How to Edit Your Own Copy: A Six-Pass Process

Editing your own marketing copy works best when you stop treating it as one final cleanup. Review the draft in separate passes for reader orientation, meaning, evidence, language, sound, and test design.

Each pass answers a different question. The process can improve the draft’s clarity and defensibility, but it cannot guarantee better clicks, qualified responses, sales, retention, or profit.

Drafting asks, “What should I say?” Editing asks, “Does the existing text do its intended job?”

A keystroke analysis distinguished editing from drafting by patterns such as longer pauses, movement to existing parts of the text, and relatively little new typing. The participants were eighth-grade essay writers, so the findings do not establish how experienced commercial writers behave. They do support treating editing as a separate attention mode rather than continuing to draft while checking every sentence (ETS report).

Before you edit:

• Duplicate the draft so you can recover intentional language.
• Write down the intended reader, offer, and next action.
• Identify the evidence available for factual claims.
• Choose one editing objective at a time.
• Record major structural changes separately from line edits.

Do not begin by polishing the opening sentence. If the argument is incomplete or the promise exceeds the evidence, elegant wording will only conceal a larger problem.

The Six-Pass Copy Edit is an original practitioner decision aid, not an experimentally validated performance method.

1. Orientation: Can the intended reader recognize the subject and its relevance?
2. Meaning: Does the copy make one coherent, unambiguous argument?
3. Evidence: Does each claim stay within what its source supports?
4. Language: Is every sentence precise, economical, and consistent?
5. Sound: Does spoken copy remain clear when heard?
6. Test: Can you evaluate the consequential change without confusing correlation with causation?

Finish the objective of one pass before moving to the next. Stop a pass when its question has been answered throughout the draft—not when every sentence feels perfect.

The opening should help the intended reader answer three questions:

• Is this for someone like me?
• What problem or opportunity is being discussed?
• Why should I continue to the explanation?

A sampled proprietary corpus contained a lead pattern in which five sources established relevance through an escalating sequence of symptoms before introducing a product. That observation describes structure in a stratified convenience sample; it does not establish that the structure retains viewers or converts them.

The same inspection principle applies outside health copy. You can establish recognizable stakes before explaining a mechanism, but each escalation must remain proportionate and supportable.

Original hypothetical example

Before:

After an orientation pass:

The revision identifies the reader’s situation and supplies a next question: where does the reporting friction begin? It is clearer, but no performance conclusion follows from that edit.

Another sampled hook pattern delayed an explanation by creating curiosity around an overlooked component or process. Use that pattern as an inspection prompt:

• Is the missing information worth waiting for?
• Does the eventual explanation answer the question plainly?
• Does the tease exaggerate what will be revealed?

Curiosity becomes confusion when the draft keeps opening loops without closing them.

Read the draft silently from the reader’s position. Ignore rhythm for now and examine what each sentence actually permits the reader to conclude.

For every paragraph, identify:

1. The point being made.
2. The support offered for it.
3. The connection to the previous point.
4. The conclusion the reader is invited to draw.

A controlled proofreading experiment found no reliable overall self-generation effect, but word and function-word errors were harder to detect than misspellings. More attention and time were associated with better detection within that task. Because the participants were students proofreading controlled texts, the findings do not establish an editing method for professional persuasive copy (proofreading study).

The practical implication is modest: slow down around small words that change relationships. Terms such as “if,” “only,” “may,” “because,” and “after” can alter the promise or logic even when every word is spelled correctly.

Original hypothetical example

This sentence hides several questions. Does adoption cause errors to disappear? Does visibility prevent mistakes, or merely make them easier to find? Does “after” describe timing while “because” asserts causation?

A bounded revision might read:

When you know what a sentence means but cannot tell what it supports, mark it for the evidence pass instead of rewriting it repeatedly.

Create a claim ledger with three fields: the exact claim, the evidence attached to it, and the narrower conclusion that evidence can support.

Pay particular attention to mechanism claims. Explaining how something could work does not establish that it produces the promised outcome.

One mechanism pattern in the sampled corpus explicitly kept bioavailability evidence separate from evidence about joint comfort. It required distinct support for absorption and for the downstream outcome. This is an evidence-boundary example, not confirmation of any product benefit.

Original hypothetical example

Unsupported leap:

Separated wording:

For each claim, ask:

• Is this a fact, inference, prediction, opinion, or instruction?
• Does the source address the same population, product, conditions, and outcome?
• Have I turned “can,” “may,” or “is designed to” into “will”?
• Does a mechanism explanation quietly become a benefit promise?
• Would a reasonable reader infer more than the evidence supports?

