May 23, 2026

Hiring WritersLOT 510 · 05.2026

How to Hire a Content Writer for Your Website

Postdex Journal
  • hiring writers
  • buying content
  • b2b content
  • freelance writers

Hiring a content writer used to be a supply problem. You posted a job, you got twelve applicants, and the work was sorting the good ones from the rest by reading their samples.

That process is broken now, and not for the reason most people assume. The problem is not that there are too few writers. It is that there are too many, their samples all read well, and the single filter everyone relied on stopped carrying information.

This is a working process for hiring someone to write your website or blog content in 2026, including what to test for now that the portfolio does not tell you much.

The Talent Pool Inverted

Two things happened at once between 2024 and 2026.

Demand for commodity writing collapsed. Upwork writing projects fell roughly 32 percent year over year, because a competent 1,000 word draft stopped being something you needed to hire for. The businesses that were buying filler simply stopped buying.

Supply did not fall to match. The writers are still there. What changed is that a large share of them now run a model, edit the output lightly, and bill it as their work. That is not a moral panic, it is just the economics: if the client cannot tell, and the rate is per word, the incentive is obvious.

So you are hiring into a market with more applicants, lower average differentiation, and a much wider gap between the best and the median than existed three years ago. The filter matters more than it ever has, and the old one no longer works.

32%year over year drop in Upwork writing projects
$100-$200cost of a paid trial piece, which settles what an interview cannot
2xwhat domain knowledge costs, and about half the time you do not need it

Portfolios Stopped Being a Filter

Here is the uncomfortable part. A portfolio proves that someone can produce clean, structured, on-topic prose. In 2026, that proves almost nothing, because clean structured on-topic prose is the one thing that became free.

Every applicant’s samples will read well. They will have good headings, varied sentence length, and a confident tone. You cannot sort applicants on those attributes anymore, because the floor rose to meet the ceiling on exactly those attributes and no others.

You may be tempted to reach for a detection tool at this point. Do not build your process on one. False positives are common enough that you will reject good human writers, and the tools are trivially defeated by anyone who cares. A detector will make you feel like you have a filter while quietly giving you a random number.

What did not become free is the thing underneath the prose. Judgment, specific experience, an opinion that costs something to hold. That is what you have to test for, and you cannot test for it by reading samples.

What to Test For Instead

Replace the portfolio review with a short conversation or a paid trial that probes for things a model cannot produce on demand.

Ask for a position, not a summary. Give the candidate a topic in your space and ask what most people get wrong about it. A model will produce a balanced overview, because balance is what it regresses toward. A person with real experience will have an opinion, and often an irritable one. The irritation is the signal.

Ask what they would cut. Hand them a brief that is slightly overloaded and ask what they would drop and why. Writers who understand the job cut things. Writers who bill by the word never cut anything.

Ask about a failure. What did they write that did not work, and what did they conclude from it? This is nearly impossible to fake convincingly, because a fabricated failure is always tidy and instructive, and real ones are neither.

Ask them to be specific about your niche. Not “do you know B2B SaaS” but “what does onboarding churn actually look like in a product like ours”. Generic competence collapses immediately under a request for specificity.

Pay for a trial piece. One real article, at your real rate, on a real brief. It costs you $100 to $200 and it settles everything the interview cannot. Anyone who will not do a paid trial is telling you something. There is a right and a wrong way to structure one, and how to run a paid trial piece covers judging it fairly rather than just cheaply.

A refusal to trial is only one of the signals worth watching. The other red flags when hiring a content writer are mostly things that show up before you ever see a draft.

Qualify the Role Before You Post It

A large share of bad hires trace back to a vague job description, because a vague description attracts everyone and selects for nobody.

“Content writer” is not a role. A B2B SaaS writer who can hold a technical conversation with your engineers, a lifestyle writer with a warm consumer voice, and a conversion copywriter who writes landing pages are three different professionals who share a job title and almost nothing else. Hiring the wrong one produces work that is competent and useless.

That last confusion is the most expensive one, and the difference between a content writer and a copywriter is worth settling before you write the job post. If you are hiring into a technical space specifically, hiring a B2B writer for blog posts has its own filter, because the piece has to survive contact with a buying committee.

Before you post, decide four things.

  • What the reader already knows when they arrive.
  • What one action the piece exists to cause.
  • Which of your competitors' pages you are trying to beat.
  • Whether the writer needs domain knowledge or can acquire it from a brief.

