30+ AI Prompts for Sales Templates for Prospecting, Outreach, and Closing

Quick Reference: AI Prompts for Sales by Category

‍

‍

That's 34 prompts across 8 categories. If you're mid-sequence with a cold prospect right now, skip straight to that section. If you're heads-down on Q3 pipeline review, jump to the bottom.

‍

AI Prompts for Sales Prospecting

‍

Most reps use AI for prospecting the lazy way: "tell me about this company." That gets you a Wikipedia summary. The better move is forcing the model to connect what it finds to your specific offer, not just describe the company in a vacuum.

‍

Best used with: ChatGPT, Claude, Perplexity, or an AI layer inside a tool like Apollo or Clay.

‍

Prompt 1 — The "Why Now" Brief

‍

"Act as a B2B researcher who's paid to find reasons a deal should happen this quarter, not eventually. Look at [company name] and tell me: one thing changing inside this company right now that creates urgency for [your solution category], the title most likely to feel that urgency first, and one detail I could reference in an opener that proves I actually looked at their business instead of running a mail merge. Output in 100 words or less, no generic filler like 'growing company' or 'innovative team.'"

‍

Prompt 2 — Reverse-Engineer the Trigger

‍

"You're a sales strategist who's tired of vague 'trigger events.' Based on this: [paste news, funding, hire, or product update], work backward and tell me the actual operational problem this event probably creates for [target title] in the next 90 days. Then tell me if that problem connects to [your product] or if I'm forcing a fit that isn't there. Be honest if it's a stretch."

‍

Prompt 3 — ICP From Pattern, Not Guesswork

‍

"Act as a GTM analyst. Here are descriptions of my 5 best customers: [paste details]. Don't just list traits, find the pattern connecting them. What's the one thing true of all 5 that isn't true of my worst-fit customers? Give me a must-have, a nice-to-have, and a red flag that should make me walk away from a prospect even if they technically match on paper."

‍

Prompt 4 — Prioritization With Teeth

‍

"You're a sales ops lead building a call list for a rep who can only make 20 calls today out of 200 leads. Using these criteria: [your ICP filters], rank my top 20 and tell me, for each, the single strongest reason to call them before anyone else on the list. If two leads look identical, tell me that instead of forcing a fake distinction."

‍

Prompt 5 — Competitive Angle, Not Competitor Trash Talk

‍

"Act as a positioning strategist, not a hype machine. Here's how [competitor] pitches [target buyer]. Tell me the one claim in their pitch that's probably true and I shouldn't try to argue against, and the one gap in their approach that [my solution] genuinely closes better. I don't want a list of reasons we're better, I want the single honest wedge that would make a smart buyer pause."

‍

‍

AI Prompts for Cold Email Sequences

‍

Signal-personalized outreach is landing 15–25% reply rates against a 3–5% industry baseline for cold email, and the gap almost always comes down to specificity and constraint, not cleverness. Every prompt below forces the model to cut, not add.

‍

Prompt 6 — The Anti-Template Opener

‍

"Act as a cold email copywriter who's been told to stop sounding like a cold email copywriter. Write an email to [job title] at [company type]. Context: [their likely pain in one sentence]. Rules: under 90 words, no question in the subject line, no sentence starts with 'I,' no phrase resembling 'I noticed' or 'I hope,' one specific detail proving research, one CTA that isn't 'do you have 15 minutes.' Make the CTA feel low-effort to say yes to."

‍

Prompt 7 — Trigger-to-Relevance Bridge‍‍

"Write a cold email that connects this trigger: [funding round, new hire, expansion, etc.] to a consequence the reader is probably dealing with right now, not a benefit of my product. Structure: trigger, consequence, one-line relevance, soft CTA. Under 75 words. Don't mention my product by name until the second-to-last sentence."

‍

Prompt 8 — The Sequence That Escalates Specificity, Not Pressure

‍

"Build a 4-email sequence for [job title] at [company size/industry]. Each email should reveal a more specific piece of relevance than the last, not repeat the ask louder. Email 1: one shared observation. Email 2: one proof point tied to a company like theirs. Email 3: preempt the objection they're most likely silently having. Email 4: a genuine exit, not a guilt trip. Each under 90 words. Give me subject lines that could pass as an internal email, not a marketing send."

