AI Prompts for Sales: 18 That Work on Real Deals
Direct answer
AI prompts for sales that work on real deals, not demos. 18 prompts for discovery, objection handling, follow-up, and closing that operators use weekly.
- AI prompts for sales that work on real deals, not demos. 18 prompts for discovery, objection handling, follow-up, and closing that operators use weekly.
- The strongest AI work starts with one operational bottleneck, one owner, and one result the team can inspect.
- Use the article as the diagnosis layer, then move into a scoped build, proof path, or commercial workflow page.
What makes AI prompts for sales actually work?
An AI prompt for sales works when it produces output a rep can send or use with a short edit, on a real deal, not a demo. Most prompt lists fail this test because they were written to sound impressive rather than to produce usable output. The 18 prompts below were iterated against actual sales conversations, and each one has been through at least 20 cycles of real-world use before appearing here. The pattern that separates a usable prompt from a wasted one is specific context in, specific output out: name the prospect, paste the exact objection, describe the signal you saw, and the model has something real to work with.
A good sales prompt does four things. It provides the specific context the model needs to produce a non-generic output. It specifies the format and length of what you want back. It states the one constraint that matters most for this use case, usually tone, specificity, or word count. And it names the outcome the output is trying to reach. "Write me a cold email" produces a cold email written to a hypothetical prospect. "Write a 4-sentence cold email for a head of operations at a 50-person logistics company who recently posted about trouble finding reliable last-mile delivery partners" produces something you can actually send. Treat the prompt as a brief, not a wish.
Outreach prompts
Prompt 1: Cold email from a research signal
"Write a 4-sentence cold email to [first name], [job title] at [company]. They recently [specific signal: posted about X / were mentioned in article about Y / are hiring for role Z]. My offer is [brief offer description]. The email should reference the signal in the first sentence, make one specific value offer in the second sentence, ask one question in the third sentence, and end with a low-friction CTA. Tone: direct, no corporate language, no superlatives."
Prompt 2: LinkedIn first message
"Write a 2-sentence LinkedIn connection request message to [first name] at [company]. I am reaching out because [specific reason: they commented on a post about X / they are hiring for Y / I noticed they are working on Z]. The message should feel like a genuine observation, not a sales pitch. Do not use the phrase 'I came across your profile.'"
Prompt 3: Follow-up after no reply, different angle
"I sent [first name] an email [X days] ago about [topic] and did not get a reply. Write a 3-sentence follow-up email that takes a completely different angle from the original. Instead of leading with my offer, lead with a specific question about the problem I help solve. Make the question specific enough to require a real answer, not a yes or no."
Prompt 4: Close-the-loop message
"Write a 2-sentence close-the-loop email to [first name] who has been in my outreach sequence for 21 days without replying. The message should indicate that I will not follow up again after this, wish them well, and leave the door open for future contact. Tone: graceful, no guilt, no passive aggression."
Discovery call prep prompts
Prompt 5: Pre-call briefing
"I have a discovery call in 30 minutes with [name], [job title] at [company]. Here is what I know about them: [company website copy / LinkedIn summary / any prior emails / relevant news]. Write a one-page briefing with: company context in 3 sentences, three likely pain points based on their industry and role, two potential objections I should prepare for, and five opening questions specific to this prospect's situation."
Prompt 6: Competitive prep for a specific deal
"I am in a deal with [company] who is also evaluating [competitor]. Write 3 talk tracks that address the most common reasons buyers choose [competitor] over us, assuming our differentiators are [list 2 to 3 differentiators]. Each talk track should be 2 to 3 sentences. Tone: confident, no disparaging the competitor directly."
Prompt 7: Stakeholder mapping
"In a discovery call, I learned the following about the buying process at [company]: [notes from call]. Based on this, write a stakeholder map summary identifying who appears to be the economic buyer, who is the technical evaluator, who is the champion, and who might be a blocker. Flag any gaps in my information."
Post-call prompts
Prompt 8: Call summary from a transcript
"Here is the transcript of a 45-minute discovery call: [transcript]. Write a structured summary covering the prospect's main problems in their own words, what they said about timing and budget, what they asked about that I should follow up on, the agreed next steps, and any concerns or objections that came up."
Prompt 9: Follow-up email from call notes
"I just finished a call with [name] at [company]. Here are my notes: [notes]. Write a follow-up email that recaps the top 3 things we discussed, confirms the agreed next steps, includes one specific resource or case study relevant to their situation (describe the type if you do not have a specific one), and ends with a clear ask for the next meeting."
