How to test a content writer before hiring

Last updated: October 11, 2026

Key Takeaways

• Test for problem-solving, not just polished prose.
• A paid test of 500 to 1,200 words should use your real brief and include at least one revision round.
• Use the format that reveals the most likely failure for the role.
• Score the work with a rubric before reading it, not after forming an opinion.

Give a content writer a small paid assignment, a clear brief, and a real deadline. That setup shows far more than a polished portfolio ever will. For how to test a content writer before hiring, it is usually the cleanest way to see whether they can think, research, and write for your audience without guessing.

A generic “write me something about X” sample misses the point. Almost anyone can make a topic look decent in isolation. The real test is harder: can they handle your tone, accuracy, SEO, judgment, and revision without wobbling?

What are you really trying to prove with the test?

You are trying to prove they can solve your content problem, not just produce 800 clean words. Need lead generation? Then the assignment should show whether they can write toward a conversion goal. Need authority? Then you want evidence they can organize facts without stuffing in fluff. Need speed? Timing matters just as much as prose.

Here is the simplest way to think about it:

Portfolios can flatter the wrong things. A bad assignment does the same. Honestly, I would rather see a smaller test that mirrors the real job than a grand sample that looks impressive and tells you little.

A good test should prove five things: they can understand a brief, they can research without inventing facts, they can write in your voice, they can edit their own work, and they can accept direction. If one of those matters less for your role, say so. A newsroom writer and a SaaS blog writer are not graded by the same yardstick.

Quick check: If the writer will touch your real audience, your real topic, or your real offers, the test should resemble real work, not a generic writing prompt.

How should I structure a paid writing test?

A paid test should be 500 to 1,200 words, use your real brief, and include at least one revision round. That is long enough to reveal judgment and short enough to keep risk contained.

I would not ask for a full article draft unless the role is clearly senior and the assignment is very close to paid work. A 1,000-word test usually gives enough surface area to judge structure, sourcing, and voice. Need only email copy? 300 to 500 words may be enough. For long-form thought leadership, though, 1,200 words is a much fairer sample than a tiny excerpt.

Use this workflow:

1. Write a one-page brief with topic, audience, goal, tone, target length, and one or two examples of what “good” looks like.
2. Give them a deadline that mirrors your real process, such as 48 hours or 5 business days.
3. Ask for an outline first if the role requires planning, strategy, or SEO structure.
4. Review the outline before they draft, so you can see how they think before you judge the prose.
5. Send one revision note set, then see whether they improve the piece without breaking it.
6. Score the final version against the same rubric every time: accuracy, clarity, voice, structure, and responsiveness.

Skip the outline step, and a structural problem can hide until the end. Skip revision, and you are hiring only for first-draft instincts. That is a gamble for any role that involves editors, clients, or stakeholders.

Need a benchmark? A normal article test can be reviewed in a Google Doc with comments, or in tools like Notion or Word. For SEO-heavy roles, I would want to see whether they know the difference between a service page, a comparison page, and an informational article. That split matters more than many applicants realize.

Quick check: If your final content will ever go through an editor, client, or SEO lead, include at least one revision round in the test.

Which writing test format works best for your situation?

The best format is usually the one that exposes the failure you most want to avoid. Weak writing? Use a clean draft test. Fake expertise? Use a research test. Slow delivery? Use a time-boxed assignment with a hard deadline.

A few formats work well:

• Brief-to-draft test: Best for general blog writers and content marketers.
• Research memo plus draft: Best for technical, medical, finance, or B2B topics.
• Editing test: Best if the role includes rewriting, repurposing, or content polish.
• Headline and outline test: Best for strategists, editors, and content leads.
• Take-home assignment plus live discussion: Best for senior hires where judgment matters as much as writing.

The standard advice says “ask for a writing sample.” Too vague. Better to hand over a real-world prompt and watch what happens when the brief is imperfect. Strong writers ask smart questions. Weak ones either write blindly or wait for the answer key.

Here is a practical rule: junior roles should check execution. Mid-level roles should check execution plus editing. Senior roles should check strategy, risk spotting, and stakeholder sense.

Do not overcomplicate the format. A 45-minute live exercise can be useful for editors, but it is a poor substitute for the kind of work a writer will actually do at home with sources, notes, and time to think. I would use live exercises as a supplement, not the only filter.

For fact-heavy work, the source standard should be clear. Look for reliable references such as government agencies, professional associations, standards bodies, or peer-reviewed summaries. The CDC, the NIH, and the U.S. Government Publishing Office are better signals than a random blog roundup, and the same is true for medical and legal content, where they should know when to avoid strong claims and when to recommend professional review. Anything else can get wobbly fast.

Quick check: If the writer will mostly create first drafts, test drafting; if they will shape or correct content, test editing too.

How do you score a content writer fairly?

Score them against a rubric before you read the test, not after you like or dislike the piece. According to the American Psychological Association, a rubric helps reduce bias by making evaluation criteria explicit, and that also helps guard against halo effect, where one strong paragraph makes the whole sample feel better than it is.

A simple rubric can use five categories on a 1-to-5 scale:

1. Brief fit — Did they answer the assignment you actually gave?
2. Clarity — Can a reader get through it without backtracking?
3. Accuracy — Are the claims supportable and sourced where needed?
4. Voice — Does it sound like your brand, not a random blog?
5. Judgment — Did they choose the right level of detail and avoid filler?

