I spent two years running outbound sales before I started using AI. The bottleneck was never writing faster — it was writing better. More specific. More relevant. Less obviously templated.
These 10 prompts are the ones I use every week. Each one has a Role → Context → Constraints → Output structure, and each one produces something I'd actually send to a prospect. I've tested every one of them on real outreach (yes, on real people — I checked the reply rates).
Full disclosure: I sell a SaaS Marketing Copy Prompt Pack ($49) and a Developer Product-Launch Prompt Pack ($9) — both include sales/outreach prompts. These 10 are a sample. If they help, the packs have 200+ more.
How These Prompts Are Structured
Every prompt below follows the same pattern:
ROLE: Who the AI is acting as
CONTEXT: The situation it needs to know about
CONSTRAINTS: What it can and can't do
OUTPUT: The exact format you want back
This structure matters. Without constraints, the AI writes generic corporate mush. With them, it writes something a human would actually read.
1. The Pain-First Cold Email (3 sentences, no fluff)
ROLE: You are a B2B SDR who has booked 500+ meetings in 2 years.
CONTEXT: I'm reaching out to {role} at {company}. They likely deal with
{specific pain point — e.g., "spending 4+ hours/week manually pulling
reports from 3 different tools"}. Our product {one-line value prop}.
CONSTRAINTS:
- Subject line: ≤5 words, lowercase, no question marks
- Body: exactly 3 sentences
- Sentence 1: the pain (observed, not assumed)
- Sentence 2: the specific outcome we produce
- Sentence 3: a soft ask (not a meeting request)
- BANNED words: "revolutionary", "seamless", "empower", "leverage",
"game-changing", "innovative", "cutting-edge"
- Do NOT mention our company name in the body
- Do NOT use exclamation points
OUTPUT: Subject line + email body. Nothing else.
Why it works: The constraint "observed, not assumed" forces you to research before the AI writes. The ≤5-word lowercase subject bypasses spam filters and pattern-matching. The soft ask (e.g., "Worth a quick look?" or "Open to seeing how?") gets 2-3x reply rates vs. "Can we hop on a call?"
2. The Discovery Call Prep Sheet (questions that reveal budget)
ROLE: You are a senior AE preparing a discovery call.
CONTEXT: I'm meeting with {persona} at {company} ({industry}, ~{revenue}
revenue). They're evaluating {category}. Our competitors in this space
are {competitor 1, competitor 2}. The deal size is typically {amount}.
CONSTRAINTS:
- Generate exactly 7 questions
- Questions must map to BANT (Budget, Authority, Need, Timeline)
but DO NOT label them or ask "what's your budget" directly
- At least 2 questions must be about their current process/workflow
(not about our product)
- At least 1 question must surface a potential deal-killer early
- No yes/no questions — every question must require a 2+ sentence answer
- Include a "trap question" — one that reveals whether they're
comparison-shopping or problem-solving
OUTPUT: Numbered list of 7 questions. After each, a 1-line note on
what answer you're listening for.
Why it works: Most discovery calls fail because reps ask product questions instead of process questions. This prompt forces workflow-first questioning, which reveals more and builds trust faster.
3. The Objection Crusher (objection → reframe, not argument)
ROLE: You are a sales trainer who specializes in objection handling
without being pushy.
CONTEXT: A prospect said: "{exact objection quote}".
They are a {role} at {company type}. Our product costs {price}.
CONSTRAINTS:
- Do NOT contradict them
- Do NOT say "I understand" or "That's a great point"
- Acknowledge the concern in ≤8 words
- Then reframe it using their own words
- Provide exactly 3 possible responses, each ≤3 sentences:
1. The logic reframe (data/perspective shift)
2. The story reframe (a relevant case study one-liner)
3. The question reframe (turn it back to them)
- Mark which one is best for a {personality type: analytical/expressive/driver}
OUTPUT: The 3 responses, clearly labeled. A 1-line recommendation
on which to use and why.
Why it works: "I understand" signals you're about to argue. Banning it forces a genuine acknowledgment. The three-response format lets you adapt in real-time based on the prospect's energy.
4. The Follow-Up That Doesn't Get Ignored (sequence of 4)
ROLE: You are a B2B SDR writing a 4-touch follow-up sequence.
CONTEXT: Prospect {name} opened my first email but didn't reply.
They're a {role} at {company}. Original email was about {topic}.
CONSTRAINTS:
- Touch 1 (Day 3): value-add, not a nudge. Share one relevant insight
or resource. ≤4 sentences.
- Touch 2 (Day 7): the pattern interrupt. A question that reframes
the conversation. ≤2 sentences.
- Touch 3 (Day 14): the breakup email. Professional, gives permission
to disengage, leaves the door open. ≤3 sentences.
- Touch 4 (Day 30): the check-in. Casual, no pitch. References
something timely. ≤2 sentences.
- BANNED: "just checking in", "circling back", "bumping this up",
"following up on my previous email"
- Each email must have a different tone (helpful → curious →
detached → casual)
OUTPUT: 4 emails with subject lines, labeled by day.
Why it works: The breakup email (Touch 3) consistently gets the highest reply rate in the sequence — psychology, not desperation. Banning "just checking in" eliminates the #1 reason prospects ghost.
5. The LinkedIn Connection Note (mutual value, not a pitch)
ROLE: You are a B2B sales professional who has a 40%+ acceptance rate
on cold LinkedIn connections.
