In house vs agency marketing: the honest two-year comparison

Aug 18
3 min read

Updated: 4 days ago

Introduction

The usual version of this comparison sets a salary against a monthly fee and declares a winner. That calculation is wrong in both directions: it understates the cost of hiring and ignores what each option leaves you holding after two years.

Here is the comparison with the missing parts restored.

1. What a hire actually costs

Salary is roughly two thirds of the real figure. Add employer contributions, equipment, software licences, recruitment, and the management time of whoever the person reports to.

Then add ramp. A marketing hire is rarely fully productive for several months, and during that period you are paying full cost for partial output.

Then add concentration risk. One person means one skill set — someone strong at paid media is rarely also strong at content, analytics and pricing — and their departure takes the capability with them.

2. What an agency actually costs

The monthly fee is more of the total cost than a salary is, but not all of it. Add your own time briefing and reviewing, and any tools retained in your name.

The larger consideration is cumulative. An agency fee does not end, and unlike a salary it does not build an asset inside your business. Two years of fees can exceed the fully loaded cost of a hire while leaving you with less retained capability.

Against that, an agency starts producing quickly, covers several disciplines at once, and can be stopped.

3. In house vs agency marketing over two years

Over one quarter the agency almost always wins on cost and speed. Over two years the comparison usually inverts, provided the hire was the right one and stayed.

The determining factor is rarely price. It is whether the workload genuinely sustains a full-time role. A hire with insufficient work becomes expensive quickly, and the role tends to expand into activity that fills time rather than produces customers.

Be honest about volume before deciding. Most owner-operated businesses do not have enough continuous marketing work to justify a full-time salary, which is the actual answer to the question they are asking.

4. What each leaves behind

This is the part omitted from most comparisons, and it usually matters more than the monthly figure.

A hire builds knowledge inside your business — provided it is written down. If it lives only in their head, their resignation resets you to zero.

An agency builds knowledge inside the agency by default. Whether any of it transfers depends entirely on how the engagement was structured: whose accounts the work sits in, whether documentation was produced, whether anyone on your side was trained.

Neither option retains capability automatically. Both can, if you require it.

5. The hybrid most small businesses end up needing

The common resolution is not either option in pure form.

Buy the building of the system as a defined project — tracking, pricing structure, campaign architecture, reporting — from whoever has that expertise. Insist it is built in your accounts, documented, and handed over with training.

Then run it internally with existing staff, since operating a working system is far less demanding than designing one. Bring in specialists for specific production work as needed.

This gets you expertise where it is scarce and avoids paying indefinitely for work your own people can do.

6. When to genuinely hire

Hire when the operating workload reliably exceeds what existing staff can absorb, when judgement work has grown beyond what the owner can cover, or when you need a skill continuously rather than occasionally.

By that point you will also hire well, because you will know precisely what the role does — you will have been doing it. Hiring before that usually means recruiting someone to invent a process, which is a harder and more expensive job than running one.

Conclusion

Compare fully loaded costs over two years rather than salary against fee, and ask what each option leaves you holding at the end.

For most small businesses the answer is neither pure option: buy the build, own the output, train your own people to operate it, and hire only when the workload genuinely justifies a permanent role. Whatever you choose, require documentation and account ownership — that is what determines whether you are building capability or renting it.