AI how-to guides Customers and sales
How to prepare sales meetings with AI
Connect an AI assistant to the calendar, CRM, email and call notes so it builds a brief before each meeting: who is coming, what has happened so far, open deals, likely needs and suggested questions and content. After the meeting it drafts the follow-up and logs next steps in the CRM. The rep checks the brief, runs the conversation and owns the relationship.
What AI does
- Reads the calendar and builds a brief per meeting: attendees, roles, account history and open opportunities.
- Summarises past emails, calls and notes so the rep knows what was promised and what is pending.
- Suggests discovery questions, likely objections and the case study or demo that fits the stage.
- Drafts the follow-up email and the CRM update after the meeting, from the rep's notes or the recording where allowed.
- Flags deals with missing next steps or that have gone quiet.
What stays with people
- Reading the room, running the conversation and building the relationship.
- Checking the brief and correcting what is wrong or out of date.
- Approving what is sent to the customer and what is committed.
How to put it in place
-
Define what a good brief contains
Agree with sales leadership on the sections every brief needs, by segment and sales stage. Use your sales methodology so preparation is consistent across the team, not only for top performers.
-
Connect calendar, CRM, email and content
Give the assistant read access to the rep's calendar, the CRM records, relevant email threads and the approved sales content. Include call recordings or transcripts only where policy and consent allow it.
-
Generate briefs grounded in sources
Have the assistant produce a brief ahead of each external meeting, with every fact linked to the record or email it came from. Public company information can be added, marked as such.
-
Close the loop after the meeting
From the rep's notes or the transcript, draft the follow-up email and the CRM update with next steps and dates. The rep reviews and approves both before anything is sent or saved.
-
Pilot with one team and compare
Start with one region or team and compare preparation time, CRM data quality and pipeline progress with a similar group. Tune the brief with the reps' feedback before widening.
What you need
- Access to calendar and email through their APIs (for example Microsoft Graph or Google Workspace).
- A CRM with reasonably current accounts, contacts and opportunities.
- Approved sales content: case studies, decks, pricing rules and objection handling.
- A clear policy on call recording, consent and personal data.
- Sales managers who define the playbook and review quality.
When it is not worth it
- Very few, very large accounts where the account team already knows every detail.
- When the CRM is barely used and email and call history cannot be connected.
- Transactional sales where meetings are rare and preparation is not the bottleneck.
Risks and how to control them
- Wrong or outdated facts in the brief: link every fact to its source and show when it was last updated.
- Personal data about contacts: process only what is needed for the relationship, follow data protection rules and respect recording consent.
- Follow-ups sent with errors or promises the company cannot keep: keep the rep's approval on every customer-facing message.
- A messy CRM produces messy briefs: start with the accounts whose data is in order and use the assistant to improve logging over time.
How Thinkia does it
We start by defining segments, methodology and must-have fields so coaching stays consistent. Opportunities, emails and call recordings, where allowed, then feed grounded summaries, next steps and content suggestions for each stage of the deal.
A region or team pilots the assistant first and we compare pipeline velocity and forecast accuracy. Playbooks, approvals and analytics then spread, so managers get visibility without extra shadow IT.
The Thinkia solution behind this guide
- AI sales assistantYour reps should be selling. Not updating the CRM.
- AI customer insightsKnow what your customers actually think. Not what the survey says.
Thinkia products involved
Frequently asked questions
What should an AI meeting brief include?
Who is attending and their role, the account and deal history, what was promised last time, open issues, likely needs and objections, suggested questions and the content that fits the stage.
Can AI update the CRM after a sales meeting?
Yes. It can draft the meeting summary, next steps and field updates from notes or a transcript. The usual approach is that the rep reviews them before they are saved.
Is it legal to record sales calls for AI?
It depends on the jurisdiction and on how you inform and obtain consent from participants. Agree the policy with your legal and data protection team before switching recording on.
Does it work if our CRM data is poor?
Partly. The brief will only be as good as its sources, so start with accounts whose data is reasonable. Automatic logging after meetings then improves the CRM over time.
Is this different from a general AI assistant?
A general assistant can summarise what you paste into it. A sales assistant is connected to your calendar, CRM and approved content and follows your playbook, so the brief is ready without searching.