Price Copywriting by Scope, Not Word Count, When Clients Say “AI Is Free”

When a client says “ChatGPT is free,” do not argue that AI is useless or pretend a human draft is automatically better. Clarify what they are buying. A text generator can produce a draft; a copywriting engagement may also include research, audience synthesis, offer decisions, source verification, stakeholder interviews, revision management, compliance checks, and a clear handoff. If the client only needs a rough first draft and can supply verified inputs, a professional service may not be the right purchase. If they need a publishable asset that makes a defensible claim for a specific audience, price the work that makes that deliverable possible.

Current market evidence is useful but easy to misuse. ProCopywriters’ 2026 survey reports 573 UK commercial-writer respondents; its public summary says 74% used generative AI and lists a £440 median freelance day rate. That is a self-reported UK survey, not a representative MENA pricing benchmark, and it does not show that AI use caused higher rates or improved client outcomes. It does show why “a free tool exists, therefore professional writing has no value” is not a reliable market conclusion. Price your own offer from its scope, operating costs, expertise, and evidence—not a single foreign benchmark. ProCopywriters Copywriter Survey 2026

AI can also produce confident but incorrect claims, fabricated references, and oversimplified answers; OpenAI’s guidance recommends verifying important information from reliable sources. If research and verification are part of your service, specify what you will check, what source standard you will use, and where your responsibility ends. “Fact-checked” should mean a documented review process, not an absolute guarantee that no error is possible. OpenAI Help Center: Does ChatGPT tell the truth?

Separate the deliverable from the production method

A client does not usually need to pay for the number of words alone. They need a defined asset that helps them make or support a decision. A landing page may require customer research, offer clarification, source checking, message hierarchy, conversion design, a review route, and variants. A technical article may require expert interviews, documentation verification, and an update owner. A three-line ad may take little space but require more research and legal review than a longer generic article.

Describe the deliverable in observable terms:

Audience and job: Who needs to understand, decide, or do what?
Asset and channel: What will be created, in which format and language?
Inputs: What documents, access, interviews, approvals, and subject-matter time will the client provide?
Work included: Research, brief, interviews, draft, editing, source checking, localization, variants, or test plan.
Acceptance criteria: What makes the work complete and usable?
Boundaries: What is excluded, what needs separate approval, and what depends on the client?
Handoff: Which source files, claim notes, usage rights, and maintenance information are delivered?

This makes it possible to compare a tool-assisted draft with a researched, reviewed, campaign-ready deliverable without pretending they are the same product.

Offer options that change the work—not the promise

Use a small set of scopes so a client can choose the support they need. The names do not matter; the boundaries do.

Scope	Client provides	Writer provides	Useful when
Editorial polish	A complete draft, approved facts, audience, and intended action	Clarity, structure, voice consistency, and a defined revision	The client already has a sound argument and needs editing
Research-backed asset	Business context, access to subject-matter input, offer details, and source restrictions	Brief, focused research, interview synthesis, copy, source notes, and editorial QA	The message or proof needs to be developed before drafting
Campaign-ready package	Stakeholder availability, approved claims, destination, and measurement context	Core asset, channel adaptations, supporting variants, implementation notes, and a proposed test plan	Multiple people need a coherent set of assets for launch

Do not sell a “higher tier” just by adding more words or file formats. Each scope should remove a real bottleneck or create a useful handoff. If the client only needs proofreading, do not upsell research. If the source evidence is missing, do not sell a publish-ready claim before the evidence owner responds.

Calculate a quote from the work and risk

For a fixed-scope project, estimate the work before naming the fee. Include briefing, preparation, interviews, research, drafting, editing, coordination, revisions, localization, quality checks, and handoff. Add the administrative time and operating costs required to deliver sustainably. Decide what complexity changes the quote: unfamiliar or regulated subject matter, multiple decision-makers, source verification, urgent turnaround, localization, or a large number of variants.

