ChatGPT for sales: Use cases, prompts, & how to close deals faster

Sales is a high-volume, high-repetition job. Prospecting emails, follow-ups, proposal drafts, objection responses, discovery prep — the average sales rep spends a significant portion of their week on work that follows predictable patterns and doesn't require deep creative thinking to execute.

ChatGPT is well-suited to that kind of work. It can draft outreach, research prospects, write proposals, prepare for calls, and handle the document work that slows down deal cycles — all faster than doing it manually. This guide covers how to use ChatGPT for sales at every stage of the sales cycle, with practical prompts you can use today

How sales teams are using ChatGPT right now

Before diving into how teams are using ChatGPT for sales, it's worth being honest about what it is and isn't good for.

It’s great at writing, structuring, and iterating. Outreach emails, follow-up sequences, proposal copy, objection responses, discovery questions — anything that follows a recognisable pattern and requires clear, professional writing is a good candidate.

That said, it’s not a replacement for knowing your prospect, your product, or your market. The best outputs come from reps who provide specific context, review the results, and treat ChatGPT as a capable assistant rather than an autopilot. The prompts below are designed with that in mind.

9 of the best uses for ChatGPT for sales — with prompts for each stage

1. Prospecting research

Understanding a prospect's business before you reach out makes every subsequent step more effective. ChatGPT can help you build a research framework, identify likely pain points, and prepare talking points based on a company's public profile.

Try these prompts: 

• I'm about to prospect into [Company Name], a mid-market SaaS company with about 200 employees. They sell [product] to [audience]. Based on that profile, what are the most likely operational pain points their sales team faces? What business pressures might make them receptive to a conversation about [your product category] right now?

• I'm targeting companies in the logistics industry with 50-250 employees. What are the most common [product area] pain points in that sector, and what language would resonate with a [persona/role] at that type of company?

What to do with the output: Use it as a starting point for prioritising your angle, not as verified fact. ChatGPT's profile of an industry is based on general patterns, not your specific prospect. Once you've done your own research — LinkedIn, the company website, CRM notes, recent news — paste the most relevant details back into the same conversation and ask ChatGPT to refine the angle based on what you've found.

ChatGPT builds context within a session, so the more you feed it, the more tailored the output gets.

2. Personalized outreach and cold email

Cold email is one of the highest-volume, most repetitive writing tasks in sales — and one of the most obvious candidates for AI assistance. ChatGPT can draft personalized outreach quickly if you give it specific context about the prospect and your angle.

Try these prompts: 

• Write a cold email to [Name], Head of Sales at [Company]. They recently expanded into the European market (I saw this on LinkedIn). I want to open a conversation about how [company] helps sales teams close deals faster by [product value]. Keep it under 100 words, lead with the expansion angle, and end with a low-friction CTA asking if they have 15 minutes this week.

• Write a three-email cold outreach sequence for a sales rep targeting HR managers at companies with 50-200 employees. The product is a [your tool]. Email 1 should open with a pain point. Email 2 should provide a specific proof point. Email 3 should be a short break-up email. Each email should be under 80 words.

What to do with the output: Read every email out loud before sending. ChatGPT's outreach tends to be well-structured but may be stiff. Edit for your voice, confirm that any specific references are accurate, and remove any language that sounds generic or like it came from a template — even if it did.

3. Summarizing call notes and extracting next steps

After a discovery call or demo, turning messy notes into a clean summary, a set of action items, and a follow-up email used to take at least 20-30 minutes. It now takes a simple prompt.

Try these prompts: 

• Here are my notes from a discovery call with [Name] at [Company]: [paste notes]. Please summarize: (1) the key pain points they mentioned, (2) the questions they asked, (3) the objections raised, (4) agreed on next steps, and (5) any red flags I should be aware of.

• Based on these call notes, draft a follow-up email to [Name] summarising what we discussed, confirming next steps, and attaching the relevant case study I mentioned. Tone should be warm and professional. 

• I'm going to paste a transcript from a sales call recorded in Gong. Please identify: (1) the prospect's top three stated pain points, (2) any objections raised and how they were handled, (3) buying signals, (4) questions that went unanswered, and (5) recommended talking points for the follow-up.

What to do with the output: Verify the summary against your own recollection of the call before sending it anywhere to ensure it didn’t miss any nuance that was clear in the conversation.

The Gong transcript prompt is particularly useful before a follow-up call with multiple stakeholders, as it helps you walk in with a clear picture of where everyone stood.

4. Follow-up emails

Following up consistently is one of the most reliable ways to close more deals, and one of the tasks sales reps most often skip when they are busy. ChatGPT removes the blank-page problem for follow-up writing.

