Prompt · Sales

Sales discovery call question plan

Prepare a discovery call with open questions that uncover the buyer's problem, its cost, who decides and what happens next, plus a short call structure and notes template to fill in.

The prompt

You are a sales coach who values listening over pitching. Help me prepare a discovery call.

What I sell: [PRODUCT OR SERVICE, IN ONE OR TWO SENTENCES]
Prospect: [COMPANY TYPE, SIZE, INDUSTRY]
Person I am meeting: [ROLE, e.g. "operations manager"]
What I already know about them: [e.g. "downloaded our pricing guide, uses spreadsheets today"]
Call length: [e.g. "30 minutes"]
Goal of the call: [e.g. "decide together whether a demo makes sense"]

Produce:
1. A call structure with timings: opening and agenda, discovery, summary, next step.
2. A one-sentence opening that sets the agenda and asks permission to ask questions.
3. 12-15 open questions grouped by theme: current situation, problem and its impact, what they have tried, decision process and timing, success criteria.
4. For the five most important questions, a natural follow-up that goes one level deeper.
5. Signals that this prospect is probably not a good fit, so I can say so honestly.
6. A notes template I can fill in during the call.

Rules: no leading or yes/no questions in the discovery section. Do not include a product pitch. Use plain, conversational language.

Example output

Example (excerpt):
Opening: "Thanks for the time. I'd suggest we spend most of it on how your team handles scheduling today, then decide together if a demo is worth it. Is it all right if I ask a few questions?"

Problem and impact: "When the schedule changes at short notice, what happens next?"
Follow-up: "How often did that happen last month, and who ends up fixing it?"

Not a good fit: fewer than five staff, or no one owns the scheduling problem.

When to use it

Use this prompt before a first call with a qualified lead, when your job is to understand rather than to sell. It helps new salespeople, founders who do their own sales and consultants who need to scope a project. A good discovery call decides whether there is a real problem worth solving, which saves time for both sides.

How to customize it

[WHAT I SELL] helps the model focus questions on problems your product actually addresses, without turning them into a pitch.
[PERSON I AM MEETING] changes the questions. An operations manager cares about daily friction; a finance director cares about cost and risk.
[WHAT I ALREADY KNOW] prevents questions the prospect has already answered, which can feel careless. If your assistant can browse, it can help summarize the company’s public website, but check what it finds.
[GOAL OF THE CALL] shapes the closing. A clear next step is better than a vague “let’s stay in touch”.

Why it works

Discovery fails when the seller talks too much or asks questions that only confirm what they want to hear. The rules against leading questions and pitching keep the focus on the buyer. Grouping questions by theme ensures you cover the decision process and timing, which are easy to forget when the conversation is going well.

Follow-up questions are where real understanding happens; the first answer is often polite and general. The “not a good fit” list encourages honest qualification, which builds trust and keeps your pipeline realistic.

Tips and variations

Do not read the list in order. Pick the questions that fit the conversation and let the buyer talk.
After the call, paste your notes and ask for a summary email that confirms the problem, the impact and the agreed next step. Our meeting notes to action items prompt works well here.
Rehearse with the model playing a busy buyer, using the role-play approach from our salary negotiation prompt.
Record calls only with consent and according to local law; see our article on AI meeting note takers.
Keep a shared list of the questions that produced the most useful answers across calls, and update this prompt with them. Over a few months, your team builds a discovery guide based on real conversations, not theory.

#Sales #Discovery Call #Questions #B2B

More Sales prompts

View all

Salesintermediate

Cold outreach email sequence that respects the reader

You are a B2B copywriter who writes cold emails people actually answer. Write an outreach sequence. What I offer: [PRODUCT OR SERVICE AND THE MAIN PROBLEM IT SOLVES] Proof I can share: [e.g. "case study: cut invoice processing from 5 days to 1 for a 40-person firm"] Target role and company type: [e.g. "finance managers at logistics firms with 50-200 staff"] Why I am contacting this person: [REAL TRIGGER, e.g. "they are hiring two AP clerks", "they spoke at an event"] My ask: [e.g. "a 15-minute call", "permission to send a short video"] Number of emails and spacing: [e.g. "3 emails over 10 days"] Tone: [e.g. "direct and friendly, no exclamation marks"] For each email produce: 1. A subject line under 7 words, without clickbait. 2. A body under 120 words: why this person, the problem in their terms, one piece of proof, one simple ask. 3. A note on what changes from the previous email, so each message adds something new. The last email should be a polite close that makes it easy to say no or to point me to the right person. Rules: use only the facts I gave you. Mark any personalization you would need me to fill in as [CHECK]. No fake "Re:" subjects, no invented mutual connections, no false urgency.

Works with: ChatGPT, Claude, Gemini, Any LLM

Salesadvanced

Proposal executive summary the buyer can forward

You are a proposal writer. Write the executive summary for my proposal. Client and contact: [COMPANY, CONTACT ROLE] Who else will read it: [e.g. "CFO who was not in our meetings"] The client's problem, in their words: [QUOTES OR NOTES FROM DISCOVERY] Impact of the problem: [COST, TIME, RISK, IF KNOWN] What I propose: [SCOPE, DELIVERABLES, TIMELINE] Expected outcome and how it will be measured: [e.g. "reduce onboarding time from 6 to 3 weeks"] Price and payment terms: [AMOUNT, OPTIONS] Why us: [RELEVANT EXPERIENCE OR PROOF] Risks or assumptions: [e.g. "client provides data access by week 2"] Next step and deadline: [e.g. "signed approval by 15 November to start in December"] Produce an executive summary under 350 words with these parts: 1. The situation and problem, written from the client's point of view. 2. The outcome they get, with the measure of success. 3. The approach in three or four steps. 4. Investment and what it includes. 5. Why us, in two sentences with evidence. 6. Assumptions and the next step. Then list any claim that needs evidence or approval before sending, and suggest one headline for the summary. Rules: use only the facts I gave you. If the impact is not quantified, do not invent numbers; write [IMPACT TO CONFIRM]. Avoid jargon the reader outside the project would not know.

Works with: ChatGPT, Claude, Gemini, Any LLM

Salesintermediate

Sales objection handling playbook with practice mode

You are a sales trainer who teaches honest, consultative selling. Build an objection handling playbook. What I sell and the price range: [PRODUCT OR SERVICE, PRICE] Typical buyer: [ROLE AND COMPANY TYPE] Main alternatives they consider: [COMPETITORS, IN-HOUSE SOLUTION, DOING NOTHING] Objections I hear most, in the buyer's words: [e.g. "too expensive", "we're happy with our current tool", "send me some info"] Real limitations of my offer: [e.g. "no mobile app yet", "setup takes 3 weeks"] For each objection produce: 1. What it may really mean (two or three possible underlying concerns). 2. A clarifying question to find out which concern it is, before answering. 3. A short response for each likely concern, in natural spoken language, under 50 words. 4. When the honest answer is that we are not the right fit. Then add: three objections I did not list but should expect, and how to prevent them earlier in the sales process. Rules: never deny a real limitation I listed. Do not use pressure tactics or false scarcity. When I type "PRACTICE", play a sceptical buyer, raise objections one at a time, and after each of my answers give one line of feedback. Stop when I type "END".

Works with: ChatGPT, Claude, Gemini, Any LLM