Separate base scope from optional work. For example, one defined interview, one core page, one review round, and a source note may be included; extra stakeholder interviews, a second language, new research, additional concepts, or a new offer may require a change order. State how quickly the client must review and what happens if approvals arrive late. A fixed fee without clear assumptions is a dispute waiting to happen.

Use hourly or day rates when discovery is uncertain or the work changes as evidence emerges. Use a fixed fee when the deliverables, inputs, review rounds, and acceptance criteria are stable enough to estimate. A retainer can fit recurring work when capacity, priorities, service levels, carry-over, and reporting are explicit. None of these fee models guarantees conversions or revenue.

Do not import the £440 UK median into a proposal for Egypt, the Gulf, or another market. Currency, buying power, type of client, specialization, scope, contract terms, and local operating costs differ. If a buyer requests a price comparison, ask whether the compared quote includes the same research, languages, rights, revisions, source checks, and implementation support.

Explain the value in terms of accountability

Avoid saying “AI cannot be creative” or “human writers always convert better.” Both are overly broad. Explain the controls and decisions in your specific workflow:

What the client must provide: approved facts, product access, customer permissions, offer terms, and subject-matter review.
What the writer will decide: audience framing, message hierarchy, channel adaptation, and the evidence needed for each claim.
What the writer will verify: source links, quotations, figures, product behavior, and consistency against supplied documentation—within the stated scope.
What remains the client’s responsibility: legal sign-off, final product accuracy, commercial terms, publication authorization, and business results.
How assisted tools may be used: agree on tool access, confidential inputs, data handling, and human review in the project terms. Do not expose client materials to a tool without authorization.

This does not require a generic public statement about AI. It requires clear client-facing scope language about how the work will be produced and reviewed where that affects confidentiality, quality, or ownership.

A direct response to the “AI is free” objection

Try a calm response that makes the decision easier:

“Yes, you can use a free AI tool to create a first draft. My quote is for [specific scope]: I’ll [research/interview/structure/edit/verify] using [approved inputs], deliver [assets], and include [review rounds and source notes]. It does not include [exclusions], and it can’t guarantee [business outcome]. If you only need a rough draft from facts you have already verified, you may not need this service. If you need the publishable asset and review process described here, I can price that against your brief.”

Then ask a qualification question: “Which part is currently costing you the most time or creating the most risk—finding the right argument, verifying claims, getting approvals, adapting the asset, or measuring what happens after launch?” The answer reveals whether to offer a project, a smaller review, or no engagement.

Protect the scope after the client says yes

Attach the accepted brief to the agreement. List deliverables, formats, languages, source standard, assumptions, access and interview requirements, review rounds, response times, client dependencies, usage rights, payment schedule, cancellation terms, and change-order process. Clarify whether drafts, research notes, raw interview recordings, and unused concepts are included. Keep privacy and data handling aligned with the client’s policy.

For performance-oriented copy, define what you can deliver and what you can only test. You can provide variants and a test plan; the client controls traffic, offer, product, analytics, sales response, and many other influences on results. Do not price as if you can guarantee conversion lift without reliable evidence and authority over the full system.

After delivery, ask what the client used, what needed revision, what questions sales or customers raised, and which claims need updating. Use that feedback to improve the next brief and estimate. Do not treat one winning variant as a universal formula or turn an anecdote into a public case study without permission.

Short-video script: “ChatGPT is free” is a scope question

Opening: “If ChatGPT is free, why hire a copywriter? Start by comparing the actual work.”

Voice-over: “A rough draft and a researched, reviewed campaign asset are different scopes. Clarify who supplies the facts, who interviews the expert, who verifies claims, what gets delivered, how many review rounds are included, and what happens after handoff. Price those steps transparently. If the client only needs a first draft, say so. If they need a publishable asset with a documented review process, quote that work—without promising sales the copy alone cannot control.”

Visual: A word-count quote expands into research, evidence, deliverables, review, handoff, and exclusions. CTA: “Save the scope checklist before writing your next proposal.”

