Monte Carlo Forecasting in Agile - Anyone Actually Using It Successfully?

[Baselinerai]

Forecasting in Agile always feels a bit fragile. Velocity shifts, scope changes, people go on leave, priorities move. Yet we’re still expected to give confident delivery dates.

Lately I’ve been digging into Monte Carlo Forecasting Agile approaches as an alternative to traditional estimation. Instead of locking onto a single forecast, it uses historical delivery data to run thousands of simulations and produce probability-based outcomes. In theory, that sounds far more aligned with how Agile teams actually work.

What I’m curious about is the practical side:

• Are teams really trusting probability ranges over fixed dates?
• What data are you feeding the simulations (cycle time vs throughput)?
• How long did it take stakeholders to “get” the results?

I’ve seen tools like Baseliner Ai applying Monte Carlo forecasting directly to Agile delivery data, which removes a lot of the manual modeling. But I’m wondering whether the real challenge is less about tooling and more about changing the conversation—from certainty to confidence levels.