Need SEO? Add a sixth category for search intent. A writer can be fluent and still miss the point if the article answers the wrong question. Good grammar, wrong page — that is a common content-team faceplant.

Here is the part many generic articles miss: a polished sample can hide a bad thinking process. Pay close attention to the notes they leave behind, especially if they explain why they cut a section, chose a source, or changed a heading. Those notes show whether they can work with an editor.

For fact-checking, a useful benchmark is whether they name the source type in the draft, not just scatter vague claims. “According to the CDC,” “per the American Medical Association,” or “as defined by ISO standards” tells you more than “studies show.” If they cite statistics, check whether the figure is current and relevant, not just impressive sounding. The NIH and CDC both publish guidance that can help you judge whether a source is primary and current.

Do not hire purely on speed unless speed is your only need. Fast writers can be wrong. Slow writers can still be right. The right balance depends on the calendar pressure around your content pipeline.

Quick check: If you could hand the same sample to three different editors and get three different emotional reactions, you need a tighter rubric.

What should you look for in a writing test sample?

Look for evidence that the writer can think in paragraphs, not just sentences. Strong samples have a clear angle, clean transitions, and no dead zones where the writing fills space.

Watch for these signals:

• They open with the reader’s problem, not a warm-up paragraph.
• They use concrete examples instead of abstract claims.
• They know when to cut a section that does not serve the goal.
• They do not stuff keywords into every heading.
• They vary sentence length without sounding choppy.
• They distinguish fact from opinion.

A weak sample often sounds smooth while saying very little. Content marketing is especially prone to that trap; applicants learn that “friendly tone” can hide vagueness. If the draft is full of broad phrases like “in today’s world” or “it’s essential to note,” treat that as a warning sign and, for anything regulated or high-stakes, ask for a source-backed rewrite or professional review.

One useful mini-test is to ask the writer to explain two choices in 2 or 3 sentences: why they used that structure, and why they chose those sources or examples. A writer who understands their own draft can usually revise it. A writer who cannot explain it may have been lucky.

For technical or regulated roles, ask for source notes. Not a bibliography for show, but a short list of the pages, standards, or documents used. That lets you see whether they know the difference between a primary source and a random roundup article.

A good test also shows whether they can follow house rules. If your brand avoids heavy jargon, the sample should too. If your brand prefers direct, plain English, see whether they can write that way without sounding flat.

Quick check: If the sample reads nicely but you cannot point to a single specific insight, the writer may be polished but shallow.

When does the standard advice break down?

The standard advice breaks down when the job is unusual, sensitive, or high-risk. In those cases, a normal take-home article test is not enough, and sometimes it is the wrong thing to ask for.

1. If the topic is regulated, then accuracy matters more than style.
   What changes: medical, legal, tax, and financial content can create real harm.
   What to do instead: test with source-heavy prompts and a fact-check review, and consider professional review before publication. A writer who is great at lifestyle content may be wrong for this work.

2. If the writer will manage clients, then communication is part of the job.
   What changes: responsiveness, clarification questions, and revision behavior become critical.
   What to do instead: include an email exchange or a short live call. A great draft does not fix poor stakeholder handling.

3. If you need a ghostwriter, then imitation matters more than originality.
   What changes: the writer has to match voice, rhythm, and point of view.
   What to do instead: give them 2 or 3 examples of the real author’s work and ask for a short section in that voice. A generic sample is nearly useless here.

4. If the work is strategic, then a writing sample alone is too small.
   What changes: topic selection, sequencing, and content priorities matter as much as prose.
   What to do instead: ask for an outline plus a short rationale. For a content lead, a 30-minute live debrief can reveal more than a 1,500-word draft.

5. If you are hiring on a tight budget, then paid tests still matter.
   What changes: the temptation is to crowdsource free samples from lots of applicants.
   What to do instead: pay a small number of serious finalists for a real assignment. Free spec work attracts more applicants, not better ones.

6. If the role is urgent, then the deadline test matters.
   What changes: speed and reliability become part of the evaluation.
   What to do instead: give a short but fair turnaround, such as 24 to 72 hours depending on scope, and see who communicates early if they hit a problem.

These cases are where generic hiring advice fails first. A writer can be excellent and still be wrong for a job that needs something else.

Quick check: If the content could affect money, health, compliance, or client trust, do not rely on a plain writing sample.

How do you avoid bad signals and false positives?

Separate presentation from performance. A clean portfolio, a confident call, or a polished sample can all hide weak judgment.

Red flags I would not ignore:

• They refuse a paid test and insist on unpaid samples.
• They answer a detailed brief with something obviously generic.
• They cannot explain how they researched the article.
• They ignore your revision notes or argue with every change.
• They use the same voice for every assignment.
• They bring claims without sources when the topic clearly needs them.

False positives often happen when a candidate is an excellent self-editor but a poor collaborator. Fine, if you want a solo freelancer with minimal oversight. Trouble, if they will work with editors, strategists, or clients.

You can also get fooled by overconfidence. Some applicants talk as if they can write anything. That sounds efficient, but it usually means they have not thought hard enough about scope. A strong writer usually has boundaries. They know when a project needs more context, and they ask for it.

One practical safeguard is to compare 2 finalists on the same brief and score them blindly, before looking at resumes again. If one sample still wins after you hide the names, the test is doing its job.

Another safeguard is to ask one follow-up question: “What would you change if this article had to rank for a different intent or convert a colder audience?”

Drafted with AI; not yet reviewed by a person.