Should You Hire an Agency, a Freelancer, or a First Marketer?

The budget gets approved and then a decision starts that nobody framed properly. Somebody says agency. Somebody else says we should just hire someone. A third person knows a freelancer who is genuinely very good. The argument runs for six weeks, and it runs that long because all three people are describing different purchases and none of them has said which one is being made.

Here is the framing we use. Hiring marketing help is a decision about cost shape, breadth, continuity and risk. It is not a decision about quality. There are excellent agencies, excellent freelancers and excellent first hires, and in each case the way the arrangement usually fails has nothing to do with talent.

So the comparison comes early, as a table. Most of the length after that goes on the parts a table cannot hold: where an agency is the wrong answer, what a first marketer needs in order not to fail, the hybrid most companies arrive at eventually, and how to brief any of the three without asking anybody to work for free.

Three purchases hide behind one budget line

Before comparing options, name what you are short of. There are only three answers, and they point in different directions.

Capacity means you know what needs doing and cannot get through it. Capability means you do not know what needs doing, or you cannot tell whether it is being done well, which is the more dangerous version of the same gap. Continuity means the work is permanent, depends on internal context, and needs somebody whose name is on it a year from now.

Capacity is a freelancer. Capability is an agency or a senior hire. Continuity is in-house. The uncomfortable part, and the reason these conversations stall, is that most companies need two of the three and have budget for one. Deciding which gap to leave open is the actual decision, and it should be made deliberately rather than discovered in month four. That is the same reasoning behind our own approach: define before you design, because scope drift is cheaper to prevent than to unwind.

The comparison, on the dimensions that decide it

Ignore capability claims for a moment. Everyone claims capability. These are the structural differences, and they hold regardless of who you end up talking to.

Read the last two rows first. The failure modes are more predictive than the strengths, because every option looks fine in the good months and the arrangement is judged on the bad ones.

Where a freelancer is plainly the right answer

When you can write the brief, a freelancer is usually the best value available. You are buying hours from somebody who does one thing repeatedly, without the overhead of an account layer you do not need. Design systems, technical implementation, a defined content batch, a specific piece of analytics work: all of these have edges, and things with edges are cheap to buy well.

The capacity trap, and how to defuse it

The failure is reliable enough to plan around. Good freelancers acquire bigger clients, and the month they do is the month you needed them most. Two habits reduce the damage. Scope deliverables rather than availability, so the contract describes work rather than presence. And never let one freelancer hold the only copy of anything: accounts, tracking configuration, ad credentials and documentation belong to you, in your systems, from the first week.

The second failure is subtler. A freelancer executes the brief you wrote, which means a wrong brief gets built beautifully. If you ask for conversion work on a page nobody visits, you will get conversion work on a page nobody visits, and the problem of optimising without traffic to test will still be there afterwards, having cost you a month.

Where an agency is the wrong answer

We are an agency, so this section is the one worth reading carefully. There are situations where hiring us, or anyone like us, is a mistake we can see coming.

The first is when nobody internally can approve work. An agency amplifies whatever decision-making already exists. Where none exists, you are buying delay at a monthly rate, and the relationship sours in the third month when everyone realises the bottleneck was never production. Someone on your side has to be able to say yes on a Tuesday without convening anybody.

The second is when the work is one narrow, continuous, well-understood task. Paying an agency for breadth you never touch is a slow leak. The third is when the real problem sits in the product, the pricing or the offer. Marketing can describe an offer accurately and it can find the people who want one, but it cannot make a proposition compelling that nobody has asked for, and an honest agency will say so in the first meeting rather than the fourth invoice.

The fourth is worth naming because it is so common: what looks like a demand problem is often a response problem. Enquiries do arrive, and then they wait, or they arrive as a form submission from somebody who had three questions the site never answered. That is not a campaign brief. It is a handling brief, and a lightweight widget such as Flidu addresses it directly by answering from what the site already says and putting contact and conversion actions in the same place. Fix that first and the demand number changes without anybody buying media.

What a first marketer actually is

The most common way a first marketing hire fails is that the job description contained three jobs. Strategy, execution across five channels, design, copywriting, events, and reporting to a board. That person exists in the market roughly never, and the version you hire will be strong in one area and quietly drowning in the others while nobody notices for two quarters.