That last question is the expensive one. Domain knowledge costs roughly double, and about half the time you do not need it. A well briefed generalist beats a poorly briefed specialist more often than anyone in this industry likes to admit.

Where to Actually Find Them

The channels sort by how much filtering they have already done for you, and the full tradeoff between an agency, a freelancer, and a marketplace comes down to how much of that filtering you are willing to pay someone else to do.

Marketplaces and job boards. Highest volume, lowest signal. You will do all the filtering yourself, against the problem described above. Workable if you run a paid trial and are ruthless about it.

Content platforms. They filter for you, and you pay for that in per-word rates plus, on most of them, a monthly subscription that is easy to miss. Verblio, WriterAccess and Compose.ly all bundle a fee that changes the real cost considerably. We broke down what a blog post actually costs including those platform fees, because the advertised rate is rarely the invoice.

Referrals. Highest signal, lowest volume, and the only channel where someone has staked their reputation on the answer. Worth asking for even when you think nobody will know anyone.

Buying finished work. Skips hiring entirely. You read the actual piece before you pay for it, which removes the whole prediction problem. It only works when what you need happens to exist already.

The Brief Is the Job

Most weak first drafts are a briefing failure wearing a writing failure’s clothes. A writer handed a topic and a word count will produce something technically correct and generically useless, and that is the brief’s fault, not theirs.

The brief carries the four decisions from earlier plus the constraints. Who it is for, the one job it does, what to leave out, what a good version looks like. If you cannot write that down, the writer cannot infer it, and you will spend the revision round discovering what you actually wanted.

This is the single highest leverage thirty minutes in the whole process. We wrote a full walkthrough on how to brief a writer so the first draft lands, and it is worth reading before you hire rather than after.

What to Pay

Rates in 2026 cluster rather than spread. Credible human writing runs $0.10 to $0.16 per word, so $100 to $160 for a 1,000 word piece. Expert and research heavy work runs $0.20 to $0.40 per word. Below about $75 a post, you are buying model output with a light pass, whatever the listing says.

Tier What you pay
Credible human writing $0.10 to $0.16 per word, or $100 to $160 for a 1,000 word piece
Expert or research heavy $0.20 to $0.40 per word
The band to avoid entirely $30 to $75 per post

The band between $30 and $75 is the one to avoid entirely. It is too expensive to be honest filler and too cheap to be a professional doing real work, which means it is almost always the former priced as the latter.

Pay unevenly. Your pillar pages and anything that has to carry authority deserve the top of the band. Supporting articles do not, and pretending otherwise just burns budget you will want later.

Frequently Asked Questions

How much does it cost to hire a content writer for a website? $0.10 to $0.16 per word for credible human work, which is $100 to $160 for a 1,000 word article. Expert or technical writers run $0.20 to $0.40 per word. Platform subscriptions, where they apply, add $39 to $99 a month on top.

How do I know if a writer is using AI? You largely cannot, and detection tools will mislead you in both directions. Test for judgment instead. Ask for an opinion, ask what they would cut, ask about something that failed. Those are hard to fake in a live conversation and easy to fake in a portfolio.

Should I hire a generalist or a specialist? A well briefed generalist beats a poorly briefed specialist most of the time and costs about half. Hire the specialist when the piece genuinely requires domain knowledge that a brief cannot transfer, which is less often than it feels.

Is a paid trial piece worth it? Almost always. One article at your real rate on a real brief costs $100 to $200 and resolves what no interview or portfolio can. Refusal to do one is itself a useful answer.

What is the difference between a content writer and a copywriter? A content writer produces articles that inform and rank. A copywriter produces pages that convert. They are different skills and hiring one for the other’s job reliably disappoints everyone.

How Postdex Fits

Postdex exists for the case where hiring is more process than the problem deserves. Every article in the catalog is finished and readable before you buy, so you are judging the actual piece rather than predicting from a sample. Each one is an edition of one, and it comes off the shelf permanently once sold.

When nothing on the shelf fits, the commissioning desk takes the brief and the piece is written to your specification, with full rights transferring on delivery. Every piece is reviewed by a named editor before it goes out, which is the part of the old portfolio filter that still carries information.

If you are hiring rather than buying, the process above still stands on its own. Qualify the role, write the brief, run a paid trial, and stop trying to read intent out of a portfolio that no longer contains any.

The Postdex Marketplace

From The Ledger

Finished articles on the shelf right now. Each one sells once, to a single owner.