‍

Prompt 9 — Deliberate Pattern Break

‍

"Write a cold email that does one specific unconventional thing: either asks a single sharp question with no pitch attached, states something counterintuitive about [their industry], or opens by telling them why this might be the wrong email for them. Target: [job title]. Pain: [one sentence]. Under 55 words. It should read like a text from a smart colleague, not a marketing email."

‍

Prompt 10 — Re-Engagement Without Apology

‍

"Write a re-engagement email for someone who went quiet after [X touches]. Don't apologize for following up and don't say 'just checking in.' Instead, bring something new: a changed circumstance, a sharper insight, or a different angle entirely. Acknowledge the silence in one sentence max, then move on. Under 85 words."

‍

AI Prompts for LinkedIn Outreach

‍

LinkedIn punishes AI-sounding copy harder than email does, because the reader can see your face and your job title right next to the message. These prompts are built to sound like a person typing, not a sequence firing.

‍

Prompt 11 — Connection Request That Isn't a Pitch in Disguise

‍

"Write a LinkedIn connection request to [job title] at [company type]. Reference: [one specific, real detail from their profile or recent activity]. This should read like genuine curiosity from a peer, with zero mention of what I sell. Under 280 characters. If you can't make it sound genuine without pitching, tell me and give me an alternative angle instead."

‍

Prompt 12 — First DM That Earns a Reply, Not a Meeting

‍

"Write the first message to send right after a LinkedIn connection is accepted. Context: [your role, their role, one relevant pain point]. The goal is a conversation, not a booked call. Ask one real question you'd actually want the answer to. Under 80 words, no bullet points, nothing that reads like a template with their name dropped in."

‍

‍

Prompt 13 — InMail That Respects the Reader's Time

‍

"Write a LinkedIn InMail to [job title] at [company]. Subject under 7 words, no punctuation gimmicks. Body under 130 words. Context: [their likely pain], [your relevant angle in one sentence]. One CTA only. Write it like you'd be embarrassed if a peer read it and thought it sounded like a vendor."

‍

Prompt 14 — Comment Strategy That Builds Recognition Before You Ever DM

‍

"Give me 3 comments I could leave on posts from [target persona, e.g. VP Sales at Series B SaaS companies]. Each should add a genuine, slightly contrarian or specific point, not agreement plus emoji. No self-promotion. Each under 40 words. The goal is that this person recognizes my name three weeks from now before I ever send a DM."

‍

AI Prompts for Follow-Up Messages

‍

The follow-up prompts below are built around one rule: every message needs a reason to exist beyond "it's been a while." If the prompt can't produce a new angle, it shouldn't produce an email.

‍

Prompt 15 — Silence-Breaker With a New Angle, Not a Nudge

‍

"Write a follow-up to a prospect who hasn't responded to my last 2 emails about [solution]. I need a genuinely new angle, not a rephrase: [a different value point, recent case study, or fresh stat]. Under 55 words. If you can't find a real new angle from what I've given you, tell me instead of forcing one."

‍

Prompt 16 — Post-Demo Follow-Up That Shows You Listened

‍

"Write a follow-up after a demo with [job title] at [company]. Reference the specific pain they raised: [pain point], and connect it directly to what they saw, not a generic recap. Include a clear next step with a real deadline. Under 110 words. Confident tone, not eager."

‍

Prompt 17 — Proposal Follow-Up That Isn't a Ping

‍

"Write a follow-up to send 3 days after a proposal. Don't ask 'any thoughts?' Instead, address the objection they're most likely sitting on silently: [insert objection], and suggest one specific next step, not an open-ended 'let me know.' Under 90 words."

‍

Prompt 18 — Breakup Email With an Actual Door Left Open

‍

"Write a final message to a prospect gone quiet after [X touches]. No guilt, no drama, no 'I understand you're busy.' Acknowledge timing might just be off, and give them one specific, easy way to come back later, like a resource or a check-in date, rather than a vague 'reach out anytime.' Under 55 words."

‍

‍

AI Prompts for Objection Handling

‍

Objection handling prompts fail when they produce defensive-sounding scripts. The fix is forcing multiple response angles so you're not stuck with one brittle line.

‍

Prompt 19 — Objection Map With Three Different Attack Angles

‍

"List the 5 objections a [job title] at a [company size/industry] company is most likely to raise about [your solution]. For each, give me three responses using different logic: one that reframes the objection, one that leans on evidence, and one that turns it into a question back to them. Under 30 words per response. No 'great question' filler."

