--- title: "Build an outreach file so AI writes cold emails from your prospect research" description: "Store prospect notes, angles by industry, email structures, proof points, and objection answers in one file, so AI outreach drafts are specific." canonical: "https://scalewithsearch.com/articles/ai-memory-for-sales-outreach" 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) # Build an outreach file so AI writes cold emails from your prospect research. You have 87 prospects on a target list. You ask an AI assistant for cold emails. It returns the same polished, generic message that every other seller now sends: "Hi [First Name], I noticed your company works in [Industry]. We help companies like yours with [Generic Value Prop]. Would love to chat about how we can help you achieve [Vague Outcome]." Prospects delete it unread. They have seen it many times this month. The assistant cannot personalize what it does not know. It does not have your prospect research, your proof, your objection answers, or your real value proposition. An outreach file gives it those. It is one Markdown file, `outreach.md`, that the assistant reads before it drafts. Every prospect and result below is a synthetic example. ## Know what a specific email needs Take the request: "Write a cold email to PROSPECT-031. She posted last week about poor mobile conversion." A specific email needs six things: - the prospect's industry, company size, and stated problem; - which proof point fits that prospect's profile; - your real results, with numbers and time frames you can support; - your email structure and tone; - your answers to common objections; - your call to action for this stage: a short call, a case study, or a resource. With these in the file, the draft names her problem, cites one relevant result, and uses your tone. You then check it and decide whether it goes out. ## Record what you learned about each prospect List more than names and companies. Record what your research found: stated problems, recent activity, tools in use, and company stage. ```markdown ## Active prospects ### PROSPECT-031: healthcare tech company - title: VP Marketing - company: Series B, 40 employees - problem: posted about mobile conversion issues on LinkedIn (Jan 15) - tools: WordPress, Google Analytics - recent activity: hired two marketers in Q4; scaling paid ads - best angle: mobile conversion - avoid: do not pitch a redesign; they launched a new site 6 months ago ### PROSPECT-032: B2B SaaS company - title: Head of Growth - company: bootstrapped, 15 employees - problem: demo request form converts at 18%; target is 25% or more - previous vendors: tried a landing page tool, cancelled after 3 months - best angle: form testing and A/B test experience - decision process: needs the CEO's approval; cost-conscious ``` Keep prospect notes to business facts that the email needs. [What client data belongs in AI agent memory](/articles/client-data-in-ai-agent-memory) applies the same limits to prospects. If your CRM already holds the contact history, point the file at the CRM record. Do not copy the whole history. [Customer conversation memory across tools](/articles/customer-conversation-memory-across-tools) shows that pattern. ## Match the angle to the industry The service is the same; the pitch changes. Healthcare buyers care about compliance. SaaS buyers care about trial conversion. The assistant needs to know which angle goes with which prospect type. ```markdown ## Angles by industry ### Healthcare tech - lead with: security, compliance, design that builds trust - stress: patient data protection, accessibility standards - proof first: medical device form project (form completion from 31% to 43%, compliance review passed) - avoid: aggressive conversion tactics, "growth hacking" language ### B2B SaaS - lead with: trial-to-paid conversion, demo request rate - stress: measurable outcomes, fast iteration, A/B tests - proof first: SaaS homepage project (6.8% conversion, demo requests up 34%) - avoid: long timelines, brand-building talk ### E-commerce - lead with: cart abandonment, mobile checkout - stress: revenue effect, average order value - proof first: retail cart recovery project (18% of abandoned carts recovered) - avoid: design talk without business outcomes ``` ## Store the structure for each