All guides

Workflow guides · 3 min read

Customer call prep in 60 seconds

How to prep for a customer call using Recon and your AI tool. Real prompts, real flow, real time savings.

The job: you have a customer call in 15 minutes. You need to know what's happened with this account recently so you walk in informed instead of asking the customer to remind you.

Without Recon: 15 minutes of tab switching across Salesforce, Stripe, Zendesk, Linear, your notes app, and Slack.

With Recon: one prompt to Claude or ChatGPT, 60 seconds, you have the brief.

The prompt

I have a call with Acme in 15 minutes. Use Recon to brief me on:
1. Their current plan and renewal date
2. Any open support escalations or engineering work
3. Recent product activity that's notable (drops, surges)
4. The last few touchpoints with them
5. Any active risk signals

Keep it under 200 words.

What happens behind the scenes

Your AI tool calls Recon's get_account tool with Acme as the identifier. Recon returns structured JSON with:

  • Plan, seats, billing info from your billing connector
  • Active claims (synthesized from the dreaming pass)
  • Recent observations (raw signals: tickets, deploys, conversations)
  • Linked engineering work (Linear issues tagged with the customer)
  • Risk score and memory_state

Your AI tool reads that and writes the 200-word brief.

A real example brief

Acme Corp · Pro · 52 seats · renews June 12 (41 days)

Status: at_risk (risk_score 0.78)

Active concerns:

  • Three Zendesk escalations this week about exports timing out. All linked to REC-447 in Linear (open, no estimate).
  • Workflows usage dropped from 4× median to flat after Tuesday. Their power user (T. Chen) hasn't logged in since.
  • Last QBR (May 8) flagged the same export issue. We promised an update.

Recommended talk track:

  • Acknowledge the export issue and give an honest engineering ETA.
  • Ask about T. Chen. Possible champion turnover signal.
  • Don't push the renewal expansion. Wait until the export issue resolves.

Sources: 3 zendesk tickets, REC-447, postgres usage data, last QBR notes.

Variations

For prep before a quarterly review (QBR):

QBR with Acme tomorrow. Use Recon to give me:
- The story since our last QBR (use search with the last QBR date)
- Top 3 wins for them this quarter
- Top 3 risks
- Two questions I should ask them

For prep before a sales-expansion call:

Use Recon to find evidence that Acme is hitting Pro plan limits. Look
at usage patterns, feature adoption depth, and any conversations
where they mentioned wanting more.

For prep before a churn-save call:

Acme gave notice. Use Recon to tell me what we missed: signals in the
last 90 days that we should have caught, themes from their tickets,
and what the closest comparable account looks like.

Start with a snapshot

Connect your first source in under 10 minutes.

Sign up free. No card required. Your AI tools get customer memory the moment discovery finishes.

Start free