Flag at-risk deals from pipeline activity
Each day the system checks open deals for stale dates, missing next steps, silent champions and competitor mentions, then sends the manager a short list with the evidence. Forecast reviews start from checked data. For sales managers who find out a deal is stuck in the last week of the quarter.
Function: Sales · Industry: Any industry
Flag at-risk deals from pipeline activity
6 steps
At a glance
- Trigger
- A daily schedule runs over open opportunities.
- Typical inputs
- Opportunity fields and history
- Email and meeting activity
- Call transcripts and competitor mentions
- Stage exit criteria
- Systems it usually connects to
- Salesforce
- HubSpot
- Clari
- Slack
- How the work splits
4 rules, 1 model, 1 person steps
How it runs, step by step
6 steps, each tagged by what does the work. Rules are fixed logic, a model reads or drafts, and a person decides.
- Rules Fixed logic in code
- Model AI reads or drafts, with a confidence score
- Person A named person decides
- 1
Run the hygiene checks
RulesRules find past close dates, missing next steps, stage jumps and deals with no activity in your set number of days.
- 2
Read recent conversations
ModelA model looks for signals in emails and calls, such as a champion going quiet, a new stakeholder or a competitor named.
- 3
Combine the evidence
RulesFindings are grouped per deal with links to the records behind each one.
- 4
Send the digest
RulesManagers receive a daily or weekly list of deals needing attention, ordered by amount and close date.
- 5
Decide the action
PersonThe manager and the rep agree what to do, such as a call with the sponsor, a mutual plan or a downgrade.
- 6
Record the outcome
RulesActions taken are logged, so you can see how flagged deals ended.
Where people approve
The decisions that stay with your team, whatever the confidence score says.
- Managers set the forecast category for each deal.
- Reps confirm facts before a deal is downgraded.
- Alert rules are owned by sales operations.
KPIs to track
What to measure, before and after. Targets come from your own baseline.
- 01Open deals with complete required fields
- 02Flagged deals that changed stage or date after review
- 03Slipped close dates caught in advance
- 04Manager time spent preparing for forecast calls
Questions about this workflow
Is this a forecasting model?
No. It highlights inconsistencies and signals for people to look at. Forecast categories are still set by managers. A statistical forecast would need its own test against your history.
Will reps resent the alerts?
They are less annoying when they are specific and right. Each alert shows the evidence, reps can dismiss with a reason, and dismissals feed the tuning.
Which steps use rules, a model and a person?
Rules: run the hygiene checks; combine the evidence; send the digest; record the outcome. Model: read recent conversations. Person: decide the action. Rules are fixed logic in code. A model reads or drafts and reports its confidence. A named person decides.
What do we need to start?
Typically: opportunity fields and history; email and meeting activity; call transcripts and competitor mentions; stage exit criteria. You also need read access to Salesforce, HubSpot, Clari and Slack, or the equivalents you use. A workflow audit measures your volumes and past cases first, then gives a fixed price for the build, so you can decide with real numbers.