Forecast vs. Actuals

A forecast that is never compared to actuals is not a forecast — it is creative writing. The discipline of forecast-to-actual reconciliation is where financial modeling becomes a learning system. Every variance tells you something: either your assumption was wrong, your data was wrong, or something changed in the environment. The reconciliation process turns misses into lessons and lessons into better assumptions for the next cycle.

1. Monthly Variance Decomposition Break the forecast miss into its component assumptions. Did new ARR miss because of fewer deals (volume), smaller deals (price), or longer cycles (timing)? Did churn exceed forecast because of a specific cohort, a product issue, or a competitive displacement? The aggregate miss is the symptom. The assumption-level variance is the diagnosis.
2. Waterfall Reconciliation Build a variance waterfall: start with the forecasted number, add or subtract each assumption-level variance, and arrive at the actual. Forecast: $4.8M. Win rate variance: -$600K. Deal size variance: +$200K. Timing variance: -$300K. Actual: $4.1M. Now you know the miss was primarily a win rate problem, partially offset by larger deals.
3. Assumption Update Cadence Update your assumption registry monthly with trailing actuals. If win rate was assumed at 24% and the trailing 3-month actual is 21%, update the base case. Do not wait for the quarterly board review to discover that your forecast has been drifting for 90 days. The earlier you catch the drift, the more options you have.