Try these prompts: 

• Write a follow-up email to a prospect who attended a demo three days ago but hasn't responded. The demo went well — they showed interest in the [features] but raised a concern about [feature]. Acknowledge the question, provide a one-sentence reassurance, and ask to schedule a follow-up call.

• I sent a proposal to [Name] at [Company] five days ago. They opened it twice but haven't responded. Write a short follow-up email that creates gentle urgency without being pushy. Reference that I can see they've reviewed the proposal and offer to answer any questions.

What to do with the output: Personalize the opening line with something specific to your last interaction. The more specific the context you provide in the prompt, the less editing you need to do on the output. Generic follow-ups get ignored, specific ones get replies.

5. Discovery call preparation

The quality of your discovery determines the quality of everything that follows. ChatGPT can help you build a question framework tailored to a specific prospect, industry, or deal type before you get on the call.

Try these prompts:

• I have a discovery call tomorrow with the [role] at a [X-person][industry] firm. I want to understand their current workflow, where the friction points are, and what success looks like for them. Give me 10 discovery questions organized by topic, starting with current state and moving toward impact and urgency.

• Give me the 5 most important questions to ask during a discovery call with a sales leader who is evaluating a [your product segment] for the first time. I want to understand their current proposal process, volume, pain points, and decision criteria.

What to do with the output: Use the questions as a framework, not a script. The best discovery is a conversation that follows the prospect's lead. Paste the questions somewhere you can reference them, but prioritise listening over following the list in order.

6. Objection handling

Preparing for objections before they come up is better than trying to handle them in the moment. ChatGPT can help you build a thorough objection library and practice responses.

Try these prompts:

• I sell a [product category] to [ICP]. What are the 5 most common objections a prospect would raise during the sales process, and what is a concise, effective response to each one?

• A prospect just told me: "We already use [competitor] and we're happy with it." Give me three different ways to respond to that objection, ranging from a soft probe to a more direct competitive challenge. I want options depending on how engaged the prospect seems.

• A mid-market prospect is stalling on a decision and said they need to discuss internally before moving forward. Write a response that acknowledges the need for internal discussion, asks a question to understand who else is involved, and moves toward scheduling a next step.

What to do with the output: Practise the responses out loud. Objection handling that reads well on a screen often sounds unnatural when spoken. Edit for your voice and delivery style, then use the written version as a reference for coaching or onboarding new reps.

7. Proposal and agreement drafting

Proposals are where the document side of sales really starts to matter. A well-written proposal that arrives quickly gives buyers confidence. A slow, generic one loses deals even when the product is right.

ChatGPT can draft proposal copy, executive summaries, and custom sections — but where it becomes most powerful for sales teams is when you combine it with a tool that can take that draft and turn it into a formatted, signature-ready agreement without leaving the conversation.

Try these prompts:

• Write an executive summary for a proposal we're sending to [Company]. They are a 200-person consulting firm looking to reduce the time their team spends on contract creation and approval. Their main pain points are [pain point] and [pain point]. Our solution addresses both through [feature]. Keep it under 150 words.

• Write a "Why us" section for a proposal targeting a [role] at a [industry] company. Focus on [benefit 1], [benefit 2], and [benefit 3] for the team. Avoid generic language.

• Draft a 3-email nurture sequence to send after a proposal has been delivered. Email 1 confirms delivery and offers to answer questions. Email 2 (sent 3 days later) provides a relevant proof point using this [case study details]. Email 3 (sent 5 days later) creates soft urgency around a timeline.

What to do with the output: Review the proposal copy against your actual product capabilities and the specific deal context before sending. Generic proposal language undermines trust. The closer the prompt context is to the real deal, the less editing the output needs.

ChatGPT gets more powerful when it's connected to the tools your team already uses. On its own, ChatGPT can draft a strong proposal — but it's working from whatever context you paste in. When you connect it to your CRM, your document platform, and your product catalog, the whole workflow changes.

A sales rep can pull deal details from HubSpot, describe the agreement they need, have PandaDoc turn it into a signature-ready document with the right template and variables filled, and send it for signature — all from one ChatGPT conversation. The draft doesn't just get written faster; it gets written with the right data already in it.

8. Drafting and sending agreements with ChatGPT

Once a deal is verbally agreed, the time between handshake and signed contract is where momentum dies. Every extra step — exporting from ChatGPT, formatting in another tool, setting up signature fields, sending — creates an opportunity for the deal to stall.

The PandaDoc ChatGPT App removes all of those steps. You can describe the agreement you need, have ChatGPT draft it, and send it for signature without leaving the conversation. The draft becomes a formatted, signature-ready document in PandaDoc automatically.