email stage A first touch, a follow-up, and a breakup email each need a different structure. Store the pattern, not the exact copy. The assistant fits the pattern to each prospect. ```markdown ## Email structure ### First touch - subject: names their problem or recent activity - opening: shows the research (their post, hiring, or launch) - problem: their pain point in their words - proof: one relevant result with numbers - call to action: low commitment (send a case study, 15-minute call, share a resource) - length: 75 to 100 words - tone: conversational, no jargon, one question at most ### Follow-up (3 to 5 days after the first email) - one line that refers to the first email - NEW information: another result, an article, or a tool suggestion - the value in one sentence - a different call to action - length: 50 words at most ### Breakup (10 days after the last touch) - acknowledge they are busy or not interested - offer one useful resource with no ask - leave the door open, no pressure - length: 30 to 40 words ``` Track which subject lines get opened, with the numbers from your email tool. A line such as `"Saw you're hiring marketers": 38% open rate, 60 sends, March` is useful. A line with no count and no date is a guess. Some mail apps load images automatically, and that inflates open counts. Treat replies as the stronger signal. ## Keep proof points short and supported Sales proof is not a full case study. It leads with the outcome and adds enough context to be believable. ```markdown ## Proof points (synthetic examples, not client results) ### SaaS homepage project one line:: homepage conversion from 2.1% to 6.8% in 60 days client type:: B2B SaaS, 50 employees work:: homepage rebuilt, value proposition clarified, interactive demo added result:: 6.8% conversion; demo requests up 34%; trial signups up 12% timeline:: 8 weeks use for:: SaaS prospects; conversion projects; prospects who mention unclear messaging permission:: [date the client approved public use] source:: [analytics export and date] ### Medical device form project one line:: contact form completion from 31% to 43%; the form still passed compliance review client type:: healthcare tech, Series A, medical device manufacturer work:: form cut from 12 fields to 4; trust signals added; mobile layout fixed result:: completion from 31% to 43%; no security or compliance issues timeline:: 6 weeks use for:: healthcare prospects; form projects; compliance-conscious buyers permission:: [date] source:: [date] ``` A proof point in an email is a claim to a stranger. If the permission or source field is empty, the assistant does not use the proof point. [AI brand memory for content creation](/articles/ai-brand-memory-content-creation) explains how to keep claims like these current. ## Write answers to the objections you hear You have heard each objection before. Store your best answer, and the assistant adapts it to an email reply. ```markdown ## Objections ### "We just redesigned our site" Not a redesign pitch. Many conversion problems come from form friction or unclear messaging. Both can be fixed without a change to the overall design. The medical device project cut a form from 12 fields to 4. ### "We don't have budget right now" Understood. Would the retail cart recovery example help? That client was also unsure about budget. The 18% cart recovery paid back the project in 6 weeks. Happy to revisit in Q2 if the timing is better. ### "We need to see results first" Two options: I can share data from a similar project in your industry, or start with a free form audit that finds the specific issues. Which is more useful? ### "Send me some information" Would the SaaS example or the healthcare example fit your situation better? (Ask them to choose, so the reply has a next step.) ``` ## Compare a draft with and without the file Without the file, the request "Write a cold email to the VP Marketing at a healthcare tech company; we do web design and optimization" gets this: ```text Subject: Improve Your Website Performance Hi [First name], I hope this email finds you well. I noticed your company is doing interesting work in the healthcare technology space. We specialize in helping companies like yours optimize their websites for better performance and user experience. Our data-driven approach has helped