The Predictive Insights section appears below the standard charts on the Analytics page. It computes four types of statistics directly from your account's own data. No AI or machine learning is involved — every number comes from your leads and transactions, using transparent formulas.
Weighted pipeline forecast
Each open transaction's assignment fee is multiplied by a stage weight reflecting how likely a deal in that stage is to close — Contract Signed starts at 50%, rising through Funded at 95%. Once your account has five or more resolved leads, the weights are automatically blended 50/50 with your own historical close rate, so the forecast improves as you accumulate data (a "blended" badge appears when active). The forecast always covers all active transactions regardless of the selected date range, because pipeline is a snapshot metric.
Lead volume trend
Compares lead intake this week versus last week (week-over-week %) and draws a trend direction using linear regression over your daily lead counts in the selected range. "Growing," "flat," or "declining" describes the direction of the regression line — it takes a few weeks of data to be meaningful.
What converts: by source and by situation
Two ranked lists showing your conversion rate per lead source and per lead situation. Conversion means a lead reached Contract or Closed status. Groups with fewer than 5 leads are marked "low sample" — the math is correct, but a single converted lead in a tiny group can look deceptively high. These panels respect your selected date range.
Avg time in current stage
Shows how long active transactions have been in each stage on average, flagging stages above the overall median as bottlenecks. One honest limitation: DispoStack does not yet record when a transaction entered its current stage, so this measures transaction age, not true time-in-stage. The limitation is shown in the UI.
All methodology notes are rendered directly under each panel, so you can always see exactly how a number was computed.