This decision usually gets made by comparing a salary to a retainer, which is like comparing a mortgage to a hotel bill. Both numbers are real and neither is the cost. Here is the comparison done properly: a fully loaded in-house figure, a weighting method that makes retainers comparable, and the volume thresholds where the answer actually changes.
Fully loaded, a US in-house content marketing manager costs about $122,000 a year, or roughly $10,200 a month — 57% more than the $77,858 average salary. At a realistic solo output of five weighted content units a month, that is about $2,040 per unit, against $495 for freelance production and $415 to $665 for a mid-market agency retainer. In-house only becomes cost-competitive above roughly 25 to 30 units a month; below eight units it is the most expensive option available.
Why the usual comparison produces the wrong answer
The standard version of this comparison lines up three prices: a freelance writer at $300 to $600 a long-form post, an agency retainer at $3,000 to $15,000 a month, an in-house content marketing manager at a US average salary of about $77,900. Then it says agencies bring a team, freelancers bring flexibility and in-house brings brand knowledge, and leaves you exactly where you started.
The comparison fails for two reasons. A salary is not a cost, it is roughly two thirds of one. And a retainer is not a price per thing, it is a price for a bundle whose contents vary wildly between proposals. Until both are converted to the same denominator, you are comparing a number to a different kind of number.
So convert them. The denominator that works is cost per published asset, weighted for effort.
Defining the unit
Not all content is the same size, so counting "pieces" flatters whoever produces the most small ones. We weight everything against one long-form article, which we call one content unit:
| Asset | Units | Why |
|---|---|---|
| Long-form article, 1,200–2,000 words, researched and edited | 1.0 | The reference asset |
| Short article or landing page copy | 0.6 | Less research, same editorial overhead |
| Short-form video, scripted and edited (30–60s) | 0.8 | Production cost offsets lower word count |
| Email or newsletter issue | 0.3 | Shorter, usually reuses existing material |
| Static social post with original design | 0.2 | Design time dominates |
| Social post repurposed from an existing asset | 0.1 | Near-zero marginal research |
The weights are not precise science and they do not need to be. Their job is to stop a proposal offering "40 pieces of content a month" from looking four times better than one offering ten, when 36 of the 40 are crops of the other four. Ask any agency to restate their deliverable in these units and the comparison becomes possible in about five minutes.
What in-house actually costs
Take the US average content marketing manager salary of $77,858 as of mid-2026 and build it out:
| Line | Annual | Note |
|---|---|---|
| Base salary | $77,858 | US average, mid-2026. Range $60,000–$87,000 at the quartiles |
| Employer taxes and benefits at 25% | $19,465 | Payroll taxes, health cover, pension, insurance |
| Tooling at $600/month | $7,200 | SEO suite, design software, stock, scheduling, AI subscriptions |
| Management and review time | $12,000 | Roughly 10% of a senior marketer's loaded time |
| Recruitment, amortised | $5,840 | 15% placement fee spread over a two-year tenure |
| Loaded annual cost | $122,363 | About $10,200 a month |
The base salary is 64% of the true figure. That gap is the single most common error in this decision, and it is why in-house consistently looks cheaper than it is at the point the decision gets made.
Now the output side, which is the part nobody wants to write down. A content marketing manager is a generalist: they set strategy, write, brief designers, publish, distribute and report. A realistic sustained output for one person doing all of that is four to six content units a month, not the ten or twelve assumed in the business case. At five units, the in-house cost per unit is about $2,040.
Before hiring, write down the expected monthly output in content units and have the hiring manager sign it. If the number is above eight, you are not hiring a content marketing manager, you are hiring a production team and calling it one job. That mismatch is the most common reason a first content hire is judged a failure within a year.