numerous clients achieve their digital marketing goals. Would you be open to a brief call to discuss how we might be able to help improve your online presence? Best regards ``` It has no research, no proof, and no reason to reply. With the file, the request is "Write a cold email to PROSPECT-031; use the mobile conversion angle and the medical device proof point." The draft reads: ```text Subject: Your mobile conversion question [First name], I saw your post about mobile conversion rates. We worked on the same problem with a medical device company. Their contact form converted at 31% on mobile. After we cut it from 12 fields to 4 and added trust signals, it reached 43%, and the form still passed their compliance review. Worth a 15-minute call to see if the same issues affect your mobile pages? [Your name] ``` The subject line has no "Re:". This email is not a reply, and a "Re:" on a first message misleads the recipient about a prior thread. Put that rule in the file. ## Work the list in batches by angle Eighty-seven prospects do not need eighty-seven separate briefings. Sort the list by the angle that fits each one, and draft in batches. 1. Add an `angle::` line to each prospect entry: mobile conversion, form testing, cart recovery, or compliance. 2. Ask the assistant to list the prospects by angle, with the proof point each angle uses. 3. Draft one batch at a time, such as the twelve prospects on the cart recovery angle. 4. Read each draft against the prospect entry. Check that the opening cites that prospect's research, not another prospect's. 5. Send the batch, and log the sends with the date and the subject line. A batch of twelve drafts on one angle is faster to review than twelve drafts on twelve angles. You check the same proof point and the same structure each time. The prospect-specific lines are the only part that changes, and those are the lines to read closely. ## Test the file with ten prospects Run this check before you send anything from the file. 1. Pick ten prospects across at least three angles. 2. In a fresh session, ask for a first-touch email for each one, by prospect ID only. 3. Search each draft for "I hope this email finds you well," "companies like yours," and "Re:" in the subject line. Pass: none appear. 4. Check that each draft cites one proof point with a permission date and a source. Pass: every proof point has both fields filled in the file. 5. Check that each draft's opening line names a fact from that prospect's entry. Fail: the opening could apply to any prospect. A fail on step 5 is the common one. It means the prospect entry holds a title and a company but no research. Add the stated problem or the recent activity, and ask again. An email with no research is the generic email you started with, whatever the file says about tone. ## Grow the file from real sends Start with three things: 1. Document 10 to 15 prospects. 2. Add your three strongest proof points, with permission and source. 3. Write your email structure for each stage. Ask for drafts for those prospects. The gaps show at once; add the missing facts. After each batch, update the file: which subject lines got opens, which proof points got replies, and which objections came up. [How corrections become part of the system](/articles/corrections-become-system-memory) shows how to make each lesson change the next draft. Measure the change. Count drafts, edits, send volume, and replies before the file and after. Do not quote a response-rate gain you did not measure. ## Keep sending with a person and inside the law The assistant drafts. A person reads each email, checks each claim, and sends it. [Who approves what an AI agent sends](/articles/who-approves-what-an-ai-agent-sends) covers how to set that rule for a team. Cold email is regulated. In the United States, the CAN-SPAM Act bans deceptive subject lines and requires a valid postal address and a working opt-out. The law makes no exception for business-to-business email. The opt-out must work for at least 30 days after the send, and you must honor a request within 10 business days. In the UK and the EU, rules for unsolicited email and for prospect data differ again. Add the rules for your market to the file, so every draft carries the opt-out text and the address. ## Related: AI memory - [Draft proposals from service, proof, and discovery files so each one fits the buyer](/articles/ai-memory-before-and-after-proposals) - [Build a proposal memory file so AI drafts proposals in your format](/articles/ai-memory-for-proposal-writing) ---- ```text .