Educational carousel: Price the job, not the word count

Cover: “ChatGPT is free.” Is that the same service you are quoting?
Define the job: Audience, decision, asset, channel, and language.
List the inputs: Facts, interviews, access, approval, and customer permission.
Show the work: Research, strategy, drafting, checking, localization, and QA.
Set the boundary: Deliverables, revisions, exclusions, dependencies, and handoff.
Choose the fee model: Fixed scope, hourly/day, or retainer based on uncertainty and recurrence.
Be honest: Sell a reviewable process, not guaranteed conversion or revenue.

Field experiment thread: Find which scope clients actually need

Proposed offer test; no result is claimed.

1/ Review recent inquiries and note the requested asset, business problem, source readiness, approval complexity, and reason the client considered AI or another provider.

2/ Create two or three clearly different scope options. Change the work included, not merely the word count or the promise.

3/ Quote a sample of comparable opportunities using the same discovery questions and a consistent estimation method. Do not change several price variables at once.

4/ Record client fit, questions, acceptance or decline reason, estimated versus actual hours, scope changes, and whether the delivered asset was approved for use.

5/ Review which tasks repeatedly exceeded estimates and which included services clients did not need. Adjust assumptions and boundaries before adjusting price alone.

6/ Repeat with enough comparable proposals to spot a useful direction. Treat a tiny or highly mixed sample as inconclusive; do not infer market-wide willingness to pay from one client.

Tool-neutral prompt pack: Turn an inquiry into a scoped quote

Prompt 1 — Discovery questions

Based on this anonymized client request, list the missing information needed to estimate copywriting work: audience, job, asset, source evidence, interviews, languages, stakeholders, approvals, rights, timeline, review rounds, implementation, and measurement. Ask only questions that could change scope, risk, or schedule. Do not invent a market price. Request: [anonymized request].

Prompt 2 — Scope options

Draft three options that differ in actual work: editorial polish, research-backed asset, and campaign-ready package. For each, specify client inputs, deliverables, exclusions, assumptions, review rounds, dependencies, and handoff. Avoid guaranteed outcomes and do not use extra word count as the only reason for a higher tier. Context: [approved brief].

Prompt 3 — Estimate and post-project review

Break this project into estimated hours by task and identify uncertainty, client dependencies, rework risks, and change-order triggers. After delivery, compare estimate with actual time and explain deviations. Do not set a currency rate or claim that the fee is standard for a market. Data: [approved estimate and time log].

Stop-doing swipe file: Replace defensive price talk with scope clarity

Stop: “AI can’t write like a human.”

Use: “A tool can help produce a draft. This scope also includes [specific research, evidence review, and adaptation]. Here is what I will check and what I cannot guarantee.”

Stop: “My 1,000-word article costs [rate] because that is the market rate.”

Use: “For this brief, the fee covers [work], [deliverables], [review rounds], and [handoff]. The estimate assumes [inputs and timing].”

Stop: “I’ll make sure your landing page converts.”

Use: “I’ll deliver [copy and variants] against the approved brief. Conversion depends on offer, traffic, design, technical setup, and other factors outside this scope.”

Stop: “Just one more small revision,” with no scope boundary.

Use: “The included review rounds are complete. I can quote this additional change as [defined scope] once we confirm the new requirement.”

Hook–Value–Template: Respond to a price objection without arguing about AI

Hook: “You can get a first draft for free. What work needs to happen before it is ready to publish?”

Value: “A useful quote makes the research, proof, review, rights, revisions, and handoff visible, then lets the client choose the support they need.”

Template: “For [audience] and [business task], I propose [asset]. The fee covers [specific tasks and deliverables] based on [inputs, assumptions, timeline, and review rounds]. It excludes [scope boundaries]. You provide [dependencies]. I will verify [defined evidence]. I cannot guarantee [outcomes outside control]. If you only need [smaller task], we can scope that instead.”

Sources: Copywriting work, AI review, and market context

The ProCopywriters survey reflects its UK commercial-writer respondents and is not a pricing recommendation for MENA. BLS data describe U.S. occupational work and pay, not freelance copywriter rates in other markets. Use both only as bounded context; set a quote from the actual scope, local operating conditions, and your own cost and capacity model.