An owner spending $40,000 a month on acquisition asks a fair question: hire someone, or pay a firm? Two tabs get opened. One holds a job posting with a salary on it. The other holds a proposal with a monthly fee on it. Both numbers look like they measure the same thing. They do not, and that is where in house marketing vs agency comparisons fall apart, because only one of the two is already fully loaded.
A retainer is the entire cost of the retainer. A salary is the first line of a cost.
The number on the job post is the smallest piece of the seat
Start with the part nobody argues about. The IRS sets the employer's share of Social Security at 6.2% and the employer's share of Medicare at 1.45%. Those two added together are 7.65% on top of wages, before a single benefit. That is a hard cost, invoiced every pay period, and it does not appear anywhere in the salary range on the posting.
There is a wrinkle worth knowing if you are hiring senior. The Social Security portion applies only up to a wage base, and for earnings in 2026 that base limit is $184,500. Pay a seat $200,000 and the employer's Social Security stops at 6.2% of $184,500, which is $11,439, while Medicare keeps running at 1.45% of the full $200,000, or $2,900. Total employer payroll tax: $14,339, rather than the $15,300 you would get by applying 7.65% to the whole salary. Small, and worth being right about when you are modelling a director-level hire.
Everything after payroll tax depends on your own plan design, so use your own figures: employer health contribution, retirement match, paid time off (which does not change the cost and does reduce the working weeks you get for it), workers compensation and unemployment insurance, payroll processing, a laptop, and the software seats an outside partner would otherwise have supplied out of their own fee.
Then there are three costs I am deliberately not going to put a number on, because the published figures I could find for them are either stale or unverifiable: recruiting, ramp, and replacement. You will spend real money and real weeks filling the role. The person will not be at full output in month one. And if they leave in month nine, you restart the whole sequence while the campaigns sit still. Treat those qualitatively in your decision, which is to say treat them as reasons the in-house number is understated rather than overstated.
A worked example you can re-run on your own numbers
This is a model, not a benchmark. None of the inputs below are published data, and nothing here says what a marketing hire costs in your market. Swap in your own figures and the arithmetic still works.
Assume one mid-to-senior generalist at a $110,000 base, a number you pulled from live postings for your city and your scope. Assume employer payroll tax at 7.65%, from the IRS rates above. Assume you contribute $14,000 a year toward health coverage and match 3% of salary into retirement. Assume $900 a month in software and hardware you would not otherwise be buying.
The math:
- Base: $110,000
- Employer payroll tax at 7.65% of $110,000: $8,415
- Health contribution: $14,000
- 3% retirement match: $3,300
- Tooling at $900 x 12: $10,800
That totals $146,515 a year, or roughly $12,210 a month. Against your assumed base salary of $110,000, the loaded figure is about 33% higher. Nothing in that $12,210 has reached a media platform yet, and nobody has written an ad.
Now put it inside the budget. On a $40,000 monthly acquisition envelope, that one seat consumes just over 30% of the total and leaves $27,790 for media, creative production, and tooling you have not counted. Add a second seat at the same load and you are at $24,420 a month of payroll with $15,580 left to spend. That is the real shape of the decision, and it is a question about allocation rather than a question about which model is cheaper.
Re-run it at your own salary assumption and your own budget before you read another comparison post. The output is specific to you, which is exactly why generic verdicts on this topic are worth so little.
What retainer pricing data tells you, and where it stops
On the other side of the comparison, published pricing is thinner than most buyers expect. In Databox's survey of agencies, most respondents (38%) reported monthly retainers of $1,001 to $2,500, a further 22% sat between $2,501 and $5,000, and fewer than 10% charged $10,001 to $15,000 a month. The same survey found that about 38% of contributing agencies prefer retainer-based pricing over other models.
Read that carefully before you use it. Those bands are agency-reported across every client size in the sample, so they are not a quote for your scope. What the distribution does tell you is that the market's centre of gravity is a long way below the loaded cost of a single senior employee. If the modal retainer sits in the low thousands and one in-house seat models out at $12,000 a month on the assumptions above, the useful question becomes how much senior attention your fee actually buys at that level, and from whom.
That is a scoping and pricing-structure question, and the fee model shapes the answer more than the dollar figure does. Our breakdown of what each agency fee model rewards covers how a retainer, a project fee, and a performance arrangement each bend behaviour in a different direction.
Where in house marketing vs agency becomes a scoping question
The binary framing is close to fictional in practice. Forrester reports that 93% of respondents to its 2026 B2B Brand And Communications Survey use agencies in some capacity. Almost every company with an in-house team is also paying someone outside. The decision in front of you is which line items sit where.
Buyers are getting pickier about those line items rather than walking away. In the same Forrester data, the share of marketers expecting to increase agency spend on digital marketing fell from 51% to 31% year over year, and for content creation and development the share dropped from 41% to 26%. Expectations of more spend are cooling in the categories where internal tooling has improved fastest. Spend is being re-scoped, not switched off.
A practical split that holds up: keep institutional knowledge inside. Your offer, your pricing, your objection handling, your sales process, your data. Those compound in whoever holds them, and you do not want them walking out the door at the end of a contract. Buy the things that need either specialist reps or an outside pair of eyes: paid media execution at scale, creative volume, technical build work, and the diagnostic read on why a channel stalled. If the thing you are actually missing is someone to own the plan rather than someone to run the channels, the comparison you want is fractional CMO against an agency, which is a different trade entirely.
Whichever way you split it, the vetting gets harder once you are buying execution rather than a whole department. Fifteen questions worth asking before you sign will tell you more about a prospective partner than their fee will.
Fund the decision from a fixed envelope, because that is how it will actually happen
Almost nobody approves a marketing hire on top of an unchanged budget. The seat gets funded by cutting something, usually media spend, which is the one line that was producing measurable pipeline. So run the hire as a swap and be explicit about it: at $12,210 a month on the assumptions above, you are trading roughly $12,000 of monthly media for one person's output. Write down what that person has to produce for the trade to pay, before you post the role.
Sizing the whole envelope first makes the swap legible. Our piece on marketing budget as a percentage of revenue walks through how owners back into a total number, and it is the right order of operations: total envelope, then allocation, then delivery model.
The practitioner version of this whole exercise takes about twenty minutes. Load one salary properly using the IRS employer rates and your own benefits numbers, divide by twelve, and express it as a percentage of your monthly acquisition budget. Then list the five jobs you need done next quarter and mark which ones need to live inside the company forever and which ones need specialist reps you cannot hire for at that price. The answer usually lands as a hybrid, the split tells you what to buy, and you will have built it from your own arithmetic rather than from someone else's verdict. If you want a second read on where the line should sit for your business, apply here.