Hiring a marketing copywriter is one of the few marketing decisions where the cheap option is reliably the expensive one. A weak writer produces pages that need rewriting, ads that never clear the learning phase, and emails that quietly train your list to ignore you. A strong one changes the economics of channels you already pay for, without you spending another dollar on media.
The problem is that the job title covers at least four different skills, rates range from $40 an hour to $400, and almost nobody's portfolio tells you whether their work actually converted. This guide is about closing that gap: what the role really involves, what the market charges, how to brief someone so the first draft is usable, and how to judge the results with numbers rather than taste.
What a marketing copywriter actually does
Copywriting is not writing. It is the job of getting a specific person to take a specific action, using words as the mechanism. Everything else follows from that distinction.
In practice the work splits into research, structure, and language, in that order of importance:
- Research. Reading support tickets, sales call recordings, review sites, and churn surveys until the writer can describe the customer's problem in the customer's own vocabulary. This is where the difference between a $60 and a $200 writer usually shows up.
- Structure. Deciding what the page argues, in what order, and what objection each section clears. A landing page is an argument with a shape, not a pile of paragraphs.
- Language. The sentence-level craft: specificity over adjectives, verbs over nouns, concrete numbers over claims.
A good writer spends more time on the first two than the third. When a draft comes back reading well but converting badly, the failure is almost always structural — the page answered questions nobody was asking.
The deliverables you will actually commission tend to be landing and product pages, paid social and search ad variants, lifecycle and campaign email, sales enablement copy, and occasionally naming and positioning work. Long-form blog content is a related but distinct discipline, closer to what we cover in our guide to lead generation content marketing than to conversion copy.
Copywriter, content writer, brand writer: not the same hire
These titles get used interchangeably and shouldn't be. Hiring the wrong one is the most common and most expensive mistake in this category.
Conversion copywriters write pages and ads that are measured on a rate — sign-ups, demos, purchases. They think in objections and proof. They should be comfortable being judged on a number.
Content writers write articles, guides, and documentation measured on rankings, time on page, and pipeline influence over months. Excellent content writers are often mediocre landing-page writers, because the two jobs reward opposite instincts: one rewards patience and completeness, the other ruthlessness.
Brand and messaging writers work upstream of both, defining positioning, value propositions, and voice. This is strategy work that produces a short document, not a page. It is worth paying for once, before you commission twenty pages that all say something slightly different.
Technical or product marketing writers translate a complicated product into claims a buyer can evaluate. In B2B SaaS this is frequently the constraint rather than persuasion skill, which is one reason specialist agencies exist — see our breakdown of what a B2B SaaS marketing agency does and doesn't cover.
Write down which of these four you need before you post a brief. Most bad hires start with a job description that quietly asks for all four.
What a marketing copywriter costs
Rates vary more than in almost any other marketing role, because the title is unregulated and the results are rarely traceable back to the writer. Here is the shape of the market you will encounter.
Freelance, hourly. Roughly $50–$90 for a generalist who can turn around clean, competent copy on a defined brief. Roughly $100–$175 for a specialist with domain knowledge and a track record of tested work. Above $200 you are usually paying for strategy and research capacity rather than words.
Freelance, per project. More common than hourly at the senior end, and generally better for you. A landing page with research and two revision rounds typically lands between $1,200 and $4,000. A full email welcome sequence of five to seven messages runs $1,500 to $5,000. A homepage plus three core pages, with messaging work included, is usually a $6,000–$15,000 engagement.
Retainers. $2,500 to $8,000 a month buys a defined output — for example, two pages plus a set of ad variants plus one email sequence. Retainers work when you have a genuine pipeline of work and someone internally to feed it. They waste money when you don't, because you end up commissioning filler to justify the invoice.
Agencies. Expect $150–$300 an hour blended, or project fees roughly 1.5 to 2.5 times a comparable freelancer. You are buying reliability, a second reviewer, and continuity when someone leaves. For a single page, that premium rarely pays for itself. For an ongoing programme across several channels, it often does.
In-house. A mid-level marketing copywriter in the US costs roughly $70,000–$95,000 base, a senior one $100,000–$140,000, plus 25–30% in employer costs. The break-even against freelance sits at around 12–15 substantial deliverables a year. Below that, you are paying salary for idle capacity; above it, in-house is cheaper and the institutional knowledge compounds.
The number that matters is not the rate. It is the cost of the copy set against what the copy moves. A $3,000 landing page that lifts conversion from 2.1% to 2.8% on 8,000 monthly visits pays for itself in weeks. The same page on 300 monthly visits is a vanity purchase. Run the traffic and rate through a conversion rate calculator before you commission anything, and you will know immediately whether copy is your constraint or whether traffic is.
How to write a brief that gets a usable first draft
Most disappointing first drafts are brief failures, not writer failures. A brief that fits on one page and answers the following will get you something close to usable on the first pass.
- The one action. What single thing should the reader do? Pages with two calls to action convert worse than pages with one, and briefs with two goals produce copy with neither.