‍

Prompt 20 — Price Objection Without Discounting

‍

"Respond to: 'Your pricing is too high.' Context: our solution typically delivers [X hours/dollars saved] with an ROI of [X]. Don't discount, don't get defensive, redirect to outcome instead of cost. Under 55 words."

‍

Prompt 21 — 'Not Right Now' Without Losing the Deal

‍

"Respond to: 'We're not ready to look at this right now.' Acknowledge their timing is real, don't push, plant something small for later, like a resource or a specific future check-in date. Under 55 words."

‍

Prompt 22 — Competitor-in-Place Objection Without Trashing Them

‍

"Respond to: 'We're already using [competitor].' Don't criticize the competitor. Acknowledge their current setup as legitimate, then introduce one specific difference relevant to [likely pain], and close with a question that opens a gap they'd have to answer honestly. Under 75 words."

‍

Prompt 23 — Verbal Objection Script That Doesn't Sound Scripted

‍

"Write a spoken response, not a written one, for handling this objection on a live call: [paste objection]. Structure: acknowledge, reframe, one piece of evidence, one question back. No filler words like 'absolutely' or 'great question.' Write it the way a person actually talks, with natural pauses, not clean sentences."

‍

AI Prompts for Sales Proposals

‍

Proposal prompts go wrong the second they start listing features. Every prompt here forces buyer-first framing.

‍

Prompt 24 — Executive Summary Written From the Buyer's Chair

‍

"Write an executive summary for a proposal to [job title] at [company]. Their stated pain: [pain point]. Our solution: [one sentence]. Expected outcome: [ROI/result]. Write this entirely from what they gain, not what we provide. If a sentence describes our product instead of their outcome, rewrite it. Under 140 words."

‍

Prompt 25 — ROI Section That Shows Its Work

‍

"Build an ROI section for a proposal. Context: [current cost of the problem], [our solution's monthly cost], [expected outcome: hours or dollars saved]. Present it as a simple before/after with a 12-month total, not a vague 'significant savings' claim. Plain language, no jargon."

‍

Prompt 26 — Proof Section That Doesn't Oversell

‍

"Write a results section for a proposal targeting [industry/persona], using this real result as the reference: [paste case study]. 3-4 sentence narrative, then 2-3 bulleted outcomes using the exact numbers given. Do not round up, exaggerate, or imply results that aren't in the source data."

‍

Prompt 27 — Proposal Cover Note That Tells Them Where to Look First

‍

"Write a short email to send with a proposal. Recipient: [job title] at [company]. Reference: [key pain from the discovery call]. Point them to the one section most relevant to what they care about, and suggest a specific time to walk through it together. Under 90 words."

‍

AI Prompts for Meeting Prep and Discovery

‍

Prompt 28 — Call Brief Built on Hypotheses, Not Just Facts

‍

"Act as a sales coach prepping me for a call. Based on this profile: [LinkedIn bio, company overview], give me 3 hypothesis-driven questions (not generic discovery questions), 2 objections I should expect and how to handle each, one case study angle that fits, and the pain point most worth probing first. One-page brief."

‍

Prompt 29 — Discovery Questions That Actually Uncover Cost of Inaction

‍

"Generate 8 open-ended discovery questions for [job title] at [company type]. Goal: uncover their current approach to [problem area], quantify what inaction is costing them, and surface their real decision criteria. No yes/no questions. Sequence from broad to specific, and make the last 2 questions uncomfortable enough that a vague answer would be a red flag."

‍

Prompt 30 — Post-Call Summary That Sets Up the Next Step

‍

"Write a post-call summary based on these notes: [paste raw notes]. Structure: what we covered, pain points surfaced, agreed next steps, proposed timeline. Under 140 words, action-oriented, no recap fluff. Include a placeholder for a calendar link."

‍

Prompt 31 — Demo Agenda Built Around Their Priorities, Not Our Feature List

‍

"Build a 30-minute demo agenda for [job title] at [company]. Their stated priorities: [2-3 priorities]. Structure the whole demo around their pain, not a feature tour. Intro (2 min), pain confirmation (3 min), solution walkthrough tied directly to their priorities (15 min), ROI framing (5 min), Q&A and next steps (5 min)."

