--- title: "Give AI a client file so email drafts start with the right context" description: "Without a client file, every AI email starts with questions and ends as a generic template. Compare two drafts and build the file that fixes it." canonical: "https://scalewithsearch.com/articles/ai-memory-before-and-after-email" 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) # Give AI a client file so email drafts start with the right context. You type: "Draft an email to John about the project delay." The assistant replies: "Who is John?" It asks every time, because nothing tells it. The comparison below uses a fictional client to show the difference a client file makes, and how to build one. ## Before: the question loop Here is the full exchange without a client file: **You:** "Draft an email to John about the project delay." **AI:** "I'd be happy to help. Could you tell me more about John and the project?" **You:** "John Miller. The website redesign for his accounting firm." **AI:** "What caused the delay, and when do you expect to deliver?" **You:** "The designer got sick. We're moving from Friday to next Tuesday." **AI:** "What tone would you like: formal or casual?" **You:** "John is relaxed. Casual but professional." Four rounds before any draft. Then you get this: > Hi John, > > I hope this email finds you well. I wanted to reach out regarding the website redesign project we've been working on together. > > Unfortunately, due to unforeseen circumstances with our design team, we're experiencing a slight delay in delivery. The new completion date will be next Tuesday instead of this Friday. > > I sincerely apologize for any inconvenience this may cause. Please let me know if you have any questions or concerns. > > Best regards It is acceptable, and it reads like AI. "I hope this email finds you well" opens it. "Unforeseen circumstances" is corporate filler. "Sincerely apologize" is a phrase you would never say to John. You spend three more minutes on edits. Total: about eight minutes for one email. ## After: the model reads the client file Same prompt. This time the assistant reads John's client file first. The file says: - client: John Miller, Miller & Associates CPA; - current project: website redesign; - original deadline: January 31; - communication style: direct, values transparency; - history: responded well to proactive updates; - context: his first website project, and he is anxious about the timeline. No follow-up questions. The draft: > Hey John, > > Quick update on the Miller & Associates site. Our designer came down with the flu, so we're moving your launch from Friday to Tuesday, February 4. That still puts us well ahead of your tax season rush. > > The extra days help us. We can test the contact form integration properly instead of rushing it. > > I'll send you the staging link Monday afternoon so you can look before we go live. > > Let me know if the new timeline creates any issues on your end. You read it, confirm the date, and send it. Total: under a minute. ## What made the difference The model did not get smarter. It got informed. This is the client file behind the second draft: ```markdown # John Miller - Miller & Associates CPA (fictional) current_project:: website redesign original_deadline:: 2026-01-31 new_launch_date:: 2026-02-04 contract_terms:: see contracts/miller-associates.md ## Communication preferences - Direct, no filler - Appreciates proactive updates - Responds to solutions, not apologies - Anxious about timelines (first website project) ## Key context - Tax season starts mid-February - Needs the site live before then for client inquiries - Values transparency over perfection ## Last interaction 2026-01-15: approved the design mockups. Wants the site to feel "modern but trustworthy." ``` The file is about 100 words. In this example it saves about seven minutes per email. If you email John twice a week, that is about an hour a month, for one client. ## The better draft is also a better decision The quality gap in the wording is obvious. The gap in judgment matters more. In the first version, you told the model to apologize, and it did. In the second, the file said John responds to solutions, not apologies. The draft framed the delay as time to test the contact