Every agency has a page arguing that hiring is a false economy, and every recruiter has one arguing the opposite, and both are written by people whose income depends on the conclusion. The genuine answer is conditional and the conditions are knowable. It comes down to three things: how much continuous work you actually have, how many of the ten disciplines you need covered at once, and how much of your marketing depends on knowing your business rather than knowing a platform. Get those three straight and the decision usually makes itself, sometimes against us.
What a hire actually costs
Most in-house-versus-agency comparisons fail at the first line, because they put a base salary next to a retainer. A salary is not what an employee costs. The comparable number is the loaded, all-in, first-year cost, and it is built from six components rather than one.
| Line item | How to price it for your own business |
|---|---|
| Base salary | Use your own market, not a national figure. The US Bureau of Labor Statistics publishes occupational wages by metropolitan area in its Occupational Employment and Wage Statistics release; look up the role and the city you are actually hiring in. |
| Employer taxes and benefits | Payroll taxes, insurance, retirement contributions, paid leave. The Bureau of Labor Statistics also publishes Employer Costs for Employee Compensation, which gives the benefits share of total compensation directly. Use that rather than a rule of thumb you read somewhere. |
| Tools this person needs | Rank tracking, a landing page builder, call tracking, creative software, an email platform, a reporting layer. Count only what the business does not already own, and count it annually. |
| Recruitment | Agency fee or job board spend, plus the hours your team spends on screening and interviews. This is a real first-year cost even when no cheque is written. |
| Ramp | The months between the start date and full productivity, priced at the loaded monthly cost. For a marketing generalist joining a business with existing accounts, plan on a quarter, more if the tracking is a mess when they arrive. |
| Management | Someone has to set the priorities, review the work and unblock it. If nobody in the business can judge the quality of this person's output, that is not a cost line, it is a reason to reconsider. |
We have deliberately not put a headline number on that table. Any figure we printed would be a national average dressed up as your situation, and the two multipliers that matter most — the local wage for the role and the benefits share of total compensation — are both published by a named public source you can check in ten minutes. Do that, and the comparison stops being an argument and becomes arithmetic.
For the other side of the comparison, here is ours, annualised, so it is on the same basis rather than a monthly figure sitting next to a yearly one.
| Package | Scope | Per month | Per year |
|---|---|---|---|
| Starter | One or two disciplines | $3,500 | $42,000 |
| Growth | Three or four disciplines | $7,500 | $90,000 |
| Scale | Five to seven disciplines | $12,500 | $150,000 |
| Dominate | Eight to all ten disciplines | $19,500 | $234,000 |
Put the two tables next to each other and the point of the exercise appears immediately: for a lot of businesses, one loaded senior hire and one of the middle packages land in genuinely similar territory. That is why the decision cannot be made on price. It has to be made on coverage.
What one person can and cannot carry
The ten disciplines are not ten flavours of the same job. Deliverability authentication has more in common with systems administration than with copywriting. Tag management is closer to software engineering than to media buying. Creative at testing volume is a production problem, not a skill problem. So "can one person do it" is really two questions — is it the same profession, and is there enough time — and the honest answer differs per row.
| Discipline | Realistic for one strong generalist? |
|---|---|
| Paid media | Yes, on one or two platforms with a stable account. No, across five platforms with weekly creative rotation and pacing. |
| SEO | Partly. On-page work and content, yes. Technical audits, migrations and earned coverage usually need a specialist. |
| AI search, AEO and GEO | Rarely. It changes monthly, it sits on top of technical SEO, and there is no settled playbook to hand someone. |
| Local and Maps | Yes, and this is the one to own internally. It is learnable, it is local knowledge, and it rewards persistence over expertise. |
| Analytics and tracking | Almost never. Tag management, server side collection, attribution and consent are a separate profession from marketing. |
| Email and SMS | Yes for campaigns and lifecycle flows. Authentication and deliverability is a different skill and fails silently. |
| CRO and landing pages | Partly. Building and shipping pages, yes. Designing an experiment that can actually be read, rarely. |
| Creative studio | Volume is the constraint, not talent. One person cannot produce a fresh testing batch every week for long. |
| Reputation | Yes. It is mostly process and follow-through, and it is genuinely better done by someone inside the business. |
| Web design and speed | Maintenance yes, a full build no. A rebuild is a project with a start and an end, which is the wrong shape for a salary. |
Count the yeses. A strong generalist can genuinely own three or four of those to a standard you would be happy with, and can hold a watching brief on two or three more. That is a real and valuable thing to have on the payroll. It is not ten. The failure that follows from pretending otherwise is not dramatic — nobody quits, nothing breaks — it is that everything gets done to about half a standard, and the deficit stays invisible for the best part of a year because nothing on a dashboard reports it.
A generalist hired for a ten-discipline problem does not fail. They do all ten at fifty percent, and nothing tells you.
When hiring is the right answer
Three conditions, and where all three hold, hire. Not "consider hiring". Hire.
- 01 Demand is steady and concentrated. One or two channels produce most of your work, the seasonality is predictable, and the account does not need reinventing every quarter. Steady single-channel work is the case an employee is built for and the case an agency is worst value in.
