Scenario Modelling Agent

At end-of-quarter planning, the agent runs once to build a fixed set of scenarios. It does not recalculate them every day, so leadership can compare three deliberate views of the quarter rather than a scenario that keeps moving.

• Trigger: End-of-quarter planning
• Cadence: Once at planning, not daily

Next, the agent pulls every opportunity that is still open, not just the deals in commit. That full scope matters because a worst case built only from commit deals would leave out the rest of the open pipeline.

• Scope: Every open opportunity
• Include: Open deals whether or not they are in commit
• Exclude: Closed opportunities

The agent then pulls the buying committee linked to each open deal. The scoring step uses that coverage to see how well each deal is threaded, which helps separate a real commit from a hope.

• Scope: The buying committee for each open deal
• Include: The contacts linked to each deal's committee

For every open deal, the agent assigns a probability from that deal's own signals, including how well it is threaded across the buying committee. It never substitutes the default probability for the opportunity's stage, so each score reflects the deal underneath it.

Using the deal-level probabilities, the agent rolls the open opportunities into best case, commit, and worst case. It names the deals that move between the scenarios, so leadership can see exactly what separates one number from the next.

Finally, the agent emails the leadership team the three scenario numbers and the deals that separate them. Keeping the deal names next to the totals makes the discussion about which opportunities move the forecast.

• Recipients: The leadership team
• Include: Best case, commit, and worst case numbers
• Include: The deals that separate the three scenarios