form. You did not ask for that framing. The model applied a preference you had recorded. A model with the right context applies your recorded judgment. It cannot apply a preference that nobody wrote down. ## The follow-up two weeks later Two weeks later, you need to check in. Without a file, you type "Draft a follow-up to John," and the model asks what the last conversation was about. You search your sent folder, find the thread, copy the context, and explain the situation again. With a file, the model reads that the site launched on February 4 and that two weeks have passed. It drafts: > Hey John, > > The Miller & Associates site has been live for two weeks. How is it working for you? Are you getting the inquiries you expected? > > On my end, [page load time from the monitoring report], and contact form submissions are coming through cleanly. > > Want a quick call to review the analytics or make changes before tax season is in full swing? Notice the bracket. The file holds no load-time data, so the draft leaves a marked gap instead of a number. Add a rule to your root context file: any fact that is not in the client file becomes a bracketed placeholder. A model without that rule may invent a plausible figure, such as "under 2 seconds," and you would send a claim you never measured. ## Build the system You need two parts: 1. A client file: one Markdown file per client. 2. A way for the AI to read it at session start. In Claude Code, the root `CLAUDE.md` file points to the client files. A client file takes about ten minutes to write. Give every file the same fields: contact, current project, key dates, communication preferences, context, and last interaction. A consistent layout lets the model find each field, and it lets you spot a missing one at a glance. The loading setup is a one-time job. After that, every email draws on the same foundation: John's file, the Peterson account, the nonprofit project. Keep one file per client, and load only the client the email names. A shared file can put one client's facts into another client's email. The guide to [separating client context before an agent touches the work](/articles/separate-client-context-before-agent-work) explains why. The article on [what client data belongs in AI agent memory](/articles/client-data-in-ai-agent-memory) sets limits for personal details. ## Write the first file from your sent folder You do not need to invent the client file. Your sent folder holds most of it. 1. Open the last ten emails you sent to one client. 2. Note the tone you used: greeting, sentence length, and how you close. 3. Note every fact you restated more than once: deadlines, project names, who approves what. 4. Note every reaction the client showed: what got a fast reply, what got a question, what got silence. 5. Write those three lists into the file as communication preferences, key context, and last interaction. The first draft of the file takes about ten minutes for a client you email often. Then run one test. Start a fresh session and ask for a short check-in email to that client. Compare it with the last check-in you sent. Where the draft differs in tone or misses a fact, the file is missing a line. Add the line and ask again. ## Keep a person on the send A better draft does not remove the review. Before any client email leaves, check three things: 1. Every date and number matches the client file or a current source. 2. Every bracketed placeholder is filled or removed. 3. The tone fits what happened since the file was last updated. If an AI agent can send email on its own, decide in advance who approves each message. The guide on [who approves what an AI agent sends](/articles/who-approves-what-an-ai-agent-sends) shows how to bind the approval to the exact message. Update the file after each meaningful exchange: the new date, the decision, the change in mood. The file is only as current as its last entry. Freelancers who manage several clients can use the same pattern; the guide to [AI memory for freelancers](/articles/ai-memory-for-freelancers) adapts it for one person with many accounts. ## Related: AI memory - [Measure AI first-draft editing time and cut it with persistent context](/articles/ai-first-draft-accuracy) - [Set up Claude Code so every client draft starts in the right voice](/articles/claude-code-for-content-writers) - [Teach AI your content with context files instead of fine-tuning a model](/articles/how-to-train-ai-on-my-content) ---- ```text .