Hire for a shape instead: deep in the one discipline that carries most of your demand, competent enough elsewhere to brief and judge specialists. The word for that is unfashionable, so ignore the word and hold the substance. What you are buying is judgement plus one craft, and the craft should be the one you would otherwise buy most often from outside.

What they need on day one

• A named approver above them. One person who can say yes, and who has agreed in advance to be interrupted for it.
• Direct access to the systems. Analytics, the CMS, the ad accounts, the email platform. Access requested through somebody else is a role defined as waiting.
• A working budget separate from their salary. A marketer with no budget is a writer with a job title, and both of you will be disappointed by the arrangement.
• A first quarter that is written down and small. Two or three outcomes, agreed before they start, so the first review discusses evidence rather than impressions.
• Somebody outside to check their thinking. A first marketer has no peers internally. Without an external counterpart they are reviewing their own homework indefinitely.

The hybrid most companies arrive at anyway

One internal owner, plus outside specialists for the disciplines that are episodic or need depth you cannot staff. This is where most companies end up after eighteen months, and arriving at it deliberately is cheaper than arriving at it after two failed arrangements.

The division that works: in-house owns the decisions, the relationships, the context and the written record. Outside owns depth, production capacity and the discipline of reading platform release notes on your behalf. That last part is undersold and it is real: staying current with search and platform changes is somebody’s full-time habit, and paying for it is more sensible than pretending you will keep up. A digital growth engine is not a team structure, but it does describe the working arrangement this division produces.

The hybrid fails in one specific way, and it is easy to spot early. The internal person becomes a message relay: forwarding requests, chasing statuses, approving nothing. When that happens the company is paying a salary for coordination it could have bought as part of a retainer, and the outside team is working without the internal context that made the arrangement worth having. The fix is to give the internal owner authority over the measurement and strategy, because whoever owns the numbers owns the decisions.

How to brief any of them without asking for free work

Free pitches are a bad instrument and they persist because they feel like diligence. A speculative audit or a mock homepage tells you how well somebody sells with no information. It tells you nothing about how they behave in month seven, which is the only thing you needed to know. Run this instead.

1. Write the problem, not the deliverable. "Enquiries dropped after the site change" is a brief. "We need a content strategy and three landing pages" is a purchase order with a guess inside it.
2. State the constraints honestly. Budget shape, timeline, what cannot change, who approves. Suppliers who hear real constraints give real answers, and the ones who cannot work inside them will say so early, which saves everybody.
3. Publish the non-goals. Say what you are deliberately not doing this year. It is the fastest way to find out whether a supplier listens or recites, and the habit is worth keeping internally too.
4. Ask for reasoning, not artefacts. A conversation about how they would approach your situation, what they would need, and what would make them decline the work, reveals more than any deck.
5. Ask about a decision they got wrong. What they thought, what happened, what changed afterwards. Anyone who has done this work for a decade has several. Anyone who has none is either new or not answering.
6. Check references on continuity. Ask past clients who actually did the work after the first month and how quickly problems surfaced. Results are context-dependent; behaviour is not.

Two of those points reward a brief written against published standards rather than opinion. Search fundamentals are documented publicly, and so are the user-experience metrics a site is measured on, which means you can name a reference instead of a feeling and hold anyone to it. On non-goals specifically, we have written about publishing them deliberately, and our parent company has written about judging an agency without spec work, which covers the evaluation side in more detail than this piece has room for.

Judge it on the first ninety days, honestly

Whichever option you choose, the review point arrives before the results do, and this is where good arrangements get cancelled and bad ones get extended. Agree in advance what the first quarter can show, because the honest list is shorter than most people expect and much more useful than a revenue target nobody can attribute. We set that out separately in what you can honestly measure in the first ninety days.

Be especially careful with channel judgements in that window. The channel that looks worst on a default report is often the one doing the introducing, and the channel that looks best is often the one that happened to be present at the end. Before anybody is praised or cancelled on those numbers, read why your best performing channel is probably miscredited, because early cancellations made on bad attribution are expensive and invisible.

If you can write the brief, buy hours. If you cannot write the brief, buy judgement. If you can write the brief and the work never ends, hire somebody.

That rule resolves most of these conversations in a single meeting, and the concession attached to it is worth stating: it assumes somebody internally has the time to own the relationship, and where nobody does, the honest answer is to do less marketing well rather than more of it badly. The constraint is almost never money; it is attention. If you want to talk through which of the three fits what you are actually short of, tell us where the work is stuck and who currently owns it.