Prompt 10: CRM deal update
"Based on this call summary: [summary], write a CRM deal note that covers current deal stage, what happened on this call, what needs to happen next, any blockers or risks identified, and a recommended probability adjustment with reasoning. Keep it under 150 words."
Objection handling prompts
Prompt 11: Objection response library entry
"Here is an objection I heard on a call: [exact objection in prospect's words]. Write 3 different responses: one that reframes the objection, one that acknowledges it and asks a clarifying question, and one that addresses it directly with a specific example. Each response should be 2 to 3 sentences."
Prompt 12: Price objection response
"A prospect said: '[exact price objection]'. Write a 3-sentence response that acknowledges the concern without apologising for the price, reframes the value in terms of the specific outcome they mentioned wanting, and asks a question that either advances the conversation or surfaces the real objection behind the price concern."
Proposal and close prompts
Prompt 13: Proposal executive summary
"Write a one-page executive summary for a proposal to [company] for [service/product]. Their main problem is [problem]. The outcome we are delivering is [outcome with specific metric if possible]. The investment is [price range]. The summary should open with their problem, describe the outcome in their language, explain briefly how we deliver it, and end with the investment and a single CTA."
Prompt 14: Deal risk assessment
"Here is the current status of a deal: [deal notes including stage, last contact date, key stakeholders, any objections raised, deal size, timeline discussed]. Write a risk assessment that identifies the top 3 risks to this deal closing as forecasted, what evidence supports each risk, and what specific action would reduce each risk."
Enablement prompts
Prompt 15: Battle card from competitor research
"Here is recent information about [competitor]: [product page / recent press / review site quotes]. Write a battle card section covering what they do well, where they fall short, the 3 most common objections we hear when we are being compared to them, and the talk track we should use for each objection."
Prompt 16: New rep onboarding summary from top calls
"Here is a transcript from one of our top-performing sales calls: [transcript]. Write a 1-page summary for a new rep that captures the key questions the rep asked that drove the discovery forward, how the rep handled the [specific objection that appeared], the structure of the call overall, and the 3 things a new rep should take from this call."
Forecasting prompts
Prompt 17: Deal health check
"Here are the details of a deal that has been in the pipeline for [X weeks]: [deal history including contact frequency, stage progression, stakeholders engaged, size, timeline]. Compare this pattern to what you know about deals that stall versus deals that close. Write a 3-sentence assessment of the deal's health and one specific recommended action to advance it."
Prompt 18: Pipeline review prep
"Here is my pipeline for the [quarter]: [list of deals with stage, value, close date, last contact]. Write a pipeline review summary that flags the three deals most at risk of slipping, identifies the two deals closest to close that need attention this week, and estimates a range for the quarter's close total based on historical close rates for deals at each stage."
How to fit these prompts into a real sales week
A prompt only earns its place if it saves time at a moment you are already busy. The strongest fit is the post-call window, when notes are fresh and the next action is unclear: Prompts 8, 9, and 10 turn raw notes into a summary, a follow-up, and a CRM note in a few minutes. Outreach prompts pay off in batches, when you have a list of signals and want consistent, specific first messages rather than 20 versions of the same template. Discovery prep prompts work best the half hour before a call, not the night before, because the context you paste in is what makes the output specific. The mistake most reps make is trying to use all 18 at once. Pick the two that match where your deals leak, build the habit, then add more. For the wider toolset that sits around these prompts, see the best AI tools for sales and the step-by-step setup in how to use AI for sales.
How twohundred approaches this
When we build a sales workflow for an operator, we do not start with a prompt library. We start with where the deal data lives and where it goes stale. A prompt that drafts a CRM note is worth little if the rep still has to paste it in by hand, so the highest-return work is usually wiring the model into the CRM directly: the call transcript comes in, the summary and the deal-stage update go out, and the rep approves rather than retypes. That is what AI CRM integration does, and it is the layer most prompt guides skip. The prompts on this page are the starting brief. The system that runs them on every call without a human copy-pasting is the part that compounds. If you want that built around your stack rather than bolted on, that is the work we do at AI CRM integration, and you can talk to us about your pipeline first.
Frequently asked questions
How do I get AI prompts for sales to produce non-generic output?
