How to use Monte Carlo forecasting
A single delivery date is a guess. A Monte Carlo forecast is a probability distribution, and that's what lets you commit honestly.
Why it matters
"We'll ship in March" sounds confident. It's also wrong almost every time. "P50 March 18, P85 April 12" tells the business what to plan against and what to budget for in case of slip. The Monte Carlo engine in Tenhaw runs against your real historical throughput, not estimates.
Step by step
- Make sure stories are pointed. Forecasts only work if there's something to count. Unpointed stories become noise.
- Make sure throughput data exists. The engine needs at least 6–8 weeks of completed work to sample from.
- Open the value tracker. For the epic you're forecasting. The engine reports P50, P85, P95 dates plus the sprint distribution.
- Quote the band, not the single date. "Most likely March 18; allow until April 12 for confidence."
- Re-forecast weekly. Throughput shifts; the forecast should follow.
- Pair with the Scenario Simulator. For what-ifs (cut scope 20%, add 1 engineer, slip the deadline 2 weeks).
What good looks like
- All in-flight stories pointed
- ≥ 8 weeks of throughput history
- Communicated dates are bands (P50 / P85), not single points
- Re-forecasted weekly
- What-if questions answered with the simulator, not vibes
How Tenhaw runs this
Tenhaw's Monte Carlo engine runs 10,000 simulations per epic against your real historical throughput. No setup, no spreadsheet maintenance. P50/P85/P95 dates land on every epic's value tracker and re-run automatically on every story state change. Pair with the Scenario Simulator for what-ifs.
- Tenhaw Monte Carlo simulation engine
- Tenhaw value tracker
- Tenhaw Scenario Simulator
- Tenhaw delivery analytics
In Tenhaw → Delivery Intelligence → any epic. Forecasts run on save and on every story-state change.
Run this in your team
The Tenhaw product enforces every rule in this playbook so it doesn't sit on a wiki gathering dust. Try Tenhaw or book a working session with the founder.