Example workflow:

You: Create a 12-month services agreement for Acme Corp — $5,000 per month, net-30 payment terms, 60-day termination notice, governing law New York. Add john.smith@acmecorp.com as the signer.

ChatGPT + PandaDoc: Here's your services agreement. I've created it in PandaDoc with signature fields placed. John Smith has been added as the signer. Ready to send?

For sales reps creating proposals and agreements at volume, cutting the steps between "drafted" and "sent for signature" meaningfully compresses deal cycle time.

What to do with the output: Review the agreement in the PandaDoc widget before sending — verify the key commercial terms, confirm the parties are named correctly, and check the signature fields are placed correctly. Once confirmed, send directly from the conversation.

9. Sales coaching and training content

Sales managers can use ChatGPT to create training materials, role-play scenarios, and onboarding content faster than building everything from scratch.

Try these prompts:

• Create a role-play scenario for training a new sales rep on handling the "we don't have budget right now" objection. Include the rep's opening line, three variations of the objection, and a model response for each. The product is a [product] sold to [ICP].

• Write a 30-60-90 day onboarding plan for a new SDR joining a B2B SaaS sales team. Include weekly milestones, key skills to develop each month, and suggested activities for each phase.

• Create a call review framework for sales managers. List 10 questions a manager should ask when reviewing a recorded discovery call to assess whether the rep uncovered genuine pain, established urgency, and qualified the opportunity effectively.

What to do with the output: Review training content with experienced team members before distributing it. ChatGPT's sales frameworks are based on general best practices — your team's specific methodology, product, and buyer profile may require adjustments.

Tips to more effectively use ChatGPT for sales 

The example ChatGPT prompts for sales listed above share a few things in common: they're specific, they include context about the prospect, and they define the output format. Here's how to get even more out of every prompt.

Give it your voice before you ask for anything

Paste a sample of your best-performing outreach email or a proposal section you're proud of and tell ChatGPT to match that tone for everything you ask it to write. This reduces the editing time significantly and produces output that sounds like you rather than a generic sales template.

Be specific about the prospect

The more context you provide — company size, industry, role, pain point, recent trigger event — the better the output. "Write a cold email to a VP of Sales" produces something generic. "Write a cold email to a VP of Sales at a 200 person fintech company that just announced a Series B" produces something you can actually use.

Use it to iterate, not just to generate

If the first output isn't right, tell ChatGPT exactly what to change: "Make the opening more direct," "Remove the third paragraph," "Rewrite this to sound less formal." Iteration is often faster than writing a better prompt from scratch.

Build a prompt library for your team

The prompts that work well for your product, your buyer, and your sales motion are worth saving and sharing. A shared prompt library means the whole team benefits from the best-performing approaches rather than every rep starting from scratch.

Use it for structure, not final language

ChatGPT is excellent at producing a complete document structure quickly. The specific language, particularly in high-stakes provisions, should be reviewed carefully and often rewritten based on your firm's standards.

Don't send AI output without reading it

This is especially true for personalised outreach and proposals. AI-generated content may include small errors, awkward phrasing, or generic language that undermines the credibility you're trying to build. Read everything before it leaves your inbox.

What ChatGPT can't do for sales teams

Understanding what ChatGPT can and can't do is essential to knowing how to use ChatGPT in sales effectively. 

• It can't replace genuine relationship building. Buyers buy from people they trust. ChatGPT can help you write better emails and prepare better for calls, but the rapport, credibility, and trust that close deals come from real human interaction.

• It doesn't know your specific buyer. ChatGPT can describe the general profile of a VP of Sales at a mid-market tech company. It doesn't know the specific person you're talking to — their priorities, their history with your company, or their personal communication style. That knowledge comes from your research and your conversations. Make sure you add that detail to your prompts.

• It doesn't have visibility into your pipeline or deal activity — unless you connect it to a tool that does. On its own, ChatGPT has no access to your CRM, your open proposals, or where each deal stands. But connected to PandaDoc, it can tell you which proposals have been opened, who hasn't signed, what changed across your active agreements overnight, and what deals need attention — all through natural language, without logging into another platform. The limitation is standalone ChatGPT, not ChatGPT with the right integrations.

• It can produce confident-sounding content that isn't accurate. Review any content that makes specific claims about your product, your pricing, or your customers before it goes to a prospect. AI-generated errors in a proposal are hard to recover from.

Ready to cut the time between drafted and signed? With the PandaDoc ChatGPT App, your next proposal is already in your ChatGPT window.