In-house, agency or freelancer: what each one is actually good at
Every growing business hits this question sooner or later. Do we hire someone, bring in an agency, or find good freelancer?
The answers online mostly agree with each other, and they are mostly right. Visual Lab, RAKO Media, Creativepool and ThreeSixty all land in roughly the same place: a freelancer buys you specialist depth for the money, an in-house hire buys you context and control, and an agency buys you breadth and a team behind the work.
We agree with most of it. We are also an agency, so we are not neutral, and we buy freelance design and video editing for our own brand. So this piece adds two things the others mostly skip: when an agency is the wrong answer, including us, and what has to stay yours whichever model you pick.
Why this decision matters more than it looks
The invoice is the smallest part of the cost
Every model has a cost that never appears on the quote.
In-house. A digital marketing specialist in Australia typically earns between $80,000 and $95,000 a year (SEEK). Add 12% superannuation (ATO) and you are close to $90,000 to $106,000, before kit, software and staff turnover, none of which shows in the salary line (Visual Lab). Then there is ramp-up: a junior hire can take six to twelve months to get good, learning on your brand as they go (RAKO Media). And the salary stays the same whether you spend $5,000 or $50,000 a month on ads.
Freelancer. The cheapest invoice, until you count your own time. Run a few freelancers at once and someone has to write the briefs, chase deadlines, check quality and make it all look like one brand, and in most small businesses that job falls to the owner (Visual Lab). A cheap option that needs you to plan, brief and chase is not cheap, it is a second job (RAKO Media).
Agency. A higher blended rate, and you share the team’s attention with other clients (Visual Lab). Ask how the fee is calculated, too. If part of it is a percentage of your ad spend, the price rises every time you scale, whether the work changes or not.
The fee only makes sense as a share of your spend
We find the most useful way to compare the three is as a share of your ad spend, not as a dollar figure. At low spend, any management fee has to produce a big improvement just to pay for itself, and the account often does not yet generate enough data for anyone to learn much from it. As spend grows, the same fee becomes a smaller share, and a good operator has room to earn it several times over.
What each one is actually good at
A freelancer is good at one job, done well. You brief the person doing the work, with no account layer in between (Visual Lab). The limits are capacity and continuity: one person can only do so much, and their holidays and busy periods become your problem (Visual Lab). A freelancer is at their best when you have a clear plan and just need it executed (ThreeSixty).
An in-house hire is good at knowing your business. Nobody outside will know it as well as someone inside it, and that knowledge builds over time (Visual Lab). The catch is breadth. One salary rarely covers strategy, production, design and ads at a strong level (Visual Lab). And with one person, holidays, sick days and turnover stop your marketing (ThreeSixty). When that person resigns, what they knew leaves with them.
An agency is good at breadth and pattern recognition. Several disciplines sit under one roof, and the work does not stall when one person goes on leave (Visual Lab). In paid media the bigger advantage is what a team learns across many accounts: spot a platform change or a tracking problem in one, and you can check the rest before it costs them. The trade-off is that you are never the only client. The classic agency failure is being sold by a founder and serviced by a junior, so ask who actually does the work before you sign (RAKO Media), and how many other accounts that person looks after.
When an agency is the wrong answer
Saying this plainly is part of the job.
Your budget is small. If your total marketing budget is a few thousand dollars a month, one excellent freelancer will usually do more with it than a thin slice of an agency (Visual Lab).
You have one clear bottleneck. Hire a specialist for it. A full retainer is an expensive way to fix one problem (Visual Lab).
You need someone in the room every day. That is what an in-house hire is for (RAKO Media).
Nobody can say what a customer is worth. No agency fixes that from inside an ad account. It is the first conversation, not the last.
What people overestimate
That one model wins everything
One of the articles above gives the agency a tick in every row of its comparison table except low upfront cost (ThreeSixty). We are an agency, and we would not trust that table. When a business picks the wrong model, it has usually misread its bottleneck, buying coverage when it needed depth, or headcount when it needed a process (Visual Lab).
That in-house means control
Control only exists if someone can judge the work, and hiring well is hard when you cannot judge the craft yourself (Visual Lab). That is the quiet failure of the in-house route: a capable-looking hire, reporting to someone who cannot check the numbers.
That it is a one-off choice
The setups that work best are usually a mix. An in-house owner holds the strategy and brand judgement, and an external team supplies the specialist skills and the volume (Visual Lab). It is how we run our own brand: the strategy stays with us, the design and editing go to specialists. Creativepool frames the whole decision as operational rather than philosophical: pick the system that will actually ship the work at the right standard (Creativepool).
What has to stay yours, whichever you choose
This is the part none of the articles spend much time on, and in paid media it matters more than the model you pick. Whoever runs your ads, three things should never leave your business.
The accounts and the history. Your ad accounts, pixels and conversion data should sit in accounts you own, with whoever runs them holding access, never the keys. That history is not just a record. Google’s automated bidding needs conversion tracking switched on, and some of its strategies rely on a minimum volume of historical conversion data to work well (Google). Lose it in a handover and the next person starts teaching the system from scratch.
The scoreboard. Every model shares one blind spot: the person running the ads also reports on them. If the only numbers you see come from the platforms, you are reading a score kept by one of the players. Meta’s default window counts a sale up to a day after someone simply saw an ad, and its attribution can only see Meta’s own touchpoints (Measured). So keep one number from your own systems: total revenue from your store or CRM against total ad spend. That is MER, and it does not care who runs the account.
The number that defines success. Someone inside your business has to own what a customer is worth and what you can afford to pay to win one. Without it, a freelancer works to a plan nobody wrote down, an in-house hire guesses, and an agency optimises towards whatever the platform reports.
Keep those three and changing your setup costs you a week. Lose them and it costs you a quarter.
Before you choose
Answer three questions before you speak to anyone. What does a customer need to be worth for your ads to pay? How much of the work is decisions, and how much is production? And if the person running your ads left tomorrow, what would you still have?
The first tells you who has to own the number. The second tells you whether you need a strategist, a pair of hands, or both. The third tells you how exposed you are right now.
Dadek Digital works inside accounts our clients own and judges performance on business numbers like MER, not just what the platforms report, so whatever setup comes next, the data and the history stay with you. If you want an honest read on which model fits your stage, including when the answer is not us, a paid strategy session is the place to start.