|########||. .|########||. .|########||. |##||. .||##|. |##||. .||##|. |##||. .||##|. |#|. .|#|. |#|. .|#|. |#|. .|#|. |#| |#| |#| |#| |#| |#| .#| |#. .#| |#. .#| |#. .#. .#| .#. .#| .#. .#| |#. .#| |#. .#| |#. .#| |# ...... #| |# ...... #| |# ...... #| .# ||#########| #| .# ||#########| #| .# ||#########| #| .#. |######||######|. .#. .#. |######||######|. .#. .#. |######||######|. .#. #. .##|###|##|#|######| .# #. .##|###|##|#|######| .# #. .##|###|##|#|######| .# || |##|#||||||||||||#||#| || || |##|#||||||||||||#||#| || || |##|#||||||||||||#||#| || # |#||||||||||||||||||||#| #. # |#||||||||||||||||||||#| #. # |#||||||||||||||||||||#| #. || |#||||||||||||||||||||||#| || || |#||||||||||||||||||||||#| || || |#||||||||||||||||||||||#| || # .#||||||||||||||||||||||||#| # # .#||||||||||||||||||||||||#| # # .#||||||||||||||||||||||||#| # || ....|||#||||||##|#|||#|#||##||||....|. || || ....|||#||||||##|#|||#|#||##||||....|. || || ....|||#||||||##|#|||#|#||##||||....|. || #. . ....|# ....#||| ||| .#||| ....#. .# #. . ....|# ....#||| ||| .#||| ....#. .# #. . ....|# ....#||| ||| .#||| ....#. .# # . ||||#| .#####|| . .#| .#||| ||||# #. # . ||||#| .#####|| . .#| .#||| ||||# #. # . ||||#| .#####|| . .#| .#||| ||||# #. .| ||||| #. |#|||||. || #. |#||. .||||| || .| ||||| #. |#|||||. || #. |#||. .||||| || .| ||||| #. |#|||||. || #. |#||. .||||| || || |.... #. ....|#. |. ...|. ....|| || || |.... #. ....|#. |. ...|. ....|| || || |.... #. ....|#. |. ...|. ....|| || #. .||||||##||||||##||####|||||||#||||||#| .# #. .||||||##||||||##||####|||||||#||||||#| .# #. .||||||##||||||##||####|||||||#||||||#| .# # .||####|###########################|. # # .||####|###########################|. # # .||####|###########################|. # # .#||||||.#.|| # |. ..# #| #|||||#| # # .#||||||.#.|| # |. ..# #| #|||||#| # # .#||||||.#.|| # |. ..# #| #|||||#| # # .#|||||| .. |# ## |#| ...#|#|||#| # # .#|||||| .. |# ## |#| ...#|#|||#| # # .#|||||| .. |# ## |#| ...#|#|||#| # # .#|||#|# .# .#| #| ##|.#..#|||||#| # # .#|||#|# .# .#| #| ##|.#..#|||||#| # # .#|||#|# .# .#| #| ##|.#..#|||||#| # # .#||||||############|######|||||#| # # .#||||||############|######|||||#| # # .#||||||############|######|||||#| # # |...||#||||||#||||#|#||||||#|||||#||| |.| # # |...||#||||||#||||#|#||||||#|||||#||| |.| # # |...||#||||||#||||#|#||||||#|||||#||| |.| # #. .. ||||# .|||##|. |#| .|| || .|||#. #| # .# #. .. ||||# .|||##|. |#| .|| || .|||#. #| # .# #. .. ||||# .|||##|. |#| .|| || .|||#. #| # .# || | ...|| ...##| | #| .|. |. ##### .. .| || || | ...|| ...##| | #| .|. |. ##### .. .| || || | ...|| ...##| | #| .|. |. ##### .. .| || || ||||| || ||||#| . |. |. |#. ||||| |#| || || || ||||| || ||||#| . |. |. |#. ||||| |#| || || || ||||| || ||||#| . |. |. |#. ||||| |#| || || .# |.....#|....|#.||||.|||##.|#|....||.#||.#. #. .# |.....#|....|#.||||.|||##.|#|....||.#||.#. #. .# |.....#|....|#.||||.|||##.|#|....||.#||.#. #. #.|||||||#############################| ||| .# #.|||||||#############################| ||| .# #.|||||||#############################| ||| .# || ##||#||#||||||#||#||#||#||##. || || ##||#||#||||||#||#||#||#||##. || || ##||#||#||||||#||#||#||#||##. || # .#|||||||||#||#|||||||||||#| # # .#|||||||||#||#|||||||||||#| # # .#|||||||||#||#|||||||||||#| # || |#||||||||||||||||||||||#| || || |#||||||||||||||||||||||#| || || |#||||||||||||||||||||||#| || .# |#||||||||||||||||||||#| #. .# |#||||||||||||||||||||#| #. .# |#||||||||||||||||||||#| #. || |#||||||||||||||||||#| || || |#||||||||||||||||||#| || || |#||||||||||||||||||#| || #. .######|#|#########| .# #. .######|#|#########| .# #. .######|#|#########| .# .#. |#####||||#####| .#. .#. |#####||||#####| .#. .#. |#####||||#####| .#. |# ||########|| #. |# ||########|| #. |# ||########|| #. |# ...... #| |# ...... #| |# ...... #| |#. .#| |#. .#| |#. .#| |#. .#. |#. .#. |#. .#. .#| |#. .#| |#. .#| |#. |#| |#| |#| |#| |#| |#| .|#|. .|#|. .|#|. .|#|. .|#|. .|#|. |##||. .||##| |##||. .||##| |##||. .||##| .||########||. .||########||. .||########||. Scale With Search 2026 [scalewithsearch.com](https://scalewithsearch.com) ```