How Ensaria's Reality Engine does it
Every task you create gets auto-tagged from its title + project + client + history. When you set an estimate, Ensaria pulls the matching reference class from your last 90 days and shows the median duration, the p25–p75 range, sample size, and confidence (low / med / high based on sample + variance).
If the sample is too thin — fewer than three similar tasks — it tells you that instead of guessing. The Money Lens never makes up a number; it just shows you the truth in your own data.
What this changes about quoting
When a client asks “how long for X?”, you no longer answer from gut. You answer from “tasks like this have taken me median 4.2 days, range 3.1–5.8, across 14 instances over the last year.” That's a quote that's harder to negotiate down and easier to deliver on time. Read more on reference-class forecasting and the planning fallacy.