- The reader, specifically. Not "marketing managers" but "a marketing manager at a 50–200 person B2B software company who has just been told to cut agency spend." Job title alone is not a reader.
- What they believe now. The single objection or misconception the copy has to dismantle. If you cannot name it, the writer will guess, and they will usually guess wrong.
- Proof you can actually use. Customer quotes, numbers you are allowed to publish, integrations, guarantees, named logos. Copy without proof becomes adjectives.
- Raw material. Three sales call recordings, twenty support tickets, or a churn survey export beats any amount of description. Writers who ask for this are the ones to hire.
- Constraints. Legal review requirements, words you cannot use, character limits, the design template the copy has to fit.
- How it will be judged. Name the metric and the measurement window before the work starts, not after the draft arrives.
That last point prevents the most common failure mode in copy projects: the draft gets judged by whoever has the strongest opinion in the room rather than by whether it worked.
Test before you commit
Never start a copywriter on your homepage. The homepage is the hardest thing to write, the slowest to measure, and the most politically loaded page you own.
Start with a paid test project — real work, paid at full rate, small enough that a bad outcome costs you a few hundred dollars instead of a quarter. Good options are a single landing page for a live campaign, a five-email welcome sequence, or eight ad variants against an existing control. Each is measurable inside a few weeks.
Judge the test on four things, in this order: the questions they asked before writing, whether the structure addresses real objections, how they respond to revision notes, and only then the sentences. Writers who argue every edit are expensive to work with. Writers who accept every edit without pushing back on the ones that will hurt performance are worse — you are paying for judgement, not compliance.
Ask candidates for one piece of work where they know the result and can explain why it worked. The answer separates people who write copy from people who understand what copy is for. A surprising number of well-reviewed writers cannot name a single conversion number from their own portfolio.
How to judge whether the copy is working
Copy is one of the few marketing investments you can measure cleanly, provided you decide the measurement in advance.
Set a real baseline. Two to four weeks of pre-change data on the exact page or sequence, on the exact traffic source. Comparing a new page against a site-wide average is how teams convince themselves that a rewrite worked when seasonality did the work.
Test one thing. If you rewrite the page and redesign it and change the offer in the same week, you have learned nothing about any of the three.
Wait for the right sample, not the right answer. Most landing page tests need several hundred conversions per variant before the result means anything. Calling a test at 40 conversions because the new version is ahead is the single most common way teams ship worse copy.
Track the downstream number, not just the click. Ad copy that lifts click-through by 30% while cutting lead quality in half has made your funnel worse and your dashboard better. Check where the change lands on acquisition cost with a CAC calculator, and for subscription and repeat-purchase businesses, check whether the customers the new copy attracts are worth as much over time using an LTV calculator. Lifecycle copy in particular is judged over months, which is why the segmentation discipline in our guide to targeted email marketing matters more than the wording of any individual send.
For ecommerce specifically, product and category page copy sits inside a wider set of levers — pricing presentation, shipping clarity, review placement — covered in our guide to ecommerce conversion rate optimization. Copy changes are easiest to attribute when nothing else on the page moved.
Red flags worth walking away from
No questions before the quote. A writer who can price your homepage from a two-line email has not thought about your homepage.
A portfolio of pages but no results. Screenshots are not evidence. "I don't have the numbers on that one" is acceptable once. Every time is a pattern.
Volume pricing. Per-word rates suit content mills, not conversion work. The best version of a headline is often shorter than the worst.
Refusal to touch structure. If a writer will only supply prose to drop into an existing layout, you have hired a typist. The structure is most of the value.
AI output with a human invoice. Assisted drafting is normal and fine. What is not fine is paying senior rates for lightly edited generic output. The tell is copy that is fluent, confident, and contains no specific fact about your business that the writer could only have learned by asking.
FAQ
How long should a landing page take? For a specialist with a complete brief and access to research material, five to ten working days for a first draft, plus a week for revisions. Anything promised in 24 hours skipped the research step, which is the step that does the work.
Should I hire a copywriter or an agency? A freelancer for one to three discrete projects a quarter, where you have someone internally to manage the brief. An agency when you need multiple channels covered continuously, or when nobody in-house has time to run the process. Agencies cost roughly 1.5 to 2.5 times more for the same deliverable, and that premium buys continuity rather than better sentences.
Can one writer cover ads, email, and website copy? Often yes, at the mid-market level, provided they are strong on structure. Specialisation matters most at the extremes — highly technical products, regulated industries, and high-volume paid social, where channel-specific instincts are worth paying for. Sales outreach is its own discipline again; the mechanics differ enough that we cover it separately in our sales email guide.
What if the new copy performs worse? Roll back, and treat it as information rather than a failure. Roughly a third of well-executed copy tests lose, and the losses usually tell you something specific about what your buyers care about. The only genuinely wasted test is one you shipped without a baseline, because you cannot learn anything from it in either direction.