‍

‍

AI Prompts for Pipeline Management and CRM Updates

‍

Prompt 32 — CRM Note That a Manager Can Actually Use

‍

"Write a CRM deal update from these notes: [paste raw notes]. Include: deal stage, stakeholders engaged, confirmed pain, next steps, risks or blockers, estimated close date. Under 110 words, categorized, not paragraph form."

‍

Prompt 33 — Stalled Deal Plan With Genuinely Different Angles

‍

"Act as a senior sales strategist. This deal has stalled: [stage, last activity, stated concern]. Give me 3 re-engagement approaches that are actually different: a new stakeholder angle, a competitive urgency angle, and a 'something changed' angle. For each, give a 2-sentence outreach message, not just a strategy label."

‍

Prompt 34 — Pipeline Review Talking Points That Flag Risk Honestly

‍

"Based on this pipeline summary: [deals with stage, value, last activity], give me 5 talking points for a weekly review. Flag deals at real risk of slipping, opportunities worth accelerating, and any deal that needs a strategy change, not just more effort. Concise bullets a manager can present in 5 minutes."

‍

‍

How to Write Better AI Prompts for Sales (The Framework)

‍

Every prompt above follows the same five-part structure, whether it's obvious or not:

‍

• Role — Who is the AI acting as? A researcher, a strategist, a copywriter. This shapes tone before you've written a word of instruction.
  
  
• Context — The specific buyer, company, and pain. This is where most people get lazy, and it's the single biggest driver of output quality.
  
  
• Instruction — What exactly should it produce? Not "help me with this email," but "write a cold email."
  
  
• Constraints — Word count, banned phrases, tone. This is what separates a usable draft from generic AI voice. Most people skip this entirely.
  
  
• Output format — Email, bullet list, script, table. Tell it the shape before it starts writing.

‍

The most common mistake is a prompt with zero constraints. Ask for "a cold email" with nothing else, and you get exclamation points, "I hope this finds you well," and 200 words that say nothing.

‍

The second most common mistake is skipping context. The more specific you are about the buyer, the company, and the exact pain, the less generic the output.

‍

The habit that actually compounds: keep a personal prompt library. Save the ones that produce output you actually send without heavy edits, and keep tightening them. Six months in, you'll have a set of prompts tuned to your exact ICP and voice that a generic list never could match.

‍

One non-negotiable rule sits above all of this: every AI draft gets a human review before it reaches a prospect. Edit for your voice, add the one detail only you'd know, and cut anything that still smells like a template.

‍

How Cleverly Uses Outbound to Book Meetings Without Prompt Guesswork

‍

‍

AI prompts speed up the drafting part of outbound. They don't build the infrastructure behind it, the deliverability setup, the list quality, the sequencing logic, or the daily optimization that separates a campaign that books meetings from one that gets ignored.

‍

That's a different job, and it's the one most B2B teams don't have spare capacity for.

‍

We built Cleverly to be the team that does that job full-time. We run LinkedIn outreach, cold email, and cold calling for over 10,000 B2B companies, and we've generated $312 million in client pipeline along the way.

‍

‍

Instead of a rep prompting and editing between calls, our specialists write the sequences, manage the technical side of deliverability, and hand qualified meetings straight to your calendar.

‍

For teams that want the smarter version of AI-assisted outreach, using it to sharpen messaging rather than replace the system entirely, Cleverly is built to be that partner. LinkedIn lead generation starts at $397/month, and our cold email and cold calling programs scale from there depending on volume and channel mix.

‍

🤝 See how Cleverly runs outbound for B2B companies and what a done-for-you system looks like next to a prompt library.

‍

Conclusion

‍

AI can draft a cold email, prep your next call, or outline a proposal in under a minute. What it can't do is know your buyer better than you do, which is exactly why the prompt matters more than the tool. Role, context, instruction, constraints, output format.

‍

Get those five right and the draft is usable. Skip them and you're sending something that sounds like every other AI-written email landing in that inbox this week.

‍

Save the prompts from this list that actually produce something you'd send without a rewrite. Adjust the brackets to your ICP, tighten the constraints as you learn what your buyers respond to, and keep the human review step non-negotiable. AI drafts. You close.

‍

‍

Frequently Asked Questions

‍