It's not two options, it's four
Framed properly, you're choosing what to buy: an employee buys you dedicated attention and company knowledge, with a slow ramp and a fixed cost that doesn't pause in quiet months. An agency buys you a managed bundle of specialists, with less context and variable seniority behind the account manager. A freelancer buys you one strong capability, with you supplying the direction. And "nobody yet" buys you time and kept money, which for plenty of businesses is genuinely the right purchase, as the full hiring guide argues at length.
Most published comparisons only argue the first two, and are written by whichever side is selling. So let's do the numbers nobody shows.
The real maths: cost per delivered hour
Salaries and retainers can't be compared directly; hours of marketing actually delivered can. A UK marketing manager on £48,000 costs about £55,900 the moment you add employer National Insurance (15% above £5,000 since the 2025 changes) and minimum pension, and a realistic £67,000 all-in once recruitment, tools, training and cover are counted. Across roughly 1,650 productive hours a year, that's about £40 per delivered hour.
| Option | Headline cost | Cost per delivered hour |
|---|---|---|
| In-house manager | ~£67,000/yr all-in | ~£40 |
| Freelancer (median) | £475–£500/day | ~£60 |
| Fractional CMO | £700–£1,800/day | £90–£225 |
| Agency retainer (mid-tier) | £2,500–£8,000/mo | £60–£150+ effective |
The agency number needs explaining, because it surprises people: divide a mid-tier retainer by the 15 to 40 hours of monthly delivery time typical at that level and the effective rate lands at £60 to £150+ an hour. The "agencies share overheads so they're efficient" pitch doesn't survive that division; what you're actually paying for is the bench (multiple specialists, cover, accountability), and where the rest of the fee goes is a story in itself.
So on pure arithmetic the employee wins. Which is exactly why the arithmetic isn't the decision.
What the salary comparison misses
Three costs never appear in the spreadsheet, and I've watched all three up close. I started my working life as a recruitment resourcer, filling exactly these vacancies, so believe me when I say the official numbers below are the optimistic end.
- Time. UK marketing roles take 40 to 51 days to fill, a new digital-marketing hire needs 3 to 6 months to reach normal capacity, and Oxford Economics puts average time to optimum productivity at 28 weeks. Stack those and the honest planning number is 5 to 9 months from opening the vacancy to a fully productive marketing function.
- Tenure. Marketing has the lowest median tenure of any industry in the UK retention data: 2.8 years overall, just 2.7 in small businesses. Your hire is a two-to-three-year chapter, not a decade-long institution, so you may pay the ramp cost more than once.
- The bad-hire risk. The most-cited UK estimate puts a failed manager-level hire around £132,000 once lost productivity, management time and team impact are counted (one REC/Oxford Economics estimate, so an order of magnitude, not gospel). Even the conservative breakdown, roughly £9,700 recruitment and training plus £9,600 lost productivity, stings a small firm.
Before agencies look smug about any of this: UK agency staff turnover hit 24.8% last year, 27.6% at creative agencies.
"Choosing an agency doesn't remove people-risk. It moves it onto their payroll."
Which is why "who exactly works on our account, and what happens when they leave?" belongs in every agency conversation.
The fill-the-seat test (and the one-channel rule)
In-house wins the maths only if there's enough consistent, useful marketing work to fill 1,650 hours a year. So here's the capacity truth nobody puts in a job ad: one marketer can run one channel well. Maybe two, if one is low-effort or heavily repurposed. At three, quality visibly slips. That rule shows up in every honest framework, and it's why the classic small-business job ad (SEO, PPC, social, email, content, design and events, one salary) is an advert for a department, not a person.
It's also why so many first hires end unhappily. The Marketing Week research on solo marketers found 80% experiencing imposter syndrome and 60% feeling overwhelmed, while 80% of CEOs say they lack confidence in the marketing work. That's not a talent problem on either side; it's what happens when one human is assigned a department's workload and then judged against it.