The single most important variable is the specificity of the context you provide. Generic context produces generic output. Specific context, including the exact words a prospect used, the precise signal you spotted, or the objection quoted verbatim, produces output that is usable on a real deal. Every prompt above asks for that context as input before it asks for the output, which is why they hold up outside a demo.
Should I edit the AI output before sending?
Yes, always. The prompts above produce strong first drafts, not final sends. The editing step should take 2 to 4 minutes per output and focus on adjusting the tone to match your voice, adding any context the model did not have, and removing any phrasing that sounds generated rather than written. A 3-minute edit that makes the output sound like you wrote it is worth more than sending the draft unchanged.
Which AI model should I use for these sales prompts?
Any capable general model handles these prompts well, because the work is reasoning over context you provide, not recalling facts. The difference between models shows up in tone control and how closely they follow length and format constraints, so test the same prompt across two models with your real data and keep the one that needs the least editing. What matters far more than model choice is the quality of the context you paste in.
Do I need a CRM to use AI prompts for sales?
No. Every prompt here works from a blank chat window with notes pasted in, so you can start today with nothing more than the tool you already use. A CRM matters when you want the output to flow back into your pipeline automatically instead of being copied by hand, which is where these prompts stop being a personal time-saver and start being a system the whole team runs on.
---
Related Services
Connecting AI to your CRM and sales stack is what AI integration services covers, from API connections to workflow triggers. For structuring the broader deployment, AI implementation services walks through the rollout process.
Related implementation paths
AI implementation services
Turn the article into a scoped first system with clear ownership, data, and measurement.
AI workflow automation
Automate one operational workflow inside the tools the team already uses.
AI agent development company
Design agents around jobs, tools, approval points, and measurable business outcomes.
Questions this article answers
What makes AI prompts for sales actually work?
An AI prompt for sales works when it produces output a rep can send or use with a short edit, on a real deal, not a demo. Most prompt lists fail this test because they were written to sound impressive rather than to produce usable output. The 18 prompts below were iterated against actual sales conversations, and each one has been through at least 20 cycles of real world use before appearing here. The pattern that separates a usable prompt from a wasted one is specific context in, specific output out : name the prospect, paste the exact objection, describe the signal you saw, and the model has something real to work with. A good sales prompt does four things. It provides the specific context the model needs to produce a non generic output. It specifies the format and length of what you want back. It states the one constraint that matters most for this use case, usually tone, specificity, or word count. And it names the outcome the output is trying to reach. "Write me a cold email" produces a cold email written to a hypothetical prospect. "Write a 4 sentence cold email for a head of operations at a 50 person logistics company who recently posted about trouble finding reliable last mile delivery partners" produces something you can actually send. Treat the prompt as a brief, not a wish.
How do I get AI prompts for sales to produce non generic output?
The single most important variable is the specificity of the context you provide. Generic context produces generic output. Specific context, including the exact words a prospect used, the precise signal you spotted, or the objection quoted verbatim, produces output that is usable on a real deal. Every prompt above asks for that context as input before it asks for the output, which is why they hold up outside a demo.
Should I edit the AI output before sending?
Yes, always. The prompts above produce strong first drafts, not final sends. The editing step should take 2 to 4 minutes per output and focus on adjusting the tone to match your voice, adding any context the model did not have, and removing any phrasing that sounds generated rather than written. A 3 minute edit that makes the output sound like you wrote it is worth more than sending the draft unchanged.
Which AI model should I use for these sales prompts?
Any capable general model handles these prompts well, because the work is reasoning over context you provide, not recalling facts. The difference between models shows up in tone control and how closely they follow length and format constraints, so test the same prompt across two models with your real data and keep the one that needs the least editing. What matters far more than model choice is the quality of the context you paste in.
Do I need a CRM to use AI prompts for sales?
No. Every prompt here works from a blank chat window with notes pasted in, so you can start today with nothing more than the tool you already use. A CRM matters when you want the output to flow back into your pipeline automatically instead of being copied by hand, which is where these prompts stop being a personal time saver and start being a system the whole team runs on.
Imraan, Founder of twohundred
Imraan is the founder of twohundred, a US AI implementation lab. Before this he built six businesses, hired more than 200 people, and sold one to a public company. He started his career at UBS in London.
Working through one of these decisions?
Book a 30-minute call. We will look at the specific workflow you are trying to put AI into, and what it would actually take to make it work in production.
Book a call