Tool
Agency vs in-house cost calculator
What a marketing hire costs once you count everything the salary line leaves out.
Most in-house versus agency comparisons put a salary next to a retainer, which is not a comparison at all. Employer costs, tooling, training, management time, the cost of hiring and the months of reduced output while someone learns your account are all real, and together they usually add half as much again. Put your own numbers in and see the figure the two models should actually be compared on.
The calculator
What the hire actually costs
- Fully-loaded annual cost, amortised over tenure
- $116,125.00
- Per month
- $9,677.08
- Multiple of the base salary
- 1.7x
- First twelve months, with hiring and ramp paid up front
- $131,200.00
Where the money goes
- Salary plus employer costs
- $87,500.00
- Management time, per year
- $10,575.00
- Output paid for and not received during ramp
- $13,125.00
- Hiring and ramp, spread across expected tenure
- $10,050.00
Cost of an hour of their time
- Days of output you get per year, after paid leave
- 235
- Cost per day worked
- $494.15
- Effective hourly cost
- $65.8865
Against the alternatives
Cost is one axis and it is the easiest to measure, which is why it gets treated as the whole decision. Read this next to what each model actually covers rather than instead of it.
| Model | Per month | Per year |
|---|---|---|
| In-house, amortised over tenure | $9,677.08 | $116,125.00 |
| In-house, first twelve months | $10,933.33 | $131,200.00 |
| Retained agency | $4,000.00 | $48,000.00 |
| Freelancer | $2,125.00 | $25,500.00 |
- In-house, compared with the agency fee
- $68,125.00 more than the agency
- In-house, compared with the freelancer
- $90,625.00 more than the freelancer
- Freelance hours a month that cost the same as the hire
- 113.8 hours
- Share of working days the function has no cover
- 9.6%
The number is not the decision
One person is one skill set with no cover, an agency is several with a handover problem, and a freelancer is neither. The full argument, including when in-house is plainly the right answer, is on the agency versus in-house comparison.
Talk it through with usMethod
How is the fully-loaded cost built up?
Salary plus employer costs, tooling, training and management time, with hiring and ramp spread across expected tenure.
Recurring costs are the easy part: salary grossed up by employer taxes and benefits, plus the software seats and data subscriptions the role needs, plus training, plus the hours a senior person spends on direction and review priced at their own loaded rate.
One-off costs are where the comparison usually goes wrong. Recruitment and the output you pay for but do not receive during ramp are both real, and both are paid once. Charging them entirely to year one overstates the ongoing cost; ignoring them understates it. They are spread across your expected tenure here, and the first-year figure is shown separately so both questions get an answer.
Paid leave does not reduce what you pay, only what you get, so the effective hourly cost is the annual figure divided by the days actually worked rather than by a full calendar. That is the number to hold against a freelance rate, because a freelance rate is already net of leave, tooling and management.
Nothing here is a benchmark and nothing is invented. Every figure is arithmetic on numbers you supplied, which means the answer is only as good as your inputs. The one people get wrong most often is management hours, and it is worth measuring for a fortnight before arguing from it. The decision itself is a wider question than cost, and the full case is on the agency versus in-house comparison.
Questions
What do people ask about in-house cost?
Why loaded cost exceeds salary, why tenure matters, what management time counts as, and why freelancers look cheapest.
Why is the fully-loaded cost so much higher than the salary?
Because a salary buys none of the things around it. Employer taxes and benefits, software seats, training, the hours a senior person spends directing the work, the cost of hiring, and the months of reduced output while someone learns your account are all real and none of them appear in the advertised figure.
Why does expected tenure change the answer so much?
Recruitment and ramp are paid once, so their cost per year depends entirely on how long the person stays. The same hire is a reasonable purchase over four years and an expensive one over eighteen months, which is why the tool shows both an amortised figure and a first-year figure.
Is this arguing that agencies are always cheaper?
No. The tool takes your numbers and reports the arithmetic, and plenty of inputs make in-house cheaper. Cost is also the easiest axis to measure and the least interesting one, because it says nothing about coverage, breadth of skill, or how quickly either model can start.
What counts as management time?
The hours somebody senior spends setting direction, reviewing work, answering questions and unblocking. It is genuinely a cost of the hire because it does not exist when the work is bought as a service, and it is the item most often left out of these comparisons.
Why does the freelancer column look cheapest?
Usually because the hours are underestimated. A freelancer at twenty hours a month is buying a fifth of a person, and comparing that against a full-time hire compares two different amounts of work. The tool shows how many freelance hours a month cost the same as the hire so the comparison can be made at equal scope.
Does anything I type get sent to you?
No. Salary figures are sensitive, so the calculator runs entirely in your browser, there is no backend, and nothing is transmitted. The only way your numbers leave your machine is if you copy the shareable link and send it yourself.
Deciding between the two?
The cost gap is rarely what settles it. Tell us what the work actually involves and we will say plainly which model fits, including when that is not us.