Poldermarketing
What should I look for in a freelance marketer for a seed-stage startup?
Look for a freelancer who has done your motion alone, product-led or sales-led, with no team under them. Name one constraint first: unclear offer, too few conversations, or weak conversion. Ask for a baseline, a decision they made, and a test they stopped. Pay for a short ship. Skip guarantees, unseen juniors, and a plan that needs a team.
Name the constraint before you open a profile
Hire the freelancer who can remove one constraint you can already name: an unclear offer, too few of the right people, or people who arrive and do not convert.
Krithika Muthukumar, in First Round Review, splits that diagnosis. If you can get prospects into a room and close them, you need more demand. If some people get the product and many do not, you need product marketing: differentiation, positioning, and a story sales can repeat.
Maya Spivak’s First Round playbook ties the split to the motion. End-user sales need familiarity with ads, email, social, search, and SEO, not mastery of each. A longer sales cycle needs case studies, credible site copy, testimonials, and education. Creative founders do not necessarily need another person whose spike is the creative side. A product nobody can explain needs a translator.
Arielle Jackson’s First Round guide catches the skipped step. Press, before a headline and a list of reporters who would care, still needs positioning, not a communications specialist. See which type of marketer fits an early-stage startup and when to hire the first growth marketer.
| What you can already say | The brief | Walk away if they propose |
|---|---|---|
| You close the people you reach, and almost nobody new arrives | One demand test: who, which channel, which qualified conversation, when you stop | A brand program, or five channels at once |
| People arrive and many do not understand the offer, or you cannot say why a buyer stayed or left | Customer conversations, a rewritten page or deck, one proof asset | More traffic into a story that does not convert |
| One channel already produces qualified conversations | A specialist on that channel. You keep the offer | A restart that leaves the channel idle |
| People buy or sign up, and they do not come back or activate | One lifecycle or activation test on current customers, with a written stop | A new channel meant to hide the leak |
| You cannot name who buys | No marketer yet. Interview buyers yourself | A campaign meant to find the buyer for you |
A growth profile assumes the page already works
A growth marketer, in Spivak’s definition, spreads materials you already have and experiments when people do not convert. Hire that profile only when a decent site and an easy signup already exist.
MarketerHire’s founding-versus-growth note separates the jobs. A founding marketer builds positioning, a first site that is not a side project, a first experiment, and the measurement that decides what continues. A growth marketer scales a channel that already produces pipeline. The miss they name is paid media hired before a message converts, because paid feels tangible.
Three stories across the homepage, the deck, and the sales call means you need the builder. For the specialist seat, see how to hire a freelance growth marketer. Spivak wants a T-shaped first hire because the motion can change within about six months. Muthukumar says a narrow specialist often fails when the diagnosis three months later is a different job. A growth candidate who has not tried your signup before the interview is out.
Stage fit is solo work, done recently
Stage fit is solo work on the same motion, a similar buyer, and a similar cycle, not a famous logo.
Jackson calls the first marketing seat single player. Ask when they last wrote the copy or set up the campaign, and whether they will do it again. She looks for about five to eight years on a first full-time hire, titled Head of Marketing or Director rather than VP. For a freelance contract, ask what they personally did in the first 90 days of the last two stage-matched engagements, and what they refused.
Muthukumar asks what they moved, what they invented, what failed, and what they left undone. Spivak adds the artifact: what is still live, and who wrote the words. A fixed sequence from a large company is a weak first hire, because so much changes. MarketerHire’s freelance guide flags the other disguise: a senior person sells, and a junior you never meet does the week. A marketer who executes can point at the file.
Proof is a baseline, a decision, and a stop
For two pieces that resemble this brief, ask what was true at the start, what they decided, and what they stopped. A gain with no starting point, no period, and no note of what else changed is not evidence.
Jackson’s test of a claimed gain is which part they owned. Copy, without running the campaign, is not ownership. Muthukumar treats engagement on a post, arrivals on a page, and email opens as vanity. With sales, look at revenue the work influenced. With a product, look at usage. Do not judge a channel on activity from its first days. Her first month is a funnel note, time on customer calls or recordings, and a hand in a real launch. Read the channel on a qualified conversation, activation, or use only after the test has run a full cycle.
MarketerHire’s freelance guide flags a freelancer who cannot cite specific results from the last three engagements, and asks for a reference at your stage or in your industry. If the numbers cannot be shared, require one reference and one artifact you can still open. Call a founder or the person who owned revenue, not their manager. Ask about a wrong assumption, a disagreement, and whether they would hire this person again for the same problem.
Ask for four answers. A failed test: hypothesis, the measure set in advance, and the stop. Product-led proof on activation, not impressions. Sales-led proof of qualified pipeline and time with sales on a good lead. Where they start tomorrow, which metric they ignore, and how they agree a qualified conversation with sales. Customers and tracking before any spend, then a message test on a channel you already have. Muthukumar flags a tactic suggested only because of the stage.
