To choose a marketer, start with the task rather than a CV: you may need a one-off diagnosis, a strategy or ongoing management. That decides the format: an in-house employee, a contractor for one channel, an agency or a full-cycle specialist. Listen to how the person explains their decisions and look at the reports. Logos tell you nothing.
01The task first, the person second
Most bad hires start the same way: first the person is chosen, then someone works out what to give them. The specialist ends up doing what they know rather than what the business needs, and both sides feel misled.
So the first step is to state the task in words that can be checked. Not we need marketing, but understand within three months why the cost per lead has doubled, or put together a plan for launching a new product.
The format follows from the task. A one-off question is closed by a consultation, an unclear situation by an audit, a long piece of work by a strategy and ongoing management. Different money, different people.
It helps to write the task down in one paragraph and show it to two or three candidates. Different answers to identical text are the best comparison available: you see who clarifies the conditions and who immediately sells a standard package.
- What has to change, and in which numbers it will be visible.
- How long you are prepared to wait for the first movement.
- What budget is allocated for the work, and separately for the channels.
- Who inside the company owns the tasks on the business side.
02Four formats: in-house, freelance, agency, full cycle
Company size has little to do with it. What decides is whether someone inside can brief marketing and review the work. Where there is no such person, a narrow contractor almost always turns out to be useless.
The table is written without advertising any one option: each of them has a situation where it is the best choice.
Formats can be combined. A common arrangement for a small company: an external specialist keeps the whole picture and writes the briefs, while narrow jobs such as advertising, design and development go to contractors.
| Format | Strength | Weakness | When it fits |
|---|---|---|---|
| In-house marketer | Immersed in the product, always at hand | One set of skills, needs a manager | A steady volume of tasks and someone to lead them |
| Freelancer for one channel | Cheap and fast on a specific task | Sees only their channel and looks no further | The task is clear and has already been defined |
| Agency | A team and many hands at the same time | An account manager between you and the person doing the work | A large volume of work in several channels at once |
| Full-cycle marketer | One person accountable for all the work and for how the channels meet | Limited in hands, part of the work goes to contractors | The whole of marketing needs sorting out, not one channel |
03Criteria you can actually verify
Years of experience and a wall of logos cannot be verified. How a person thinks can be: ask them to take your situation apart out loud and watch whether they ask about economics and the product before proposing a solution.
The second verifiable trait is specificity. A good answer sounds like this: first we capture the starting numbers, then we test two hypotheses, and we judge the result by cost per lead. A poor one sounds like comprehensive promotion that will increase your sales.
Another trait is how the person talks about past projects. Sentences about what exactly did not work and why are rarer than success stories, and they say more about qualification.
A practical way to choose a marketer without long interviews is to give a small paid task: a review of one channel or an audit of one area. That shows the thinking, the discipline and the reporting format at once.
- Asks about numbers, the product and lead handling before offering services.
- Can explain why this channel and not the one beside it.
- Shows the scope of work and the format of the result before payment.
- Names a price or a range up front, not after the third call.
- Is willing to say the task would be cheaper to solve in-house.
- Leaves you the access, the documents and the data instead of keeping them.
04Ten questions to ask before paying
These questions are worth asking any candidate — in-house, freelance or agency. The answers tell you more than a portfolio.
- What will you count as the result of this work, and in which numbers?
- What starting data do you need from us in the first week?
- What is included in the scope, and what is not?
- Who exactly will do the work, and who will I be talking to?
- What does the report look like, and how often does it arrive?
- What will you do in the first month, and why that?
- In which case will you say the channel has not worked, and what happens then?
- Which jobs do you not do, and who do you bring in for them?
- How is the relationship documented, and what happens if it ends early?
- Who keeps the access, the accounts and the materials once the work is over?
05Red flags
None of the signs below proves bad faith, but each one needs an explanation. Where there is no explanation, the risk outweighs the expected benefit.
The most common flag is a promise of a result the contractor does not control. Search positions, a specific revenue figure by a date, a guaranteed flow of leads: all of it depends on algorithms, competitors, price and the quality of lead handling.
Another signal is pressure on time: the offer is valid today only, places are running out, the price goes up tomorrow. In work whose result is measured in months, that urgency has no basis.
- A guarantee of top 10 or of revenue growth written into the contract.
- A closed method that cannot be explained in words.
- A report about activity: posts, tasks and hours instead of leads and money.
- An offer to fake behavioral signals, reviews or followers.
- A price that only appears after several calls.
- Access and accounts registered to the contractor rather than to you.
- There was no diagnosis, yet a solution is already proposed at the first meeting.
06How to read case studies and reviews
A case study with percentages and no starting figures means nothing: tripling two enquiries produces six enquiries. Read it step by step: what the task was, what was found, what was done and what failed to work.
A useful sign is honesty about limits. A case that states which hypothesis failed and why is usually more credible than a collection of successes.
Reviews are useful for the detail: how fast people replied, how disagreements were resolved, what happened after the work ended. Praise on its own tells you nothing.
Few public cases is not a verdict: part of the work is covered by non-disclosure. Then ask the person to take someone else's example, or your own situation, apart out loud — how they think comes through without company names.
07What a normal start looks like
The sequence in which both sides understand what is happening is usually the same: diagnosis, a written proposal with scope and price, a contract, work in stages, acceptance of the result.
Four things matter in the contract: the scope of work, the deadline, the payment terms and what counts as delivered. If even one is missing, an argument about the result is inevitable.
Agree on reporting separately: format, frequency and metrics. A report should connect what was done with enquiries and money rather than list activity.
08What it costs and where not to economise
The spread of prices on the market is wide, and a low price is not a problem in itself. The problem is a price with no scope attached: it is unclear what is being paid for and what will exist at the end.
My guidance is visible before you enquire: a consultation is $100, a marketing audit from $730, an audit of the business, audience and product from $490, ongoing marketing $1,090 per month. The total depends on scope: the number of products, channels, markets and the depth of the data.
It is sensible to economise on volume, not on diagnosis. Working blind costs more: a month of management without understanding the cause is more expensive than the review that finds it.
09In short: the order of choosing
Choosing a marketer is easier in order: state the task and the numbers, pick the format that fits it, ask the ten questions, listen to how the plan is explained and put the agreements into a contract. That alone filters out most bad hires.
A good sign when picking a contractor or an outsourced marketer is a willingness to name limits: what the person does not do and when you do not need the work at all. Someone who takes on everything usually answers for nothing.
If you would rather test the format than sign a long contract, start with one consultation: a single meeting shows how the person thinks and whether it makes sense to go further.
