Skip to main content
Deal Analysis reads the conversations behind each deal in your CRM, including WhatsApp chats and email threads with the deal’s contacts, and records what it finds on the deal record: who owes the next reply, whether engagement is fading, what could block the close, and when the customer expects to sign. It runs automatically every 6 hours, so every deal you open is already current. There is nothing to set up or trigger.
Requirements: a connected CRM, with contacts associated with your deals. A deal with no contacts attached is analysed from CRM data only. On HubSpot, reading email activity also requires the sales-email-read scope.

​
How It Works

Each analysis draws on two sources. The two sources are compared. Where the CRM record and the conversation differ, the difference is recorded rather than resolved. Each run refreshes deals that have had new activity since they were last analysed. Every deal records the time of its most recent analysis.
Results are written to separate, dedicated fields. Existing deal fields such as stage, amount, and close date are never modified.

​
Where to Find It

Open any deal in your CRM. The analysis fields appear alongside the deal’s existing fields on the record.

​
What the Analysis Records

​
Filtering and Reporting

Results are stored as standard deal fields, so they work in filters, saved views, dashboards, and reports like any other field. In your CRM’s field list they appear together, under names beginning with AI. Views teams commonly save:
  • Deals awaiting a reply from your team
  • Deals with a declining engagement trend
  • Deals where the stage and the conversation disagree
  • Deals expected to close in the current month

​
Behaviour to Be Aware Of

  • An empty field means no signal was found. Values are written only when there is supporting evidence. An empty field indicates the information was not present in the CRM or the conversations.
  • Inactivity is measured in days. A deal is not marked inactive as soon as a reply is late. The customer must have left your last message unanswered for at least two days.
  • Values can change without new messages. Inactivity and engagement trend are based on elapsed time and on comparison with earlier activity, so results can change during a period of silence.
  • Stated timings are standardised. Phrases such as “in a few weeks” are converted into a fixed set of timeframes, calculated from the current date. Only closing commitments are counted, not scheduling for meetings or demos.
  • Manual edits are overwritten. You can edit an analysis field directly, but the next scheduled run replaces the value. Use your own fields or a deal note to record your own assessment.

​
Troubleshooting

​
FAQ

No. Results are written to their own dedicated fields. Deal stage, amount, close date, and any other field your team maintains are never edited or overwritten. Where the analysis disagrees with the CRM record, it records the difference in its own field.
Every 6 hours. Each run refreshes deals that have had new activity since they were last analysed, and the results are written back to your CRM automatically. Each deal shows when it was last analysed.
Each result is linked to the message or activity it came from. If a value looks wrong, check the supporting evidence recorded on the deal.
Yes, but with less data behind the results. Each deal records how many conversations were read, so you can judge how much a result is based on.

BEA Radar Overview

The WhatsApp side of BEA Radar, applied to individual chats

AI Filters

Filter your Team Inbox by what the analysis has identified