At $5,000 a month you are already paying for a part-time person without getting one. Here is the actual cost comparison across in-house, agency, and self-managed tooling — and the site-size thresholds where each one stops making sense.
Pricing current as of August 2026, drawn from published industry studies. Treat every figure as a band rather than a quote.
The three routes get compared on price per link, which is the wrong axis. They have different cost structures, different fixed and variable components, and different points at which they stop scaling.
Comparing them properly means modelling each at your actual volume rather than reading an average.
The three cost structures
Agency. Mostly variable. You pay per link or per retainer, and the cost rises roughly with output. Little fixed overhead on your side.
In-house. Mostly fixed. A salary and a tool stack cost the same whether you place five links or twenty-five. Marginal cost per additional link is low once the fixed cost is covered.
Self-managed with tools. Low fixed, high time. A subscription plus your own hours, which are real even when they are not invoiced.
Those structures cross at different volumes, which is why the right answer changes as you grow rather than being a fixed preference.
The numbers, modelled
Agency route. Reported per-link averages around $1,459 for agency-mediated placements, or retainers of $3,000 to $10,000 monthly producing roughly five to fifteen links. At the mid-point: $5,000 for ten links, so $500 each.
In-house route. $3,000 to $8,000 monthly in salary and tooling, before placement fees. A competent mid-level person plus a backlink tool subscription sits around $5,000 to $6,000 all-in. Output depends entirely on the person, but 10 to 20 links a month is realistic for someone whose whole job it is.
At 15 links a month on $5,500 fixed cost plus, say, $200 average placement fees on half of them: roughly $467 a link. Comparable to the agency figure, with the difference concentrated in what else you get.
Self-managed route. A tool subscription at $100 to $500 monthly, plus your own time. If link building takes eight hours a month and your time is worth $75 an hour, that is $600 in time plus $200 in tooling on $800 total. At four links a month, $200 each — the cheapest per link and the lowest ceiling.
What per-link cost hides
Three things, all of which change the ranking.
Retained knowledge. An in-house person accumulates partner relationships, niche understanding, and process. An agency accumulates the same things and keeps them when the engagement ends. Over two years that difference compounds considerably.
Attention. An agency account manager runs several clients. An in-house person runs one site. Whether that matters depends on whether your programme needs continuous judgement or periodic execution.
Opportunity cost of your own hours. Self-managed looks cheapest because the time is not invoiced. If those eight hours would otherwise go to something with higher return, the real cost is that return rather than $600.
And the cost nobody models in any route: decay. Every figure above is cost per link, which is the wrong unit because links are not permanent. The honest unit is cost per surviving link-year, which the link building cost comparison breakdown covers in full — and it changes the ranking, because routes differ in how well they monitor and replace losses.
If you go in-house, what you are actually hiring for
The most common in-house failure is hiring the wrong profile, and it happens because link building looks like an outreach job.
What the role actually requires: analytical judgement to assess a site in five minutes, process discipline to maintain standards under pressure, enough writing ability to produce a genuinely specific request, and enough patience to keep pace when nothing visible is happening for six weeks.
What it does not require: a large existing network, which is what agencies charge for and what a hire will not bring at a normal salary.
Two practical points on this:
A generalist SEO frequently outperforms a link building specialist for sites under DR 40, because at that stage the highest-return work — internal routing, content structure, technical fixes, relationship sweeps — sits outside link building narrowly defined.
The role is easier to hire junior and train than to hire senior. The judgement is teachable from a written standard, and a written standard is something you need anyway. What is not teachable is the willingness to run the same audit every quarter, which is the actual predictor of whether the programme works.
The thresholds where each route breaks
Self-managed breaks when your constraint becomes capacity rather than process. The signal is specific: you know exactly what to do next and cannot find the hours to do it. Before that point, hiring or outsourcing buys you execution of a plan you have not made, which is expensive.
Rough threshold: somewhere around 5 to 8 links a month, or when the audit and monitoring work alone exceeds a few hours monthly.
Agency breaks when the retainer approaches a salary. At $5,000 a month you are spending $60,000 a year for someone’s partial attention across several accounts. At that point hiring gets you full attention, retained knowledge, and someone who can also do the adjacent work — content, technical fixes, internal routing — that an agency scoped out of the contract.
Rough threshold: $5,000 to $6,000 monthly, adjusted for what a competent hire costs in your market.
In-house breaks in two directions. Below roughly 8 to 10 links a month the fixed cost is not justified — you are paying a salary for part-time output. And at the top end, in-house cannot reach placements that require established publisher relationships or genuine PR capability, which is where a specialist agency retains an advantage regardless of your budget.
By site size, which is the more useful cut
Under DR 20. Self-managed, without question. The highest-return tactics at this stage — relationship links, unlinked mentions, lost link reclamation, internal routing — are ones nobody can do better than you, because they depend on relationships you have and knowledge of your own site. Paying an agency here buys execution of work you should be doing directly.
DR 20–40. Self-managed with tooling, or a small retainer for a specific gap. This is the band where a repeatable process matters more than access, and where most sites overspend on agencies out of impatience.
DR 40–60. The genuine decision point. Volume is high enough to justify fixed cost, and the tactics that work are increasingly relationship-dependent. Either in-house or agency is defensible; the deciding factor is usually whether you want the knowledge retained.
Above DR 60. In-house for the sustained work, plus specialist agency or PR for high-tier placements. At this level the two are complements rather than alternatives, and the price gap between tiers is steep — each additional 10 DR points reportedly raises placement cost by around a third, which the backlink pricing 2026 breakdown covers in detail.
The split that works better than any single route
The pattern that rarely gets proposed because neither agencies nor tool vendors benefit from it: divide the work by whether access is the constraint.
Keep internal: relationship links, unlinked mention sweeps, lost link recovery, internal routing, exchange programmes, monitoring and audits. None of these require access you do not have. They require consistency, which is a process problem, and an agency has no structural advantage.
Outsource: digital PR, high-tier publication placements, anything that genuinely depends on relationships built over years.
Splitting on that line usually costs less than a full retainer and produces more, because the internal half is the half where paying someone buys you nothing except the fact that it gets done.
The obvious failure mode: it gets done only if someone owns it. A split arrangement with no internal owner becomes a retainer with extra steps.
The question to answer before any of this
Before choosing a route, work out whether external acquisition is your constraint at all.
A site with authority pooling on an old blog post while its commercial pages sit starved does not have a link building problem. It has a distribution problem, and no amount of new acquisition fixes it efficiently — routing what you already have is faster, free, and requires nobody’s cooperation.
The diagnostic takes twenty minutes: check which of your pages hold the most external links, check which pages you actually want ranking, and see whether the two lists overlap. On most sites they barely do.
Answer that first. It determines whether you are choosing between three ways to spend money or discovering that the cheapest option is one you have not tried — and it is the same diagnostic the how much do backlinks cost breakdown recommends running before committing to any acquisition budget.
Build the comparison with your own numbers
- Target links per quarter, from the referring domain gap on your target pages — not a round number
- Your quality floor, in DR and minimum traffic
- Expected pass rate at that floor
- Cost per attempt under each route, including failures
- Your own hours, priced honestly
- Assumed survival rate — and note that this differs by route, since monitoring quality varies
Run it for all three. The answer is frequently not the one that looks cheapest per link, because the routes differ most in the columns that per-link pricing leaves out.
And revisit it annually. The thresholds above move as your profile grows, which means the route that was right at DR 25 is frequently wrong at DR 45 — and staying with a route past its threshold is the most common and least visible way link building budgets get wasted.
