I run an agency, so weigh this accordingly. I've also watched businesses pick each of these three and get burned by each of them, and the reasons are more predictable than the options.
What you're actually buying
Paid social for a local business needs, at minimum:
- Media buying — structure, budgets, bidding, testing
- Creative — the images and video, which is where most results come from
- Tracking — pixel, server-side conversions, and a booking signal sent back
- Landing pages — somewhere the click goes that isn't your homepage
- Reporting — cost per customer, not cost per click
Add SEO and a website and that's seven skills. The comparison is really about how many of those seven each option gives you.
In-house hire
What you get: one person, fully focused on you, who learns your business deeply.
What it costs: a salary, plus payroll taxes, benefits, software and the time you spend managing them. The total is always meaningfully above the salary.
Where it breaks: one person rarely has all seven skills. A strong media buyer is often a weak designer; a good designer rarely sets up server-side tracking. You end up paying for a person and still outsourcing half the work. And when they leave, the knowledge walks out with them.
When it's right: once your marketing is large enough to keep a specialist busy, and important enough that you want it inside the building.
Freelancer
What you get: usually one strong skill at a good price.
Where it breaks: single point of failure. Vacations, illness, a better client — any of them stops your marketing. And a freelancer buying media usually isn't building your creative or fixing your tracking.
When it's right: a specific, bounded job — a landing page, a creative batch, a tracking audit — where one skill is exactly what you need.
Agency
What you get: the full set of skills, spread across clients so you're not paying for all of them full time, and continuity when someone is out.
Where it breaks: you can get the B-team. Junior account managers, a templated strategy, and reporting that shows impressions rather than customers. Some agencies take a percentage of your ad spend, which pays them more when you spend more — whether or not it's working.
When it's right: when you need several of the seven skills and don't have the volume to employ them.
The best version of each option is fine. The question is which option's worst version you can live with.
The questions that decide it
| In-house | Freelancer | Agency | |
|---|---|---|---|
| Skills covered | Usually 1–3 | Usually 1 | Usually most |
| Continuity if someone leaves | Poor | Poor | Good |
| Depth on your business | Best | Varies | Varies |
| Management time from you | High | Medium | Low |
| Who owns the ad account | You | Should be you | Should be you — ask |
That last row matters in every column. Whoever you pick, you should own the ad account, the pixel and the domain. If you don't, you can't leave without starting over.
How to tell a good one from a bad one
The same test applies to all three: ask for cost per customer rather than cost per lead, and ask what they'd turn off. We wrote out nine questions that separate real expertise from a good pitch, and they work just as well in an interview for a hire.
If you're weighing a hire against an agency, bring the salary you had in mind to a call. We'll lay the two side by side for your business specifically — including the cases where hiring is the better answer.