Grade a live read, then pay for one ship
Skip the 30, 60, or 90 day strategy document. Muthukumar’s reason is that anyone with a search box can write one. Whiteboard the funnel live. Muthukumar’s mark of a strong candidate is that you learned something that changes what you are already trying to do.
Use one scenario for every finalist, drawn from the real engagement. Score problem framing, prioritization, and clarity on a sheet you write before the calls. Then pay for one ship. MarketerHire’s trial is about two weeks: they raise a problem before you find it, the first week shows a deliverable, the choice is in plain language, and they change course when you push back. Before any login, agree the artifact, the date, the measure, who decides, who approves, and what you hand over in the first week.
- They name one bet, the metric, and the stop, then ship one page, sequence, or test.
- Extend only if that ship changed the next decision. A long minimum before any work ends the process.
Jackson puts a full-time search at about three to six months unless the person is a friend or former colleague you already know well. See freelance or a full-time hire.
End the call when the promise comes first
End the call when a promise arrives before they ask who buys.
- A guaranteed ranking, or a guaranteed return.
- A channel menu that does not change after you describe the buyer.
- A first month on a brand system, a stack migration, or a listening tour that ends in a deck.
- A first step that hires more people, or briefs an agency before the page exists. Spivak says the first marketer still does hands-on work for at least a few quarters.
- Messages to people who never asked to hear from them. Spivak asks which tactics they refuse.
- Clicks, when you asked for pipeline or activation.
You still own the buyer, the number, and the logins
You keep the customer, the offer, and the one number. Jackson’s test is whether you can name problems you cannot cover, rather than a wish to grow because an investor said so. She hires sooner on a marketing-led path, a consumer brand or a signup driven by content and paid, around the time the company reaches about 10 people. An enterprise sale waits for a repeatable sales motion, around the search for a sales leader. Freelance is the other door: one defined project now, while that longer search runs.
You keep the site, the ad accounts, the analytics, and an export of the list. Access ends with the trial unless you extend, and one person is never the only login. If the file includes people in the EU, you decide why it is used, so you are the controller, and a freelancer who only stores or sends it on your instructions is the processor. Before the first send, Article 28 of the GDPR requires a contract: the file and its purpose, action only on your written instructions, no other processor without your written permission, and deletion or return when the engagement ends. A list they brought is a reason to walk away.
Five freelancers who never talk to each other leave the work uncoordinated, until managing them is the full-time job. Hire in house then. A growth marketer or an agency is the comparison when you need several channels at once.
What I screen for when the seat is mine
On a seed brief I start from one constraint you can already name, then one ship with a written stop. If the offer is hard to say, I use customer conversations and rewrite the page before a channel test. If a channel already produces conversations, I instrument that path and run one test with a written stop.
I am Piet Baudoin, one person under the name Poldermarketing, an AI-native growth marketer. I work fully remote, in Dutch and English. You can hire me freelance, for a bounded project or a few days a week. The offer is an all-in GTM solution for freshly funded startups anywhere: marketing, AI, and automation, from the first message to the first customers, without a separate specialist for every part. I am strong in AI, content, automation, and workflows, and I build that work in your accounts. I do the work, not only the advice. Google Ads and Meta Ads are newer for me. I can set them up and review them. I am the wrong hire when the core job is scaling a large paid program.
The free growth scan shows how the current site reads. How I work is the shape of an engagement. Positioning and messaging is where I start when the words are the constraint.
Questions people ask
Should I hire a generalist or a channel specialist?
Hire a generalist when nobody owns marketing yet and no channel produces qualified conversations on its own. That person writes the offer, stands up measurement, and runs one test. Hire a specialist only when one channel already produces those conversations and the job is to push that channel, not to invent the message. A paid-media specialist before the offer converts spends the test learning that the page is the problem.
How long should the paid trial be?
About two weeks is enough to see whether they narrow the scope, ship something real, and say what they learned. Agree the deliverable, the date, and the measure before day one. A full-time search often takes three to six months, so a freelancer is how you move now, not how you skip the test. Extend only if the shipped piece changed what you will do next. A long minimum before any work is a reason to walk away.
What if every result they cite is confidential?
Treat a fully confidential book as incomplete. Ask what they personally wrote, launched, and stopped, and whether a founder or the person who owned revenue will take a short call. A strong operator can still describe the decision. If they cannot name one person and one page you can open, you are grading a logo.
When should I not hire a freelance marketer yet?
Do not hire a marketer to invent the buyer. If you cannot name who buys, interview buyers yourself. A discrete project is a fair freelance brief. A pile of projects you cannot direct is not. Several freelancers who never speak to each other produce an uncoordinated site. Hire full-time when those marketing problems have multiplied and managing the freelancers has become the job.