--- title: "Write voice rules that make AI drafts sound like you" description: "AI writing sounds generic because the model writes toward the average, and explicit, testable voice rules fix it where tone adjectives fail." canonical: "https://scalewithsearch.com/articles/ai-writing-sounds-generic" date: "2026-01-28" modified: "2026-09-25" --- ## Site navigation - [Scale With Search](https://scalewithsearch.com/) - Real estate - Real estate - [Real estate](https://scalewithsearch.com/for/real-estate) - Work - Start here - [Send your brief](https://scalewithsearch.com/work#send-your-brief) - [Prepare your six-question brief](https://scalewithsearch.com/work#prepare-your-six-question-brief) - Build - [Site build, content library with SEO, signal desk](https://scalewithsearch.com/work) - For your business - Trades and home services - [Auto body and collision shops](https://scalewithsearch.com/for/auto-body-and-collision-shops) - [Foundation and home repair contractors](https://scalewithsearch.com/for/foundation-and-home-repair) - [Garage door and fencing contractors](https://scalewithsearch.com/for/garage-door-and-fencing-contractors) - [HVAC contractors](https://scalewithsearch.com/for/hvac-contractors) - [Janitorial and commercial cleaning companies](https://scalewithsearch.com/for/janitorial-and-commercial-cleaning) - [Locksmiths](https://scalewithsearch.com/for/locksmiths) - [Moving companies](https://scalewithsearch.com/for/moving-companies) - [Pest control companies](https://scalewithsearch.com/for/pest-control-companies) - [Plumbing and electrical contractors](https://scalewithsearch.com/for/plumbing-and-electrical-contractors) - [Restoration and water or fire damage companies](https://scalewithsearch.com/for/restoration-and-water-fire-damage) - [Roofing companies](https://scalewithsearch.com/for/roofing-companies) - [Towing companies](https://scalewithsearch.com/for/towing-companies) - [Tree services and landscaping companies](https://scalewithsearch.com/for/tree-services-and-landscaping) - [Solar installers](https://scalewithsearch.com/for/solar-installation) - [General contractors](https://scalewithsearch.com/for/general-contractors-and-construction) - [Paving, concrete, and flooring contractors](https://scalewithsearch.com/for/paving) - Practices and professional services - [Bookkeeping and tax practices](https://scalewithsearch.com/for/bookkeeping-and-tax-practices) - [Dental practices](https://scalewithsearch.com/for/dental-practices) - [Family and criminal defense law firms](https://scalewithsearch.com/for/family-and-criminal-defense-law-firms) - [Med spas and aesthetics practices](https://scalewithsearch.com/for/med-spas-and-aesthetics) - [Personal injury law firms](https://scalewithsearch.com/for/personal-injury-law-firms) - [Veterinary clinics](https://scalewithsearch.com/for/veterinary-clinics) - [Gyms and fitness studios](https://scalewithsearch.com/for/fitness) - [Therapy and outpatient health practices](https://scalewithsearch.com/for/therapy-and-outpatient-health) - [Medical billing companies](https://scalewithsearch.com/for/medical-billing) - [Insurance agencies](https://scalewithsearch.com/for/insurance-agencies) - [Financial advisors](https://scalewithsearch.com/for/financial-advisors) - [Property management companies](https://scalewithsearch.com/for/property-management) - [Recruiting and staffing agencies](https://scalewithsearch.com/for/recruiting-and-staffing) - [Architects and interior designers](https://scalewithsearch.com/for/architects-and-interior-designers) - [Logistics and supply chain companies](https://scalewithsearch.com/for/logistics-and-supply-chain) - Agencies, MSPs, and manufacturing - [IT and managed service providers](https://scalewithsearch.com/for/it-and-managed-service-providers) - [Machine shops and precision manufacturers](https://scalewithsearch.com/for/machine-shops-and-precision-manufacturing) - [Marketing agencies and freelancers](https://scalewithsearch.com/for/marketing-agencies-and-freelancers) - [SEO agencies and consultants](https://scalewithsearch.com/for/seo-agencies-and-consultants) - [Small manufacturers and fabricators](https://scalewithsearch.com/for/small-manufacturers-and-fabricators) - Restaurants, shops, studios, and nonprofits - [Restaurants and hospitality businesses](https://scalewithsearch.com/for/restaurants-and-hospitality) - [Retail stores and ecommerce sellers](https://scalewithsearch.com/for/retail-and-ecommerce) - [Photographers, event planners, and travel agents](https://scalewithsearch.com/for/photographers) - [Churches and nonprofits](https://scalewithsearch.com/for/churches-and-nonprofits) - [All industries](https://scalewithsearch.com/for/) - Learn - For your office - [Office job guides](https://scalewithsearch.com/guides/) - [Browser calculators](https://scalewithsearch.com/tools/) - Start here - [How it works](https://scalewithsearch.com/how-it-works) - [Free Starter Kit](https://scalewithsearch.com/kit/business-memory-starter-kit.zip) - [Synthetic specimen](https://scalewithsearch.com/specimen/working-session-specimen.zip) - Guides - [The Complete Guide to Business Memory for AI Agents](https://scalewithsearch.com/articles/business-memory-for-ai-agents-guide) - [The Complete Small-Business Guide to AI Agent Governance](https://scalewithsearch.com/articles/ai-agent-governance-guide-small-business) - [The Complete Guide to Leaving Vendor AI Memory](https://scalewithsearch.com/articles/leaving-vendor-ai-memory-guide) - Articles by cluster - [Business memory](https://scalewithsearch.com/articles/business-memory-for-ai-agents-guide) - [Agent governance](https://scalewithsearch.com/articles/ai-agent-governance-guide-small-business) - [Migration and ownership](https://scalewithsearch.com/articles/leaving-vendor-ai-memory-guide) - For machines - [llms.txt](https://scalewithsearch.com/llms.txt) - [llms-full.txt](https://scalewithsearch.com/llms-full.txt) - [Machine view](https://scalewithsearch.com/?view=machine) - Company - Evidence - [Proof](https://scalewithsearch.com/proof) - Company - [About](https://scalewithsearch.com/about) # Write voice rules that make AI drafts sound like you. You ask an AI assistant for an email. It returns three paragraphs that could fit any business, any client, and any situation. The tone is "professional" in the way bulk marketing email is professional: polished, empty, and forgettable. You ask for a blog post. It returns the same bullet structure as every other AI-written article. The transitions, the emphatic words, and the rhythm all tell a reader that a model wrote it. The cause is not a weak model. A model without specific direction writes toward the average of its training data. Most of that data is ordinary, so the default output is ordinary too, delivered with full confidence. ## Count what generic writing costs Generic output needs a rewrite. You read the draft, see that it sounds like everyone else, and spend 20 minutes editing it into something usable. The time the draft saved goes into the edit. Generic writing also changes how readers treat you. When your emails sound like everyone else's, clients skim them. When your content follows the same pattern as the AI content in every industry, your expertise is no longer visible. The tool meant to extend your voice flattens it. Then a second problem follows. You no longer use AI for work that matters. Quick drafts are fine. Proposals, client messages, and anything with your name on it, you write yourself. The assistant ends up on low-stakes tasks only. ## See why the default voice is generic Models learn patterns from training data. That data includes very large amounts of corporate websites, marketing email, business blogs, and professional documents. Most of it follows one safe template: friendly but not personal, informative but not specific, helpful but not direct. The model reproduces those patterns. Ask for a professional email, and it writes toward the statistical center of professional email. That center is dull by definition. Anything distinctive is uncommon, so the model rates it as less likely. This is why drafts open with "I hope this email finds you well," "I wanted to reach out about," and "Please don't hesitate to contact me." Those phrases appear in huge numbers of training examples. They are central to professional email, and they carry no information. ## Know which fixes do not work **Tone adjectives.** Most people add instructions such as "write in a casual tone," "be more direct," or "sound confident." The output moves a little but stays in the same range. "Casual" drops some formality and keeps the structure. "Direct" shortens sentences and keeps the empty phrases. The draft still needs heavy editing. **Pasted samples.** Others paste an email they wrote and say "write in this style." The model copies surface features, such as sentence length and paragraph shape, and misses the voice underneath. The result reads like an imitation of you. **Persona descriptions.