- 02 The marketing depends on product knowledge. If credible copy requires knowing how the thing is actually installed, specified, regulated or serviced, an employee who sits near the people who do it will beat an outside team indefinitely. No amount of onboarding closes that gap.
- 03 There is enough work to keep them busy. This is the one people skip. A full-time hire with three days of real work a week is not a bargain, it is an idle asset, and the usual response — inventing work to justify the salary — is how businesses end up with a podcast nobody asked for.
Where those three hold, the employee also wins on things that never appear in a cost model. They are in the room when a decision gets made. They hear the sales calls. They know which job types are profitable and which ones the crew hates. They can turn a customer complaint into a landing page revision the same afternoon. An agency can be told all of that and will still be told it second-hand, a week later, in a summary. If you can hire a strong operator and keep them genuinely busy, hire them. We would rather write that sentence than have somebody discover it after a year of paying us.
What to actually look for in that hire
If you have decided to hire, the shortlist criteria that predict success are not the ones on most job adverts. Platform certifications tell you somebody sat an exam that the platform wrote and marked. What tells you more is whether the candidate can describe, unprompted, a campaign they switched off and why, because knowing when to stop is the rarest skill in this job and the one that most directly protects your money. Ask them to walk you through a month where the numbers went the wrong way and listen for whether they diagnosed it or narrated it.
Second, hire for the disciplines you actually need rather than for a generic title. "Marketing manager" is four different jobs depending on the business, and a candidate who is excellent at lifecycle email and mediocre at paid media is the right hire or the wrong one entirely depending on which of those you are short of. Write the three or four disciplines into the advert and screen against them.
Third, decide up front who judges the work. A marketing hire reporting to somebody who cannot tell good work from confident work will drift toward whichever activity is easiest to present, and that drift is not a character flaw — it is what happens to anyone whose output is only ever assessed on how it was described. If nobody in the business can make that judgement, either buy that judgement in alongside the hire or accept that you are trusting the first year entirely.
When hiring goes wrong
Two failure modes, both common, both slow. The first is the coverage gap described above: one generalist, ten disciplines, everything at half a standard. The second is the single point of failure. One person holds the accounts, the logins, the vendor relationships, the reasoning behind every past decision and the only working understanding of how the tracking is wired. That is a fine arrangement right up until they resign, go on leave, or get poached, at which point the marketing stops and the recovery takes a quarter. The mitigation is not distrust, it is documentation and shared ownership of accounts, and almost nobody does it until after the first time.
There is a third, quieter one. A single in-house marketer sees exactly one account — yours. An operator working across many accounts sees a failure mode in somebody else's business three months before it reaches yours, which is a genuine advantage of the outside model and one of the few that is hard to replicate. It is also the reason an isolated in-house hire benefits enormously from a budget for conferences, communities and peer contact, which is a line item most businesses cut first and should cut last.
The hybrid, and how to keep it honest
The arrangement that works for most businesses past the first hire is not either column. It is an internal owner plus outside specialists, with the ownership sitting inside. The employee holds the brand, the offer, the customer knowledge, the CRM and the relationships, and runs the disciplines that reward proximity to the business — typically local and maps, reputation, and the campaign side of email. The outside team takes the ones that need either specialist depth or production volume: tracking and attribution, technical SEO, creative at testing volume, and deliverability.
Two rules keep it from degenerating. First, the internal person owns the outcome and the accounts — every ad account, analytics property, pixel and audience in the company's name, with the agency as a user rather than an owner. Second, the agency reports to them rather than around them. A specialist team that reports past the internal owner to the founder will, without anyone intending it, start optimising for the founder's attention. The hybrid works when there is exactly one person accountable and they are on your payroll.
The decision, in four questions
How many disciplines do you actually need covered in the next twelve months? Is there enough continuous work in them to fill a week, every week? Does the work depend more on knowing your business or on knowing a platform? And is there anyone in the building who can judge whether the work is good? If the honest answers are "three or fewer", "yes", "my business" and "yes", hire, and do not let anyone talk you out of it. If they are "six or more", "not yet", "the platforms" and "no", an outside team is the cheaper way to get coverage. Anything in between is the hybrid.
One disclosure, because it bears on how much weight to give any of this. CurrentAds has no paying clients yet. The one case study published on this site is an engagement with a company under common ownership with CurrentAds, published anonymised for that reason; no revenue increase is attributable to the work and none is claimed. So this piece is not evidence that hiring us beat hiring somebody. It is the framework we would use ourselves, and our prices are published precisely so you can run the comparison without a sales call. The terms are on the guarantee page, and the full discipline list is there to be cut down rather than bought whole.
Two pieces sit either side of this one. If the salary column is settled and the question is now what the outside number should look like, the four models an agency can charge on, and what each of them quietly rewards, are pulled apart in what an agency actually costs. And if you get as far as a contract, the clauses that decide whether leaving is cheap or expensive are listed in the contract clauses worth reading first, with the fair version of each written next to it. If you would rather see the two columns totalled against each other rather than argued at, that is what the side by side cost comparison does.