Assumption Chain Calculator
Multiply a chain of assumptions — a market size, a reach, a conversion, a retention — into one number, and see which link would move the result most if it were 20% wrong.
Any top-down estimate is a chain: total market, times the share you can reach, times the fraction who convert, times what each is worth.
How the assumption chain calculator works
Any top-down estimate is a chain: total market, times the share you can reach, times the fraction who convert, times what each is worth. The result is the product, and every link is a guess.
Because it is a product, a 20% error in any link is a 20% error in the answer — but the links are not equally uncertain. This ranks them by how confident you say you are, so the least certain link is the one you go and check.
Formula: result = start × Π rate_i; a 20% error in link i moves the result 20%
Worked examples
| Inputs | Result of the chain | Note |
|---|---|---|
| A five-link funnel | 9,450 | about 9,450 from two million |
| A weaker middle | 5,040 | the result nearly halves |
| One link measured | 9,450 | the weakest link changes |
FAQFrequently asked questions
What is a top-down estimate?
A total, narrowed by a chain of fractions: market, times reachable, times convert, times retain. Every step is an assumption and the result is their product.
Why rank by confidence rather than size?
Because in a product every link has the same leverage — 20% wrong anywhere is 20% wrong overall. What differs is how likely each one is to be wrong.
What does the confidence scale mean?
One is a guess, five is measured data. Three is a reasoned estimate from something comparable.
How wrong are these chains usually?
Badly. Five links each 20% optimistic compound to a result 2.5 times too high. That is why checking the weakest link first matters.
Should I use ranges instead?
When it matters, yes — the Monte Carlo page takes a low, likely and high for each link and gives you a spread rather than one number.
Where these figures come from
- Hubbard (2014) — How to Measure Anything — the case for ranges over point estimates
- Vose (2008) — Risk Analysis: A Quantitative Guide — Monte Carlo and the triangular distribution
- Saltelli et al. (2008) — Global Sensitivity Analysis: The Primer — one-at-a-time sensitivity and its limits
- NIST — Uncertainty analysis guidance — the US measurement standards body
Last checked: September 2026. The methods are textbook decision analysis: triangular distributions for three-point estimates, tornado ranking for sensitivity, and simple additive weighting for decision matrices.