** A custom GPT with a personality description helps more. "Be conversational and avoid corporate jargon" beats nothing. Without explicit rules on what to avoid and what to prefer, the model drifts back to its defaults. The drift is worst on long or complex drafts. [ChatGPT custom instructions are not hard constraints](/articles/chatgpt-custom-instructions-not-hard-constraints) explains why a stated preference is not a control. ## Replace descriptions with rules A voice guide is a list of explicit rules about how you write. It does not say "be authentic" or "sound professional." It states concrete, testable rules: banned words, banned phrases, structure preferences, terminology, and tone rules with examples. Put the rules in a context file that the assistant reads at the start of each session. For Claude Code, that file is `CLAUDE.md`. The assistant then follows them in each draft. A banned word stops appearing. A banned phrase gets your preferred alternative. The structure follows your stated preferences. This works because an explicit rule overrides the statistical pull. The model reaches for "leverage" because the word appears in countless business documents. Your guide bans "leverage," so the model picks another word. Models still slip, especially in long drafts. Run a text search for the banned list on every draft that matters. A rule you can search for is a rule you can enforce. ## Put four kinds of rule in the guide **Banned words.** List the jargon and AI-pattern words you want gone: "synergy," "paradigm," "robust," "leverage," "ecosystem," "cutting-edge." Write them all down. **Banned phrases.** List the patterns that mark generic writing: "In today's world," "It's worth noting," "The fact of the matter is," "At the end of the day," "Game-changing," "Next-level." The assistant then routes around them. **Structure preferences.** If you dislike bullet points, say so. If you want short paragraphs, give the maximum sentence count. If you always open with a direct statement instead of a preamble, write that rule. The assistant follows structure rules as closely as word lists. **Positive patterns.** Full sample documents transfer poorly, because the model copies their surface. Name the patterns you want instead: "Start emails with a one-sentence reason for writing." "Use questions as section transitions." "End with one concrete next step, not a vague offer to help." A short guide might look like this: ```markdown ## Voice guide banned_words:: synergy, paradigm, robust, leverage, ecosystem, cutting-edge banned_phrases:: "In today's world", "It's worth noting", "At the end of the day", "I hope this email finds you well", "reaching out" structure:: - Open with a direct statement. No preamble. - Paragraphs: three sentences at most. - No bullet lists in emails. - No summary paragraph at the end. patterns:: - Emails start with one sentence on why you are writing. - Articles use questions as section transitions. - Every message ends with one concrete next step. ``` For a full brand record that keeps these rules apart from changing facts and claims, see [AI brand memory for content creation](/articles/ai-brand-memory-content-creation). ## See the rules at work You ask Claude to draft a client email. The voice guide loads: banned words, banned phrases, structure rules. The draft opens with a direct statement, uses your terms, contains no jargon, and ends with a specific next step. It has no "I hope this finds you well," no "reaching out," and no empty courtesy. It sounds like you because it follows your rules. You ask for a blog post. The guide bans bullet-point rhythm and tidy concluding paragraphs, and it asks for varied sentence structure. The structure varies, and the ideas move without formula. The draft carries fewer of the marks that tell a reader a model wrote it, because the model worked inside your rules. ## Grow the guide from your edits The first version of the guide will miss things. That is expected. Each time you edit a draft for voice, ask whether the edit reflects a rule. If you make the same edit twice, write the rule down. [How corrections become part of the system](/articles/corrections-become-system-memory) shows how to make each fix reach the next draft instead of only the current one. Test the guide on a fixed set of tasks, such as one email, one post, and one proposal section. Run them after each change to the guide. Search the output for banned words and phrases, and read for the patterns. If a rule no longer holds after a model update, you will see it in the fixed set before a client does. ## Know what the guide changes over time A voice guide makes AI-assisted writing consistent. Every email follows the same rules, and every document matches the same preferences. The output sounds like it came from you, not from a generic assistant. Over months, clients and readers see the same voice in every message, including AI-assisted ones. Your messages no longer sound like the rest of their inbox. The time saving becomes real at that point. The 20-minute tone edit shrinks, because the draft already follows your rules. The first draft sounds like you because you wrote down, in explicit and testable terms, what "sounds like you" means. [A brief is not a prompt](/articles/brief-vs-prompt) shows how to attach that guide to each recurring writing job. ## Related: AI memory - [Choose an AI for business writing that keeps your voice across projects](/articles/best-ai-for-business-writing) - [Compare ChatGPT and Claude for writing by testing voice drift across sessions](/articles/chatgpt-vs-claude-for-writing) - [Set up Claude Code so every client draft starts in the right voice](/articles/claude-code-for-content-writers) ---- ```text .