The three-way comparison at three volumes
Costs below use mid-market 2026 figures: freelance long-form at $450 per piece, internal briefing and editing at $60 an hour loaded, agency retainers at the published mid-market range, and the loaded in-house cost above.
| Monthly volume | Freelance-led | Agency retainer | In-house | Cheapest per unit |
|---|---|---|---|---|
| 4 units | ~$1,980 total $495/unit | $2,500–$4,000 $625–$1,000/unit | $10,200 $2,550/unit | Freelance, by a distance |
| 12 units | ~$9,480 total with a part-time internal owner $790/unit | $5,000–$8,000 $415–$665/unit | Not achievable by one person | Agency |
| 30 units | Coordination overhead dominates | $12,000–$20,000 $400–$665/unit | Two-person team plus overflow, ~$20,000 ~$665/unit | Roughly level |
Three conclusions fall straight out of the table, and they are not the conclusions the category usually reaches.
- Below about eight units a month, in-house is the most expensive option available, often by three to five times per unit. A single salary is a large fixed cost spread across modest output. If your content plan is four articles a month, hiring someone to write them is the worst-value structure of the three.
- Between roughly eight and twenty-five units, agencies win on cost per unit, because a retainer buys fractional access to a writer, an editor, a designer and an editor's judgement without paying for any of them full time. The specialism coverage is the actual product, not the word count.
- Above roughly twenty-five to thirty units a month the two converge, and the decision stops being financial. At that volume you are choosing between control and coordination, and cost per unit no longer separates the options.
That crossover point — call it 25 to 30 content units a month — is the number worth remembering, because it is the only place in this decision where the money genuinely stops mattering.
What you are actually buying
Cost per unit decides the shortlist. It does not decide the choice, because the three models do not deliver the same set of capabilities.
| Capability | Freelance | Agency | In-house |
|---|---|---|---|
| Strategy and prioritisation | Rare, and usually not what you are paying for | Yes, and the main reason to pay a retainer | Yes, with the deepest business context |
| Specialist coverage (writing, design, video, SEO) | One specialism per person | Several under one contract | Whatever the one person happens to have |
| Brand and product knowledge | Builds slowly, lost on churn | Builds with account tenure | Deepest, and compounds |
| Turnaround on a same-week request | Depends entirely on availability | Contracted, but queued | Fastest |
| Volume flexibility month to month | High | Medium, within retainer scope | None — the cost is fixed |
| Key-person risk | Medium | Low | High — one resignation stops everything |
| Distribution and measurement | Almost never included | Sometimes, check the scope | Yes, if the person has time |
Read the last row carefully when you compare proposals. A large share of content programmes fail not because the content was bad but because nothing happened after publication, and distribution is the line most quietly absent from both freelance arrangements and cheap retainers.
The hybrid that usually wins
For most businesses between four and twenty units a month, the structure that produces the best cost per unit is not on the list of three. It is one internal owner — often a marketing generalist who already works there, at perhaps a third of their time — who holds strategy, priorities and brand judgement, plus external production for everything else.
The internal owner is doing the work that cannot be outsourced well: deciding what to make, knowing which claims the business can stand behind, and being accountable for whether it worked. Production, design and video go out. The failure mode of pure outsourcing is a programme nobody inside the business owns; the failure mode of pure in-house is one person doing six jobs adequately. It is also the shape most of our own content engagements take, for the straightforward reason that it is the only one where both the judgement and the production sit with whoever is best placed to do them.
Two practical notes on making the hybrid work. Give the internal owner authority over the brief, not just approval over the output — briefs written by the person who understands the customer are the cheapest quality intervention available. And do not split video across the boundary: production quality on video is discontinuous in a way it is not for text, which is why what a brand video actually costs is a separate conversation with a separate budget.
The costs everyone forgets
- Ramp time. A new in-house hire produces little of value for six to ten weeks. Amortised over a two-year tenure that is roughly 6% of the loaded cost, and it lands entirely in the quarter you can least afford it.
- Your own time. Every model consumes internal hours: briefing, review, approvals, subject-matter interviews. Agencies reduce it, they do not remove it. Budget two to four hours per content unit of somebody's real time and the comparison shifts, usually against whichever option you were about to choose for being "hands-off".
- The rework rate. Ask any prospective partner what proportion of drafts need a second substantive round, and treat an answer below 20% as a sales number rather than an observation.
- Asset ownership. Confirm in writing that you own the output, the source files and the raw footage. This is boring until you change supplier, at which point it is the only thing that matters.
