Contents
- What a marketing team actually costs
- What an agency or freelancer actually costs
- When building in-house is the better decision
- When buying it in is the better decision
- The hybrid model most companies land on
- A checklist before you decide
- Running the comparison on cost per output
- Contract terms worth insisting on
- Frequently asked questions about marketing team cost
Most companies compare an in-house salary against an agency retainer, see the retainer is higher, and conclude that building internally is cheaper. That comparison is wrong on both sides. Working out the real marketing team cost requires counting everything the headcount drags along with it — and everything the retainer already includes.
This article breaks the cost into its real components, sets out when building beats buying and when it does not, describes the hybrid model most companies actually end up with, and gives a checklist for deciding without relying on instinct.
What a marketing team actually costs
Salary is the visible part and rarely more than two thirds of the total. The rest is real spending that appears in different budget lines and therefore tends to go uncounted in the comparison.
- Employment overhead: statutory contributions, insurance, benefits, leave. Depending on jurisdiction this adds a meaningful percentage on top of gross salary.
- Tooling: analytics, automation, design, SEO and ad management licences, charged per seat. Small teams pay the worst per-person rates.
- Ramp time: a new marketing hire is rarely productive before the second or third month, and that period is paid at full cost.
- Management time: someone senior spends hours each week directing the work. If that person is the founder, it is the most expensive hour in the company.
- Turnover: recruitment, handover and the gap. Marketing tenure is short in most markets, so this is a recurring cost, not a one-off.
A useful rule of thumb: the loaded annual cost of an in-house marketer is meaningfully higher than their salary once all five lines are included. Compare that number against a retainer, not the salary alone.
What an agency or freelancer actually costs
The retainer looks expensive because it is a single visible number. It usually already contains tooling, several specialisms, cover during absence, and no ramp period — items you would otherwise pay for separately.
- What is included: multiple skill sets at partial allocation, licences already paid for, continuity when one person is unavailable.
- What is not: media spend, which is separate and usually larger than the fee itself.
- The hidden cost: your own time briefing, reviewing and correcting. An under-managed agency is expensive regardless of the rate.
- The exit cost: if the knowledge lives with the agency, changing partners means paying twice for the same learning curve.
When building in-house is the better decision
Building wins on a specific pattern, not on a general preference for control. The pattern is: the work is continuous, deeply product-specific, and the accumulated knowledge is itself an asset.
- The work never stops. Daily content, community, lifecycle messaging and customer support overlap poorly with an external retainer.
- Product knowledge is the constraint. If explaining the product takes longer than doing the task, the explanation should only happen once.
- Speed of iteration matters. A change that must go through a briefing cycle is slow at any price.
- Volume passes the break-even point. Once you need a specialism more than roughly half-time, buying it by the hour stops making sense.
When buying it in is the better decision
The inverse pattern is just as clear: work that is intermittent, highly specialised, or that you cannot yet evaluate should be bought rather than hired.
- Project-shaped work: a rebrand, a site migration, a campaign burst. Hiring for a peak leaves you paying through the trough.
- Narrow specialisms: technical SEO, video production, complex ad operations. You cannot keep a specialist busy or current at small scale.
- Skills you cannot assess. Hiring into a discipline nobody internally understands means you will not know for a year whether it worked. A short engagement buys that judgement cheaply.
- Unproven channels. Test externally, bring in-house once it works — not the other way round.
The hybrid model most companies land on
In practice the question is rarely all-or-nothing. The stable arrangement is a small internal core that owns direction and continuity, with execution capacity bought around it.
- Keep in-house: strategy, brand and messaging ownership, the customer relationship, and measurement. These are the things that must not walk out of the door.
- Buy in: production volume, specialised technical work, and anything seasonal.
- Hold the data internally. Analytics accounts, ad accounts and the domain stay in your name regardless of who operates them.
- Require documentation as a deliverable. What was tried, what worked and why, written down where you can read it after the engagement ends.
Which functions belong in that internal core, and in what order to add them, is a structural question rather than a purely financial one — see our guide to marketing team structure.
A checklist before you decide
Run these five questions before committing either way. Most bad decisions here come from answering only the first one.
- Is this work continuous or project-shaped over the next twelve months?
- Can anyone internally evaluate whether it is being done well?
- What is the fully loaded internal cost, not the salary?
- If this person or partner left in six months, what would remain with us?
- Who will manage the work, and do they have the hours?
The last question sinks more arrangements than the budget does. Both models fail the same way when nobody owns direction, and no amount of spending fixes an unmanaged function.
Running the comparison on cost per output
The abstract build-versus-buy debate resolves quickly once you convert both options into a cost per unit of work. The arithmetic is simple; the discipline is in counting the same things on both sides.
- Define the unit. Published articles per month, campaigns launched per quarter, emails shipped — whatever the role actually produces. Vague units produce vague comparisons.
- Divide loaded cost by realistic throughput. Not theoretical capacity. A full-time marketer does not produce forty hours of output, and assuming otherwise flatters the in-house side.
- Count management hours on both sides. They exist either way, and they are frequently higher for the internal option than people expect.
- Add a quality adjustment. If one option needs a revision round the other does not, the effective cost differs from the nominal one.
This calculation regularly reverses an intuition. Buying is often cheaper per unit at low volume and more expensive above a threshold — and knowing roughly where that threshold sits tells you when to revisit the decision, rather than leaving it fixed for years.
Contract terms worth insisting on
When you buy rather than build, most of the risk is contractual rather than financial. These terms cost nothing to negotiate at the start and are effectively unobtainable once a relationship is underway.
- Accounts in your name. Ad accounts, analytics properties, the domain and the CMS. An agency operating your accounts is normal; owning them is not.
- Named allocation. Who works on this, and for how many hours. Without it you are buying a firm’s capacity, not its people, and the seniority you met at the pitch may not appear again.
- A defined exit. Notice period, handover scope, and what documentation transfers. Assume the relationship ends, because eventually it will.
- Reporting on your definitions. The metrics you agreed, at an agreed frequency — not the platform’s default dashboard, which is designed to make the platform look effective.
- Case study permission, both ways. Agree in advance what each side may publish about the work. Cheap now, contentious later.
The first and third points do most of the work. A partnership that ends with your data and your accounts intact is a change of supplier; one that ends without them is a restart, and the restart is usually more expensive than any fee you negotiated down.
Frequently asked questions about marketing team cost
What share of revenue should go to marketing?
Benchmarks vary widely by sector and growth stage, and copying one is usually worse than reasoning from your own numbers. Work backwards instead: from the target, the current conversion rate and the acceptable acquisition cost. A benchmark tells you what other companies spend, not what your funnel needs.
Is one generalist better than two junior specialists?
At small scale, yes, in almost every case. A capable generalist can identify which specialism you actually need, which is exactly the judgement that is missing when a small company hires two juniors and has nobody to direct them.
How long before a marketing hire pays for themselves?
Plan for two to three quarters in most B2B and considered-purchase contexts, faster in high-velocity e-commerce. If the plan assumes payback inside one quarter, it is usually a plan for paid acquisition rather than for a marketing function — and those are different purchases.
You can find additional benchmarking data at Gartner Marketing.
You can find more hands-on marketing guides at https://marketing365.vn.
Content from marketing365 is created for SMEs, online shop owners, and new marketers.
You can also read more articles on the same topic in the Analysis Corner category.
