TL;DR

Ann Handley's Everybody Writes (second edition, Wiley, 2022) treats writing as a process of thinking, rewriting and empathy for the reader. As AI makes fluent text cheap, those principles become the differentiator: the ugly first draft where you work out what you think, the rewrite from the reader's side, a consistent brand voice, and careful sourcing. Handley's own test is simple: use AI freely for how-to information, but be cautious wherever the writer's own insight is the point.

Who Ann Handley is, and what Everybody Writes argues

Ann Handley is a principal at MarketingProfs, often described as the world's first Chief Content Officer, and a Wall Street Journal bestselling author. She also writes Total Annarchy, a fortnightly newsletter on writing and marketing.

Her best-known book, Everybody Writes: Your New and Improved Go-To Guide to Creating Ridiculously Good Content, appeared in a revised second edition from Wiley in October 2022. As the title suggests, its starting point is that in a digital business almost everyone writes, from the founder's LinkedIn post to the support team's emails, so writing well is a business skill rather than a specialist one.

The book's structure shows where her emphasis lies. Part I sets out writing rules, built around a step-by-step "Writing GPS" framework and chapters titled "Embrace The Ugly First Draft", "Cross Out the Wrong Words", "Swap Places with Your Reader" and "Develop Pathological Empathy". Later parts cover grammar and usage, brand voice ("Voice Doesn't Change, Tone Does"), storytelling, and publishing rules on credibility, sourcing and fact-checking.

Why her principles matter more in an age of AI content

When anyone can produce competent-sounding paragraphs in seconds, fluency stops being a signal of quality. What remains scarce is what Handley's book focuses on: knowing what you think, seeing the topic from the reader's side, sounding like yourself, and being right.

Handley has written about AI directly on her blog. Her position is not anti-technology. In "When To Use AI: 21 Good Uses and 10 Terrible Uses" she describes using AI to unstick workflow, improve efficiency and take on dull parts of the job. But she draws a line at the writing itself, because for her, writing is how she works out what she thinks.

"I don't use AI to write." — Ann Handley, annhandley.com

In "Should We Use AI to Write First Drafts?" she offers two questions. Is the goal how-to information? If so, particularly for SEO-driven content, AI can reasonably produce an outline or first draft. Does the writer matter? If the piece depends on memorable, personal or persuasive insight, she suggests limiting AI to research, if using it at all, because an AI-shaped first draft can steer your thinking without you noticing.

In "AI Slop, Accusations, And A Better Way" she argues that hunting for supposed "AI tells" in punctuation or rhythm is misguided, since models learned those patterns from people. The better question is whether doing the work yourself would help you grow or make something that matters.

Six Handley principles, applied to AI-era marketing

  1. Keep the ugly first draft human when the thinking matters. For thought leadership, opinion pieces and founder posts, write the rough version yourself. Use AI afterwards to tighten, check structure or suggest alternatives.
  2. Swap places with your reader. Before publishing, reread every piece, human or AI-assisted, as your customer. What do they need to know? What would make them stop reading? AI drafts tend to be generic; this step forces specificity.
  3. Practise empathy, not just accuracy. Handley's "pathological empathy" chapter title is a reminder that good content starts from the customer's problem, not the company's message. Brief AI with real customer language from reviews, support tickets and calls.
  4. Write down your brand voice. Her distinction between voice (constant) and tone (situational) is the most practical AI brief you can give. Document three or four voice traits with examples, and include them in every prompt and review checklist.
  5. Cross out the wrong words. AI output often carries filler, hedging and buzzwords. Edit ruthlessly: cut the throat-clearing opening, replace vague verbs, remove claims you cannot support.
  6. Treat sourcing and fact-checking as non-negotiable. The book gives publishing rules on credibility and sourcing their own section. With AI capable of inventing facts and citations, every statistic, quote and claim needs a source someone has actually opened.

A practical workflow for marketing teams

  1. Classify the piece. Is it how-to and informational, or does the writer matter? Use Handley's two questions.
  2. For how-to content: allow AI outlines and drafts, then apply a human edit for voice, accuracy and reader relevance.
  3. For insight content: the subject-matter expert writes or dictates the ugly first draft; AI assists only with research and polishing.
  4. Run a reader pass. Someone who did not write it reads it as the customer and marks anything vague, generic or unproven.
  5. Run a source pass. Every factual claim links to a source; anything unverifiable is cut.
  6. Sign it. A named person takes responsibility for each piece. This also supports transparency expectations from search engines and regulators.

Limits and criticism

Handley's stance reflects her own craft as a professional writer, and not every business has the time or skills to write every first draft by hand. Her own framework acknowledges this by treating how-to content differently. There is also a cost question: human-led drafting is slower, so teams need to be deliberate about which pieces truly need a human voice. Finally, the principles are about quality, not about rules; they will not by themselves satisfy legal or platform requirements on AI disclosure.

The deeper risk Handley points to in "Don't Outsource This Part of Writing" is losing what makes human writing distinctive: if more and more text is machine-generated, the quirks of how real people write fade, and human words become scarcer and more valuable. Her advice in the AI slop post sums it up: "if you skip the struggle, you miss the growth".

Key takeaways
  • Everybody Writes (2nd edition, 2022) treats writing as thinking, rewriting and empathy for the reader.
  • Handley uses AI for workflow and efficiency, but not to do her writing.
  • Her test: AI drafts are fine for how-to information; be careful when the writer's insight matters.
  • A documented brand voice, a reader pass and a source pass make AI-assisted content worth reading.

FAQ

Is Ann Handley against using AI for content?
No. She lists many good uses of AI and says it can help with workflow and efficiency. Her caution is about handing over the thinking, especially in the first draft of pieces where the writer's perspective is the value.
What is the "ugly first draft"?
It is a chapter title in Everybody Writes and shorthand for a rough first version in which you work out what you think, before shaping it for the reader in later drafts.
How do we keep AI-assisted content on brand?
Document your brand voice with a few clear traits and examples, include it in every prompt, and have a human edit each piece against that voice before publishing.

Sources

Content that still sounds like you

JRK can help you set up a brand voice guide and an AI-assisted content workflow that keeps human insight, accuracy and your reader at the centre.

Book a free growth session →

Want this working for your brand?

Book a free 30-minute growth session. We'll look at your funnel and tell you what we'd do first — honestly.

Book a growth session →