Build vs buy marketing capability: a decision per function

Aug 22
3 min read

Updated: 3 days ago

Introduction

The question usually arrives as a binary. Do we hire someone, or do we outsource it?

Framed that way it has no good answer, because marketing is not one job. It is six or seven different jobs with different frequencies, different skill requirements and different consequences if they stop.

1. Build vs buy marketing capability is a decision per function, not per department

Split the work before deciding anything: strategy and prioritisation, channel setup, ongoing channel operation, creative production, measurement and reporting, customer follow-up, and documentation.

Each of those has a different answer. Setting up a measurement layer is specialist, technical and happens once — a poor candidate for hiring. Answering enquiries within an hour is constant, context-dependent and impossible to outsource well.

Treating the whole set as one decision is why both answers feel wrong.

2. Frequency decides more than skill level

The most useful test is how often the task recurs.

Anything that happens daily or weekly and depends on knowing your customers belongs inside the business. Anything that happens once, or once a year, and requires expertise you would otherwise have to acquire from scratch is worth buying.

Hiring for a one-off build means paying to develop a skill you will use once. Outsourcing a daily routine means paying a premium for someone who will always know your business less well than you do.

3. Ask whether the output is a thing or a service

A useful distinction: some engagements leave you with an asset, others leave you with a month of activity.

A build produces something that persists — a working measurement layer, documented procedures, a functioning follow-up sequence. When it ends, you still have it.

A retainer produces activity. When it ends, the activity stops and nothing remains except whatever was learned informally. Neither is wrong, but they are not substitutes and should not be priced against each other.

4. Count the true cost of building

A salary is the smallest part of the internal option.

Add recruitment time, the months before someone is productive, management attention, tooling, and the risk that the person you hire has one channel's experience and you need four. Add the cost of the role sitting empty for three months when they leave.

For most small businesses the honest comparison is not salary against fee. It is salary plus overhead plus ramp against a fixed-scope project with a defined end.

5. Count the true cost of buying

The external option has its own hidden line items, and the largest is dependency.

If the accounts, data, creative files and reasoning live with the supplier, the real cost includes what it would take to leave. A cheap monthly fee attached to assets you do not own is expensive.

Check three things before signing: who holds the accounts, who owns the data, and what you are left with if it ends.

6. Never outsource the judgement layer

Whoever decides which lever to pull needs to understand the business, and that is not transferable.

You can buy execution, technical setup, creative production and analysis. Deciding what matters this quarter and what to stop requires knowing your margins, your capacity and your appetite for risk.

Businesses that outsource judgement end up funding whatever their supplier is good at, which is a different thing from what they needed.

7. The hybrid that usually works

For most owner-operated businesses the pattern is: buy the build, own the operation.

Bring in expertise to construct the system and train the team, then run it internally with documented procedures. Specialist skill is applied where it is scarce and expensive; daily operation stays with the people who know the customers.

This also converts a recurring cost into a one-off one, and leaves the capability inside the business rather than inside a supplier.

8. Write the exit into either choice

Both options need a defined ending.

For a hire, that means documented procedures so the role survives the person. For a supplier, it means account ownership, data export and a handover document as contractual conditions rather than goodwill.

If neither choice has an exit, you have not decided between build and buy — you have created a single point of failure and given it a different job title.

Conclusion

Split marketing into functions and decide each separately. Frequent, context-heavy work belongs inside; one-off technical builds are worth buying.

Compare true costs on both sides, keep the judgement layer internal, favour buying the build and owning the operation, and make sure whichever you choose leaves you holding the accounts, the data and the documentation.