We are an agency, so the obvious incentive here is to talk you out of this. Instead: above roughly fifteen to twenty acquired links a month sustained for a year, in-house economics genuinely win, and a company at that volume should hire.
What follows is the part that does not appear in the business case — the ramp, and the four fixed costs that were previously spread across an agency's roster and are now yours alone.
The four fixed costs you have just taken on
1. A prospect index
Knowing which publications in your category accept contributed content, who edits them, what they have published, what they have refused, and when you last contacted them.
Building this from nothing takes six to nine months of work that produces very few links. An agency amortises it across every client in the vertical; you now carry it for one company.
2. Sender reputation
Cold outreach from a domain with no sending history goes to spam. Establishing an identity editors recognise takes months of low-volume, carefully written correspondence.
This is the most underestimated line in the whole transition, and the easiest to destroy. One enthusiastic person with a sequencing tool undoes a year of it permanently.
3. Scoring infrastructure
Deciding which of two hundred candidates are worth pitching requires traffic verification, topical overlap scoring, footprint clustering and outbound hygiene checks. Manageable by hand for ten domains a month; not for two hundred.
4. Editor relationships
The reason a pitch gets read in week one rather than week nine. These accrue slowly, belong to individuals rather than companies, and leave when the person does.
None of these four appear on a salary line. All four are the reason the first year costs more and produces less than the spreadsheet said.
The realistic first year
| Months | What happens | Links |
|---|---|---|
| 0–2 | Hiring and onboarding. Tooling procurement. Gap analysis. | 0 |
| 2–4 | Prospect list built from scratch. First pitches. Sending domain warming. | 0–4 |
| 4–6 | First placements land. Reply rates low — no reputation yet. | 3–8/mo |
| 6–9 | Relationships forming. Reply rates climbing. Process settling. | 8–14/mo |
| 9–12 | Approaching steady state. | 12–20/mo |
Cumulative first-year output: roughly 70–120 placements, most of them in the second half. Compare with an agency delivering from month two at a steady rate — the agency wins year one comfortably and loses year three.
The honest cost comparison
| Line | Year one, in-house |
|---|---|
| Salary, one experienced outreach specialist | $65,000–$95,000 |
| Employer costs, roughly | +20–30% |
| Tooling (backlink suite, outreach, verification) | $4,000–$9,000 |
| Publisher fees and placement costs | $25,000–$60,000 |
| Content production | $15,000–$35,000 |
| Management time | Real, and rarely counted |
Against 70–120 placements, that lands somewhere between $1,100 and $2,200 per link in year one — versus a market average of roughly $500. By year two, with the fixed costs already paid, the same spend produces 200+ placements and the unit cost falls below agency rates.
That crossover is the actual argument for hiring, and it only pays if you are still funding the function in year two.
Who to hire
Two profiles work and one does not.
Works: an outreach specialist with existing relationships in your vertical. You are buying the relationships, which is the expensive part. Ramp compresses to two or three months.
Works: a former journalist or trade editor. They know how publications decide, they write to a publishable standard, and editors recognise the register. Slower on the technical side, faster on everything that matters.
Does not work: a generalist marketer given link building as one of five responsibilities. This is the most common version and it fails predictably — the ramp never completes because the role never has enough of anyone's week.
What to do in months 0–4, when nothing is landing
The ramp is survivable if the first four months produce something other than zero. Four things that do:
- Claim the partner ecosystem. Integration directories, associations, conferences, customer case studies. Twenty to fifty domains in weeks two to six, at almost no cost.
- Fix internal linking. Two weeks, no third party involved, movement in four to six weeks.
- Audit and correct round-up listings. Correction requests convert at 40–60% and produce referral demos immediately.
- Recover your own broken inbound links. Pull 404s with referring domains and redirect them. Frequently ten to sixty domains you already earned.
All four are available to a new hire in week one and none of them require a prospect index or a sending reputation. They are also what makes month four look like progress rather than an expensive mistake.
The hybrid most companies should actually run
Keep an agency running acquisition during the hire's ramp, then taper.
Months 0–6: agency at full volume; new hire builds the prospect index, runs the four cheap workstreams above, and learns the vertical.
Months 6–12: agency at half volume; hire takes over editorial acquisition.
Month 12+: hire owns acquisition; agency retained for digital PR and category placement, which need relationships that take years.
It costs more in year one than either pure option and it is the only version where the gap does not widen during the transition.
What stays hard in-house, permanently
Two things do not get easier with time, because they depend on breadth rather than depth.
Digital PR reach. A journalist list built across many clients in a vertical is structurally wider than one built for a single company. This is the workstream most companies keep outsourced permanently, and sensibly.
Category placement at scale. Auditing and maintaining relationships with 60–200 round-up publishers is a lot of surface area for one person who also has to acquire.
The signal that the hire is working
Not link count in month four. Reply rate.
A specialist building genuine relationships sees reply rates climb from single digits to 15–25% between months three and nine. That curve is the leading indicator, and it appears months before the placement count does.
If reply rates are flat at month eight, something is wrong with the pitching or the target list — and that is diagnosable and fixable, unlike "we have not got many links yet".