As of 13 August 2026, AI can only partly book sales meetings with prospects.

What to actually do

1. Hand it to a person

   The route this page recommends

   A person who owns the outcome does this end to end, worth it when the failure is dear.

2. Use a tool built for this

   Second choice

3. Do it yourself

   The distant third

   A chat interface, power-user skill, and roughly 1 hour until you can act on the result.

   How to actually do it

   1. Open your CRM, calendar and approved product documentation, then gather the ideal customer profile, exclusions, verified claims, meeting purpose, availability and contact-permission rules.
   2. Create a clean prospect list with company, role, contact details, source, existing-customer status, suppression status and any available buying-situation information.
   3. Paste the campaign information and prospect-list specification into the prompt, then ask the chatbot to produce the qualification rules, sequence, reply tree and handover template.
   4. Compare every drafted claim and personalisation point against your current product documentation and prospect records, deleting anything unsupported or uncertain.
   5. Configure Apollo to import or identify the permitted audience, enrich the records and run the approved outreach sequence, with exclusions and suppression rules applied before sending.
   6. Connect the booking workflow to the correct calendar, test an interested reply and confirm that a meeting appears in the calendar before treating it as booked.
   7. Send a small approved test campaign, inspect the messages and replies, then adjust the qualification rules and sequence before wider use.
   8. Review each booked meeting against the qualification definition and give the salesperson the verified handover note before the meeting.

   Prompt

   Act as a UK B2B sales operations assistant. Design a prospecting campaign that aims to book qualified sales meetings without inventing facts or contacting people outside the permitted audience.
   
   Offer: [describe the product or service]
   Ideal customer profile: [industry, organisation size, location, job roles and buying situation]
   Exclusions: [competitors, existing customers, unsuitable sectors or roles]
   Customer evidence and approved claims: [paste only claims that have been checked]
   Meeting purpose: [what the meeting should decide or explore]
   Meeting length and availability: [paste calendar availability or booking-link details]
   Brand and tone: [describe the permitted tone]
   Prospect data and source: [describe the fields, source and permission status]
   
   Produce:
   1. A clear qualification definition for a meeting worth booking.
   2. A prospect-list specification with required fields and exclusion checks.
   3. A short, plain-text outreach sequence with an initial message and follow-ups.
   4. A personalisation plan that uses only supplied or verifiable information, and marks anything needing human confirmation.
   5. A reply-handling decision tree for interested, unsure, not interested, unsubscribe and out-of-office replies.
   6. A booking workflow that asks for the minimum information needed and does not claim a meeting is booked until the calendar confirms it.
   7. A handover note template for each booked meeting.
   8. A pre-launch checklist covering factual accuracy, data protection, contact permissions, suppression lists, sender identity, unsubscribe handling and calendar access.
   
   Do not invent prospect details, case studies, results, legal permissions, availability or product features. Do not send anything. Flag decisions that require a sales owner or professional advice.

   Open it prefilled in ChatGPT or Claude, or copy it into Gemini, which takes no prefill link.