CONTEXT: I found {name} who is a {role} at {company}. We have
{mutual connection or shared group/interest}. They recently
{observable activity — posted, commented, launched, hired}.
CONSTRAINTS:
- ≤200 characters (LinkedIn limit is 300, but shorter = better)
- Reference the observable activity specifically
- Do NOT pitch anything
- Do NOT ask for a call or meeting
- Offer one specific thing of value (an intro, a resource, an insight)
- End with a question mark (questions get higher reply rates)
- No emojis
OUTPUT: Just the connection note text.
Why it works: LinkedIn connection notes that pitch get ignored. Notes that reference specific recent activity + offer value get accepted. This prompt forces both.
6. The Competitor Battle Card (honest, not trash-talk)
ROLE: You are a sales enablement lead creating a battle card.
CONTEXT: We sell {our product}. Main competitor is {competitor}.
Our pricing: {our price}. Their pricing: {their price}.
CONSTRAINTS:
- Do NOT trash-talk the competitor
- Do NOT make claims you can't prove with a URL
- Create a table with 4 columns: Feature/Aspect | Us | Them |
What to say (if asked)
- Cover exactly 6 rows: pricing model, onboarding time, key
differentiator, integration ecosystem, support model,
ideal customer profile
- For "What to say": frame as a question back to the prospect,
not a statement
- Include a "DO NOT SAY" section with 3 things that will backfire
OUTPUT: The table + the DO NOT SAY section.
Why it works: Prospects who mention competitors are 60% of the way to buying — they're validating. Trash-talking the competitor signals insecurity. This prompt produces honest comparisons that build trust.
7. The Post-Demo Recap Email (recap that moves the deal forward)
ROLE: You are an AE writing a post-demo follow-up.
CONTEXT: Just finished a demo with {name} ({role}) at {company}.
Key things discussed: {3 bullet points of what they cared about most}.
Next step discussed: {agreed next step}. Timeline mentioned:
{their timeline if any}.
CONSTRAINTS:
- Subject line: "Recap: {topic}" — no clickbait
- First line: what THEY said mattered most (not what you pitched)
- Second section: the 3 key takeaways in their words
- Third section: exactly one clear next step with a date
- Fourth section: one relevant resource (case study, doc, etc.)
- Do NOT include pricing unless they asked about it on the call
- Do NOT say "great meeting you" or "thanks for your time"
- ≤200 words total
OUTPUT: Subject + email body.
Why it works: The recap email is the most-forwarded email in B2B sales — prospects share it internally to justify the purchase. Leading with their priorities (not your pitch) makes it forwardable.
8. The Pricing Objection Reframe (value anchor, not discount)
ROLE: You are a sales coach specializing in value-based selling.
CONTEXT: Prospect said "{pricing objection}" about our product
that costs {price}. The alternative they're considering is
{alternative + its price or "doing nothing"}.
CONSTRAINTS:
- Do NOT offer a discount
- Do NOT justify the price with features
- Calculate the cost of the status quo (what they're losing
by not solving the problem)
- Reframe using one of: cost-per-month, cost-per-seat-per-day,
ROI timeline, or opportunity cost
- Provide the math explicitly (show your work)
- End with a question that shifts focus from price to value
OUTPUT: The reframe (3-5 sentences) + the explicit math.
Why it works: Price objections are rarely about price — they're about unproven value. This prompt forces you to quantify the cost of doing nothing, which is almost always higher than your price.
4 more prompts (abbreviated):
9. The Lost-Deal Post-Mortem Email — a graceful "what could we have done better?" email to lost deals. Gets 30%+ reply rate and occasionally reopens deals. Constraint: no re-pitching, genuine curiosity, ≤4 sentences.
10. The Upsell Trigger Detector — paste a customer's recent support tickets / feature requests / usage data, and the prompt identifies upsell opportunities. Constraint: must map to a specific product tier or add-on, must include the exact trigger phrase to listen for.
11. The Sales Forecast Reality Check — paste your pipeline, and the prompt stress-tests each deal's probability. Constraint: must identify deals that are "happy ear" (you're too optimistic) vs. "at risk" (you're not paying attention). Outputs a revised forecast with confidence intervals.
12. The ICP Refinement Prompt — paste your last 20 won deals + last 20 lost deals, and the prompt identifies the pattern that separates them. Constraint: must produce a 1-page Ideal Customer Profile with firmographic + behavioral signals, not generic personas.
How I Test These Prompts
- Run each prompt 3 times with the same inputs. If the outputs vary wildly, the prompt is too loose — tighten the constraints.
- Generalize: after testing on one scenario, swap in a different industry/product. If it still works, the prompt is robust.
- Read it aloud. If it sounds like an AI wrote it, add a constraint that bans the telltale phrase.
- Cut anything generic. If a line could apply to any company in any industry, delete it and replace with something specific.
The Prompt Packs
If these were useful, I've packaged 200+ more into structured packs:
- Developer Product-Launch Prompt Pack ($9) — 7 prompts for shipping and launching dev products
- SaaS Marketing Copy Prompt Pack ($49) — 25 prompts for landing pages, ads, emails, and sales copy
- Developer Productivity Prompt Library ($49) — 30 prompts for dev workflows
- AI Startup Operations Prompt System ($199) — 50+ prompts + 10 workflows for running a startup with AI
- AI Business Transformation Playbook ($999) — 100+ prompts + 12 workflows, the flagship
Or browse the full store: innovate01.gumroad.com
What sales prompts have worked for you? I'm always looking to test new ones.
Top comments (0)