If you cannot verify a material claim, narrow it, label the uncertainty, defer it for source review, or remove it.

Only now should you perform the conventional line edit. Preserve useful specificity and deliberate tension while removing language that obscures meaning.

Clarity

• Replace pronouns with unclear referents.
• Break sentences that contain competing ideas.
• Put conditions and exceptions next to the claims they limit.
• Define necessary technical terms on first use.

Specificity

• Replace vague quantities such as “many” or “dramatic” with supported wording.
• Name the action instead of relying on abstractions such as “optimization.”
• Distinguish a feature, mechanism, benefit, and measured outcome.

Consistency

• Use the same term for the same concept.
• Check names, capitalization, tense, and point of view.
• Make headings describe the material beneath them.

Claim scope

• Remove absolutes that the evidence does not justify.
• Keep comparisons tied to an identified alternative.
• Do not turn historical observations into current performance claims.

Stop when every change has a reason. “Shorter” is not automatically better if a cut removes a necessary qualification, transition, or explanation.

Read-aloud review is useful for questions about sound: difficult pronunciation, accidental repetition, crowded clauses, abrupt transitions, and places where a speaker cannot find a natural pause. It should not replace silent review of logic and evidence.

A counterbalanced study of multilingual university writers found that the value of reading aloud varied with English proficiency, with greater attentional demands for lower-proficiency writers. It did not test advertising delivery or conversion (Purdue study).

A separate study of university students associated read-aloud revision more with surface evaluation and silent revision more with semantic evaluation. Its small sample, short drafts, and fixed revision sequence limit broader conclusions (comparative revision study).

Original hypothetical VSL opening

A silent review finds an ambiguous promise: “what most teams overlook” is unsupported. An aloud review finds an overloaded sentence with no comfortable pause.

Revised:

The revision improves interpretability and delivery. It does not prove that viewers will listen longer or take action.

Separate editorial confidence from performance evidence. You can determine that a sentence is grammatically sound or that a claim matches its source without knowing whether the new version will improve a business outcome.

Fifteen large Facebook field experiments found that commonly used observational approaches often failed to recover the advertising effects estimated through randomized experiments. That work examined advertising measurement, not copy-editing methods, but it supports caution when attributing a metric change to an edit (Marketing Science article).

Use this hypothesis template:

Keep a record of concurrent changes to targeting, placement, offer, price, page speed, seasonality, and measurement. If several factors change together, the result cannot cleanly isolate the copy edit.

Daily Intel is designed to research ads, VSLs, transcripts, offer context, and funnel paths in one workflow, so you can compare a draft with relevant source material before editing its claims or structural assumptions. Record the source context, observed structure, editorial question, and evidence boundary; do not treat a recurring corpus pattern as proof of effectiveness.

A compact worksheet can use four columns:

Use this sequence for an ad, landing page, email, or VSL script:

1. Prepare: Preserve the original and state the reader, offer, action, and available evidence.
2. Orient: Confirm that the opening establishes relevance and creates a legitimate next question.
3. Read silently: Map the logic, implied promises, and missing steps.
4. Audit evidence: Verify each material claim and separate mechanisms from outcomes.
5. Line edit: Improve precision, consistency, and economy without deleting necessary context.
6. Read aloud: Mark difficult phrases and spoken transitions, then revise only the problems you hear.
7. Classify unresolved issues: Revise now, verify the source, preserve intentionally, request another reader, or move the question into a controlled test.
8. Write the hypothesis: Document the consequential change and how it will be evaluated.

Stop editing when further changes are matters of taste, when you need evidence you do not have, or when two plausible versions require comparison. Ask another reader for help when familiarity prevents you from explaining what the text says without supplying missing context from memory.

Should you edit copy silently or read it aloud?

Use both for different jobs. Silent review is better suited to meaning, logic, evidence, and implied promises; read-aloud review can expose rhythm, pronunciation, repetition, and crowded phrasing. Neither mode has been established as universally superior.

How many editing passes should marketing copy receive?

Use as many passes as the draft’s risk and complexity require. The six-pass sequence is a way to separate objectives, not a mandatory quota. A short ad may need brief passes, while a claim-heavy landing page may require source verification and another reviewer.

Can a cleaner edit improve conversion?

It may create a clearer variant worth testing, but cleanliness alone does not establish a performance effect. Define the change, comparison, outcome, guardrails, and decision rule before interpreting results.

How do you edit claims without overstating evidence?

Write the exact claim beside its source, then compare the source’s population, conditions, mechanism, and measured outcome with your wording. Narrow or remove any conclusion that travels beyond those boundaries.