800 verified contacts worked every month. That volume is the fixed thing. What changes is who does it and what it costs.

The outsourced bar is a mid-range multichannel programme at $6,000 a month against eight meetings, which works out at $750 each and sits inside Cleverly's own published $300 to $900 band for mid-market appointments. The hire bar is $115,000 of year-one employer cost spread over nine productive months, so about $12,700 a month, which is $1,600 a meeting before the software. The third bar is $530 of stack plus ten hours a week from someone already on payroll at roughly $45 an hour loaded, so about $2,465 a month and $310 a meeting.

Two honest corrections to that picture. The cheap bar carries a build you have not paid for yet, whether you build it over six weekends or buy the build once. And the cheap bar assumes those ten hours a week actually exist, which on a founder-led team is the assumption that fails most often. If the hours are imaginary, the real comparison is between the $750 bar and no outbound at all.

Who owns the LinkedIn account and the prospect data?

You do, and both facts limit what an outsourced provider can legitimately do on your behalf.

LinkedIn's User Agreement is explicit on each. Section 2.2 asks you not to "share or transfer your account or any part of it", and section 8.2 bans using "bots or other unauthorized automated methods to access the Services, add or download contacts, send or redirect messages". A provider running outreach from your profile is doing one of those two things, and the account that gets restricted is yours, not theirs. That is the main reason the LinkedIn half of an outsourced programme tends to be the half that quietly stalls. The practical rule for any route is that sends happen from seats your people control, with a queue someone approves, which is also the subject of the guide on building the outbound system itself.

The data side is a contract question with a legal floor underneath it. Under UK and EU data protection rules the organisation that decides why and how personal data gets processed is the controller, and the ICO defines controllers as "the main decision-makers" who "exercise overall control over the purposes and means of the processing", with processors acting only on the controller's instructions. If a vendor chooses your targeting, buys the data and decides the cadence, the role is not obviously a processor's. Pin it down in writing before launch, along with what happens to the records at the end.

Ask for four things specifically: an export of every contact and every reply in CSV, monthly, with no post-termination restriction; a named transfer path for the sending domains; the suppression list; and the notice period in days rather than months.

How do you decide between the three routes?

Pick by what you are short of, because each route buys back a different scarce thing.

1. Short of time, not money, and you need meetings this quarter. Rent. Take a month-to-month arrangement, write the export clause in, and treat it as a paid test of whether your offer works cold rather than a permanent channel.
2. Short of money, with one person who has ten real hours a week. Build it yourself. Start lean, email only, one signal, 20 to 30 sends per mailbox per day, and read the sizing guide on domains and inboxes before you buy anything.
3. Short of bandwidth but not of commitment, and you want the asset. Have the system built inside your business and run for you. This is the route that answers the ownership problem without the ramp, and it is the one the vendor tables leave out.
4. Short of a clear buyer or an offer that sells cold. None of the three works yet. Fix the offer first, because a weak offer fails identically on all three routes and costs the most on the rented one.

Ourix builds the third route for founder-led B2B businesses, agencies first: a working outbound system on the client's own domains and accounts, run by Ourix, with verified lists of the buyers they want, outreach on email and LinkedIn, every reply in one inbox and a pipeline view of what got booked. Campaigns go live in about 30 days, everything is the client's to take over at any time, and there is no lock-in. The fit is a business with clients, a team and an offer that already sells, typically $30K to $300K or more a month; it is the wrong call for anyone still finding product-market fit, selling B2C or low-ticket, or needing meetings inside two weeks. Ourix does not publish prices. The build is a one-time fee scoped to your channels and your market, and you get the exact number on the call.

When is outsourcing the right call?

When speed matters more than ownership, which is a real situation and not a failure of nerve.

Three cases make renting the better decision. You have a seasonal window and need conversations inside four weeks, so the warmup and the learning curve are both time you do not have. You are testing a new vertical or a new offer and want to know whether strangers respond before you buy infrastructure for it. Or the channel is one you have no intention of operating ever, such as outbound calling into a market where a dialer team and local numbers are the whole advantage.

A good provider in those cases looks specific. Month-to-month terms, a written definition of a qualifying appointment, a named person who works the replies, and a willingness to hand over the data monthly. SalesHive's published terms are close to that shape. Most are not, and the ones that open with a six-month minimum and a setup fee are pricing their own churn rather than your result.

What makes renting the wrong call is using it as a permanent substitute for a channel you need every month. Twelve months at $6,000 is $72,000 spent on meetings you then have to keep buying, against the same money building something that still runs in year two. The referral problem this is usually meant to solve is the same one covered in the guide on getting clients without relying on referrals: a channel you cannot turn up on demand is not a channel.

What to do next

1. Write down the number of booked meetings a month you need, derived from your close rate and average deal size, not from a round figure that sounds ambitious.
2. Multiply it by 100 at a 1 percent meeting-booked rate to get the monthly contact volume any route has to produce. Sanity-check it against the 0.5 to 2.5 percent band before you believe it.
3. Price all three routes at that volume: the published retainer range for your channels, the stack shapes above, and a build you own. Put the numbers in one table.
4. Before you sign anything outsourced, ask for the exit terms in writing: data export format and frequency, domain transfer, suppression list, notice period in days.
5. If the offer has never been tested cold, test the offer before the channel. A rented programme is the most expensive place to discover a positioning problem.
6. If you want the volume math and the route choice done against your actual numbers, the free growth audit produces both as a document you keep.