- The AI discount that is not one. Every model now uses AI assistance and every model is quoting as if that is a differentiator. It compresses drafting time, not research, interviews, judgement or editing, and the visible gap between AI-assisted content that gets cited and content that gets ignored is entirely in the parts AI did not compress.
A decision rule you can apply today
- Write the plan in content units, weighted with the table above, before you talk to anyone. Most plans shrink by a third at this step, which changes the answer.
- Under eight units a month: freelance production with an internal owner. Do not hire. Do not sign a retainer larger than the work.
- Eight to twenty-five units: agency retainer, with the deliverable restated in units and distribution explicitly in scope.
- Above twenty-five units: run the comparison on control, risk and speed rather than cost, because the money is level. In-house becomes defensible here and not before.
- Recalculate at twelve months. Volume needs change, and a structure chosen for last year's plan is the most common reason a content programme feels expensive without anyone being able to say why.
One last framing. Industry guidance putting content at 20% to 30% of a marketing budget is a useful sanity check on the total, not a method for choosing a structure. A programme that spends 25% of the budget producing content nobody distributes is still a failed programme, and the structural question is only worth answering once the plan is worth executing.
Frequently asked questions
How much does an in-house content marketer really cost?
In the US in 2026, about $122,000 a year fully loaded, against an average base salary of $77,858. The difference is employer taxes and benefits at roughly 25% of salary, tooling at around $600 a month, management and review time worth about $12,000 a year, and recruitment costs amortised over the expected tenure. Base salary is only about 64% of the true cost.
How much does a content marketing agency cost in 2026?
Published mid-market retainers run from roughly $1,500 to $10,000 a month for most US businesses, with B2B programmes commonly quoted between $3,000 and $15,000 and full-service strategy, production and distribution between $5,000 and $20,000. The useful comparison is not the retainer but the retainer divided by weighted output, which typically lands between $400 and $665 per long-form-equivalent unit.
What does a freelance writer charge per article in 2026?
Commonly $150 to $300 for a short post of 500 to 800 words and $300 to $600 for long-form of 1,200 to 2,000 words, which works out at roughly $0.10 to $1.00 per word depending on subject expertise. Add your own briefing and editing time, usually one to two hours per piece, to get the real cost per published asset.
At what volume does in-house content become cheaper than an agency?
At roughly 25 to 30 weighted content units a month. Below eight units a month, a single salary spread across modest output makes in-house three to five times more expensive per unit than freelance production. Between eight and 25 units, an agency retainer generally wins because it buys fractional access to several specialisms. Above 30 units the two converge and the decision should be made on control and risk rather than cost.
What is a content unit and why weight assets?
A content unit is one researched, edited long-form article of 1,200 to 2,000 words. Other assets are weighted against it: short article 0.6, scripted short-form video 0.8, email 0.3, original static social post 0.2, repurposed social post 0.1. Weighting stops a proposal offering 40 assets a month from appearing four times better than one offering ten, when most of the 40 are crops of the same few pieces.
Is a hybrid content model better than choosing one?
For most businesses producing four to twenty units a month, yes. One internal owner, often a marketing generalist at about a third of their time, holds strategy, priorities and brand judgement; production, design and video go outside. It avoids both failure modes: an outsourced programme nobody internally owns, and one in-house person doing six jobs adequately.
Does AI reduce what content should cost?
It compresses drafting time, which is a minority of the real cost. Research, subject-matter interviews, editorial judgement, fact-checking and distribution are unchanged, and those are precisely the parts that determine whether content gets cited or ignored. Treat a quote discounted purely on AI assistance as a claim about drafting speed, and ask what happens to the other stages.
The takeaway
Convert everything to one denominator before you decide. Load the salary properly, restate every retainer in weighted content units, and add your own hours to all three columns. Do that and the shape of the answer is consistent: under eight units a month, hiring is the most expensive thing you can do; between eight and twenty-five, an agency's fractional specialists win on cost per unit; above thirty, the money is level and you are really choosing between control and coordination. Most businesses land in the middle and are best served by a hybrid — one internal owner who decides what gets made, and external production for everything else.