Prompts for Salespeople: Prepare a Meeting, Follow Up, Handle Objections

A salesperson who sells scheduling software to veterinary clinics lands a meeting for Thursday with a clinic he knows almost nothing about. He has thirty minutes between two trips. He can skim the clinic's website, or prepare the meeting with the three questions that will get the director talking.

AI helps in moments like this, as long as you don't ask it to sell in your place. It structures, rephrases, practices with you. It knows neither your client, nor your product, nor the history of your relationship: you bring those facts.

Prepare a meeting in ten minutes

Paste what you know, including what seems trivial: the AI will tell you what's missing and which questions to ask. Don't put in personal data about individuals, only professional and public information.

I'm a salesperson at [COMPANY], I sell [PRODUCT/SERVICE IN ONE SENTENCE] to [TYPE OF CUSTOMER].
I have a 45-minute meeting on Thursday with [CONTACT'S TITLE] at [CLIENT COMPANY NAME].

What I know about them: [SIZE, ACTIVITY, NEWS FROM THEIR WEBSITE, WHAT THE CONTACT TOLD ME ON THE PHONE]
What I don't know: [LIST]
My goal for this meeting: [E.G. GET A 15-DAY TRIAL]

Prepare for me:
1. Five open questions to understand their current situation, from the broadest to the most specific.
2. The three most likely reasons they wouldn't change anything, worded the way they would say them.
3. A three-part meeting plan with the amount of speaking time I should leave them.

Don't attribute any problem to them that I haven't mentioned: put forward hypotheses, marked as such.

The result is a conversation base, not a script. The sentence about hypotheses matters: without it, the AI describes the industry's "classic problems" as if they were your client's, and you risk announcing them to him with a confident air.

Following up without "just circling back"

Most follow-ups fail because they contain nothing new. A useful follow-up brings one thing: an answer to a question asked, a similar example, a document.

Write a follow-up email of 80 words maximum, professional and formal tone, for [FIRST NAME LAST NAME / TITLE].

History: we spoke on [DATE]. What they told me: [QUOTE OR SUMMARY OF THEIR NEED].
What I'm bringing them today: [NEW ELEMENT: A SIMILAR CLIENT CASE, A FIGURE FROM THEIR INDUSTRY TAKEN FROM A SOURCE I CITE, AN ANSWER TO THEIR QUESTION]
Proposed next step: [15-MINUTE CALL, DEMO, QUOTE]

Subject line of 6 words maximum, without "follow-up" or "following up." No "just circling back," no "don't hesitate to reach out." End with an offer of two time slots.

You supply the figure or the client case: a follow-up that contains an invented statistic is quickly spotted. Email prospecting is governed by rules on consent and opting out; check them before automating anything (in France, on the CNIL's website). For more general messages, look at the prompts for professional emails.

Practicing objections

This may be the most interesting use: the AI plays the difficult client, and you practice before facing the real one.

Play the role of [TITLE], director of a 12-person veterinary clinic. You're in a hurry, you already have software that works, and you don't like having your time wasted.

I'm a salesperson and I'm proposing [PRODUCT] to you. Run the exchange like a real contact: raise realistic objections (price, switching time, fear of losing data), one at a time, and react to what I answer. Don't help me.

After six exchanges, step out of the role and give me: what was convincing, what was vague, the answer I couldn't give.

This exercise takes a bit of discipline. The AI is naturally accommodating and ends up giving in too fast: ask it to stick to its refusal when your answer doesn't convince it. To reuse this kind of scenario with changing cases, also read the article on giving AI a role.

Meeting notes ready for the CRM

After the meeting, you have ten minutes of scattered notes. Turn them into a record that someone else can read.

Here are my meeting notes, scattered:
[NOTES]

Turn them into a CRM record: need expressed, budget mentioned (or "not discussed"), decision makers and the role of each, objections, next steps with dates, main risk. Quote only my notes and write "not discussed" when the information is missing.

What doesn't work

The AI doesn't know whether a client is ready to sign, and it doesn't know the discounts you can grant. It also shouldn't promise a feature your product doesn't have: reread every sentence of a message before sending. To find out whether a follow-up prompt really works, test it on ten cases before adopting it as a routine.

To try it, take your next meeting and prepare it with the first prompt. Afterward, note which of the AI's questions were useful to you, and which one you wouldn't have asked on your own.