Prepare a sales meeting with Claude: research, discovery and objection practice

Set a realistic meeting objective

Choose the decision the conversation should enable: understand a difficulty, clarify scope or agree a pilot. A first meeting may not establish budget, timing and contract approval at once. Define the expected outcome before researching the prospect.

Write down your own limits: available services, questions requiring a colleague and commitments you cannot make during the meeting. A role-play should not train you to promise nonexistent features.

Build a brief with dated sources

Use the company website, relevant company publications and information the contact has voluntarily provided. For each important fact, retain the source and date. An old announcement does not prove that a project remains active.

Ask Claude to separate sourced facts, hypotheses to test and open questions. “The company lists three offices” is a verifiable observation. “Its teams coordinate badly” remains a hypothesis. Avoid personal research unrelated to the business discussion.

When assisted research supplies citations, open the pages and locate the supporting passages. A relevant URL alone does not substantiate the claim. Our document-summary workflow explains how to preserve this traceability.

Prioritize discovery questions

Start with the current process, then explore the consequences of a problem and the desired improvement. Avoid a question that already assumes the need for your solution.

Choose five priority questions and a few follow-ups. Thirty questions can turn a conversation into an interrogation. Plan a midpoint restatement so the prospect can correct your understanding.

Practice objections as fictional scenarios

Define a role-play: a prospect interested in a pilot, concerned about confidentiality and with an undisclosed budget. Make clear that generated objections are practice scenarios, not predictions about the real person.

After each reply, ask for feedback on listening, precision and respect for the offer’s limits. “We have no time” may call for clarification of the constraint rather than a promise of immediate installation. Practice saying that you will verify a technical answer when necessary.

Repeat with a skeptical interlocutor and one who provides contradictory information. Look for places where you rush to conclusions. A flattering model-generated score is not proof that the meeting will succeed.

Create a prospect brief separating dated sourced facts, hypotheses and unknown information. Suggest five open questions about the current process, main difficulty, success criteria and pilot approval. Then role-play a fictional skeptical prospect. Invent no budget or commitment, and flag promises outside my stated offer.

Prepare the follow-up before the meeting

Open a record template with confirmed needs, constraints, unknowns, actions, owners and deadlines. Take notes relevant to the conversation. If you intend to record or transcribe it, follow your organization’s procedure for informing participants and handling the data.

Afterwards, distinguish suggestions from decisions. “We could try in November” does not mean a pilot was approved. Review the follow-up before sending and ask for confirmation of important points. A meeting-notes Skill can package this procedure.

Download the meeting preparation sheet. Once the need is confirmed, use the sales proposal workflow.

Sources

• →Claude Academy.
• →Anthropic: create and manage Projects.
• →Claude Code: verification practices.