|########||. .|########||. .|########||. |##||. .||##|. |##||. .||##|. |##||. .||##|. |#|. .|#|. |#|. .|#|. |#|. .|#|. |#| |#| |#| |#| |#| |#| .#| |#. .#| |#. .#| |#. .#. .#| .#. .#| .#. .#| |#. .#| |#. .#| |#. .#| |# ...... #| |# ...... #| |# ...... #| .# ||#########| #| .# ||#########| #| .# ||#########| #| .#. |######||######|. .#. .#. |######||######|. .#. .#. |######||######|. .#. #. .##|###|##|#|######| .# #. .##|###|##|#|######| .# #. .##|###|##|#|######| .# || |##|#||||||||||||#||#| || || |##|#||||||||||||#||#| || || |##|#||||||||||||#||#| || # |#||||||||||||||||||||#| #. # |#||||||||||||||||||||#| #. # |#||||||||||||||||||||#| #. || |#||||||||||||||||||||||#| || || |#||||||||||||||||||||||#| || || |#||||||||||||||||||||||#| || # .#||||||||||||||||||||||||#| # # .#||||||||||||||||||||||||#| # # .#||||||||||||||||||||||||#| # || ....|||#||||||##|#|||#|#||##||||....|. || || ....|||#||||||##|#|||#|#||##||||....|. || || ....|||#||||||##|#|||#|#||##||||....|. || #. . ....|# ....#||| ||| .#||| ....#. .# #. . ....|# ....#||| ||| .#||| ....#. .# #. . ....|# ....#||| ||| .#||| ....#. .# # . ||||#| .#####|| . .#| .#||| ||||# #. # . ||||#| .#####|| . .#| .#||| ||||# #. # . ||||#| .#####|| . .#| .#||| ||||# #. .| ||||| #. |#|||||. || #. |#||. .||||| || .| ||||| #. |#|||||. || #. |#||. .||||| || .| ||||| #. |#|||||. || #. |#||. .||||| || || |.... #. ....|#. |. ...|. ....|| || || |.... #. ....|#. |. ...|. ....|| || || |.... #. ....|#. |. ...|. ....|| || #. .||||||##||||||##||####|||||||#||||||#| .# #. .||||||##||||||##||####|||||||#||||||#| .# #. .||||||##||||||##||####|||||||#||||||#| .# # .||####|###########################|. # # .||####|###########################|. # # .||####|###########################|. # # .#||||||.#.|| # |. ..# #| #|||||#| # # .#||||||.#.|| # |. ..# #| #|||||#| # # .#||||||.#.|| # |. ..# #| #|||||#| # # .#|||||| .. |# ## |#| ...#|#|||#| # # .#|||||| .. |# ## |#| ...#|#|||#| # # .#|||||| .. |# ## |#| ...#|#|||#| # # .#|||#|# .# .#| #| ##|.#..#|||||#| # # .#|||#|# .# .#| #| ##|.#..#|||||#| # # .#|||#|# .# .#| #| ##|.#..#|||||#| # # .#||||||############|######|||||#| # # .#||||||############|######|||||#| # # .#||||||############|######|||||#| # # |...||#||||||#||||#|#||||||#|||||#||| |.| # # |...||#||||||#||||#|#||||||#|||||#||| |.| # # |...||#||||||#||||#|#||||||#|||||#||| |.| # #. .. ||||# .|||##|. |#| .|| || .|||#. #| # .# #. .. ||||# .|||##|. |#| .|| || .|||#. #| # .# #. .. ||||# .|||##|. |#| .|| || .|||#. #| # .# || | ...|| ...##| | #| .|. |. ##### .. .| || || | ...|| ...##| | #| .|. |. ##### .. .| || || | ...|| ...##| | #| .|. |. ##### .. .| || || ||||| || ||||#| . |. |. |#. ||||| |#| || || || ||||| || ||||#| . |. |. |#. ||||| |#| || || || ||||| || ||||#| . |. |. |#. ||||| |#| || || .# |.....#|....|#.||||.|||##.|#|....||.#||.#. #. .# |.....#|....|#.||||.|||##.|#|....||.#||.#. #. .# |.....#|....|#.||||.|||##.|#|....||.#||.#. #. #.|||||||#############################| ||| .# #.|||||||#############################| ||| .# #.|||||||#############################| ||| .# || ##||#||#||||||#||#||#||#||##. || || ##||#||#||||||#||#||#||#||##. || || ##||#||#||||||#||#||#||#||##. || # .#|||||||||#||#|||||||||||#| # # .#|||||||||#||#|||||||||||#| # # .#|||||||||#||#|||||||||||#| # || |#||||||||||||||||||||||#| || || |#||||||||||||||||||||||#| || || |#||||||||||||||||||||||#| || .# |#||||||||||||||||||||#| #. .# |#||||||||||||||||||||#| #. .# |#||||||||||||||||||||#| #. || |#||||||||||||||||||#| || || |#||||||||||||||||||#| || || |#||||||||||||||||||#| || #. .######|#|#########| .# #. .######|#|#########| .# #. .######|#|#########| .# .#. |#####||||#####| .#. .#. |#####||||#####| .#. .#. |#####||||#####| .#. |# ||########|| #. |# ||########|| #. |# ||########|| #. |# ...... #| |# ...... #| |# ...... #| |#. .#| |#. .#| |#. .#| |#. .#. |#. .#. |#. .#. .#| |#. .#| |#. .#| |#. |#| |#| |#| |#| |#| |#| .|#|. .|#|. .|#|. .|#|. .|#|. .|#|. |##||. .||##| |##||. .||##| |##||. .||##| .||########||. .||########||. .||########||. Scale With Search 2026 [scalewithsearch.com](https://scalewithsearch.com) ```