|########||. .|########||. .|########||. |##||. .||##|. |##||. .||##|. |##||. .||##|. |#|. .|#|. |#|. .|#|. |#|. .|#|. |#| |#| |#| |#| |#| |#| .#| |#. .#| |#. .#| |#. .#. .#| .#. .#| .#. .#| |#. .#| |#. .#| |#. .#| |# ...... #| |# ...... #| |# ...... #| .# ||#########| #| .# ||#########| #| .# ||#########| #| .#. |######||######|. .#. .#. |######||######|. .#. .#. |######||######|. .#. #. .##|###|##|#|######| .# #. .##|###|##|#|######| .# #. .##|###|##|#|######| .# || |##|#||||||||||||#||#| || || |##|#||||||||||||#||#| || || |##|#||||||||||||#||#| || # |#||||||||||||||||||||#| #. # |#||||||||||||||||||||#| #. # |#||||||||||||||||||||#| #. || |#||||||||||||||||||||||#| || || |#||||||||||||||||||||||#| || || |#||||||||||||||||||||||#| || # .#||||||||||||||||||||||||#| # # .#||||||||||||||||||||||||#| # # .#||||||||||||||||||||||||#| # || ....|||#||||||##|#|||#|#||##||||....|. || || ....|||#||||||##|#|||#|#||##||||....|. || || ....|||#||||||##|#|||#|#||##||||....|. || #. . ....|# ....#||| ||| .#||| ....#. .# #. . ....|# ....#||| ||| .#||| ....#. .# #. . ....|# ....#||| ||| .#||| ....#. .# # . ||||#| .#####|| . .#| .#||| ||||# #. # . ||||#| .#####|| . .#| .#||| ||||# #. # . ||||#| .#####|| . .#| .#||| ||||# #. .| ||||| #. |#|||||. || #. |#||. .||||| || .| ||||| #. |#|||||. || #. |#||. .||||| || .| ||||| #. |#|||||. || #. |#||. .||||| || || |.... #. ....|#. |. ...|. ....|| || || |.... #. ....|#. |. ...|. ....|| || || |.... #. ....|#. |. ...|. ....|| || #. .||||||##||||||##||####|||||||#||||||#| .# #. .||||||##||||||##||####|||||||#||||||#| .# #. .||||||##||||||##||####|||||||#||||||#| .# # .||####|###########################|. # # .||####|###########################|. # # .||####|###########################|. # # .#||||||.#.|| # |. ..# #| #|||||#| # # .#||||||.#.|| # |. ..# #| #|||||#| # # .#||||||.#.|| # |. ..# #| #|||||#| # # .#|||||| .. |# ## |#| ...#|#|||#| # # .#|||||| .. |# ## |#| ...#|#|||#| # # .#|||||| .. |# ## |#| ...#|#|||#| # # .#|||#|# .# .#| #| ##|.#..#|||||#| # # .#|||#|# .# .#| #| ##|.#..#|||||#| # # .#|||#|# .# .#| #| ##|.#..#|||||#| # # .#||||||############|######|||||#| # # .#||||||############|######|||||#| # # .#||||||############|######|||||#| # # |...||#||||||#||||#|#||||||#|||||#||| |.| # # |...||#||||||#||||#|#||||||#|||||#||| |.| # # |...||#||||||#||||#|#||||||#|||||#||| |.| # #. .. ||||# .|||##|. |#| .|| || .|||#. #| # .# #. .. ||||# .|||##|. |#| .|| || .|||#. #| # .# #. .. ||||# .|||##|. |#| .|| || .|||#. #| # .# || | ...|| ...##| | #| .|. |. ##### .. .| || || | ...|| ...##| | #| .|. |. ##### .. .| || || | ...|| ...##| | #| .|. |. ##### .. .| || || ||||| || ||||#| . |. |. |#. ||||| |#| || || || ||||| || ||||#| . |. |. |#. ||||| |#| || || || ||||| || ||||#| . |. |. |#. ||||| |#| || || .# |.....#|....|#.||||.|||##.|#|....||.#||.#. #. .# |.....#|....|#.||||.|||##.|#|....||.#||.#. #. .# |.....#|....|#.||||.|||##.|#|....||.#||.#. #. #.|||||||#############################| ||| .# #.|||||||#############################| ||| .# #.|||||||#############################| ||| .# || ##||#||#||||||#||#||#||#||##. || || ##||#||#||||||#||#||#||#||##. || || ##||#||#||||||#||#||#||#||##. || # .#|||||||||#||#|||||||||||#| # # .#|||||||||#||#|||||||||||#| # # .#|||||||||#||#|||||||||||#| # || |#||||||||||||||||||||||#| || || |#||||||||||||||||||||||#| || || |#||||||||||||||||||||||#| || .# |#||||||||||||||||||||#| #. .# |#||||||||||||||||||||#| #. .# |#||||||||||||||||||||#| #. || |#||||||||||||||||||#| || || |#||||||||||||||||||#| || || |#||||||||||||||||||#| || #. .######|#|#########| .# #. .######|#|#########| .# #. .######|#|#########| .# .#. |#####||||#####| .#. .#. |#####||||#####| .#. .#. |#####||||#####| .#. |# ||########|| #. |# ||########|| #. |# ||########|| #. |# ...... #| |# ...... #| |# ...... #| |#. .#| |#. .#| |#. .#| |#. .#. |#. .#. |#. .#. .#| |#. .#| |#. .#| |#. |#| |#| |#| |#| |#| |#| .|#|. .|#|. .|#|. .|#|. .|#|. .|#|. |##||. .||##| |##||. .||##| |##||. .||##| .||########||. .||########||. .||########||. Scale With Search 2026 [scalewithsearch.com](https://scalewithsearch.com) ```