I live this rule myself, for what it's worth. This site and its daily letters are a one-person operation built in the evenings, and it only works because I run it exactly as above: one channel done properly (the email list), one experiment at a time, and a hard no to everything else however tempting. The moment I've tried to do three things at once, all three got worse.
So the test: write down the marketing work you actually need monthly, in hours, honestly. Enough to fill a focused role around one or two channels? In-house maths works. A patchwork of specialist tasks across five disciplines? You're describing a hybrid, not a hire.
If you do hire: generalist first, and measure month three fairly
The evidence on first hires is unusually one-sided: hire a commercially minded generalist, not a channel specialist, unless one channel already demonstrably produces customers and simply needs more hands. The decision rule is one sentence: if you can't name your proven channel and its economics, you need diagnosis before depth, and diagnosis is generalist work. A specialist without a strategy produces channel activity that connects to nothing.
And agree what "good" looks like at 90 days before they start, because it isn't revenue. A fair first-quarter scorecard: a clear picture of your ideal customer, working tracking and CRM handoff, an audit of what exists, a prioritised channel plan, and the first tests live. Measure a capable person against "transform revenue in month one" and you'll fire them just before the work compounds.
The middle options: freelancer and fractional
A freelancer at the £475 to £500 median day rate is the best value in the market for a defined channel or project: near-employee cost per hour, no employment risk, senior craft. The catch is that you supply the strategy, coordination and quality control, so a freelancer suits a business that knows what it needs, not one still finding out.
Fractional marketing leadership (a senior head, one or two days a week) has settled into a remarkably consistent UK band: £700 to £1,800 a day, £3,000 to £10,000 a month. It buys the judgement layer (prioritisation, supplier oversight, measurement design) without a director's salary. Two honest cautions: below about £3,000 a month you're usually buying manager-grade support wearing the CMO label, and too few days produces advice with no delivery momentum. It works best sitting above juniors or suppliers who do the hands-on work.
Where most businesses actually land: hybrid
Follow the logic to its end and you get the model most growing firms eventually reach, usually after learning the hard way that one person can't be a department: one accountable in-house owner, plus bought-in specialists. The split that works:
| Keep in-house (the thinking) | Buy in (the depth) |
|---|---|
| Strategy and priorities | Paid media management |
| Positioning and brand voice | Technical SEO |
| Budget and supplier briefs | Design and video |
| Quality control and approvals | PR and outreach |
| Deciding what the numbers mean | Development work |
The inside person owns the left column because it depends on knowledge no outsider accumulates. The right column is where currency and depth beat context.
That split also fixes the most common supplier failure mode, and I say this having spent twelve years on the selling side of these relationships: an unbriefed, unmanaged supplier drifts toward whatever is easiest to report, every time, not out of malice but because nothing is pulling them anywhere better. An agency with a strong internal owner on the other side is a different, far better product than an agency handed a credit card and hope.
The five-minute decision
- No proven channel yet, or budget under ~£500/month? Hire nobody. Owner-led marketing plus the free stuff on this site until something works. And take comfort: I asked thousands of owners "do you do any marketing?" across twelve years of sales calls, and nine out of ten said no or described something abandoned. The baseline you're competing with is lower than you think.
- One clear channel or project, strategy known? Freelancer. Best value per hour in the market.
- Enough work for one focused role, and someone senior to point it? In-house generalist, judged on a fair 90-day scorecard and given 5 to 9 months honestly.
- Plenty of budget for delivery but nobody senior to direct it? Fractional lead first, delivery second.
- Proven channels needing scale across several specialisms? Agency, with itemised billing, named staff and an internal owner managing them.
And whichever door you pick, run the numbers on what it must produce to pay for itself before you sign anything. The Pipeline Calculator does that arithmetic in about a minute.
Figures are 2026 UK benchmarks from named sources: GOV.UK employer National Insurance rates, ITJobsWatch contractor data, Vestd's UK retention and FTSE tenure analyses, the IPA Agency Census 2025, REC/Oxford Economics failed-hire estimates, Oxford Economics productivity research, converged fractional-CMO rate data and Marketing Week's solo-marketer survey. Where a figure is a single estimate or self-reported, the text says so.