Filling gaps by inference
When you don’t have a number, the engine can often work it out from your other answers. It labels the result as an inference, shows its working, and waits for you to confirm or reject it.
Not knowing a number shouldn’t stop the run
Most owners can’t say off the top of their head what one seat costs them each month. The engine can often work it out from answers you’ve already given. It fills in the missing number, shows its working, and waits for you to say yes or no.
An inference, worked
In the studio screenshot, the sample charges about $25 a seat each month and keeps 64% of it as gross margin. What’s left is the cost of serving that seat:
When the carrier invoices arrived, the finished report read the real figure from them: $10.01 a seat. The inference was close, and the document replaced it.
“That is right”
The number becomes yours. It counts in full toward the share resting on your numbers, and it’s tested at ±5% instead of ±15%.
“Reject”
The inference is removed and the question goes back to the industry median for your kind of business, counted at zero, until you answer it.
The rules an inference follows
- It is always labelled. An inference never passes for your number.
- It shows its reasoning in one line under the answer.
- It counts at half weight until you confirm it, and it is tested at ±15%, three times as wide as your own figure.
- It is listed in the report’s appendix as “inferred by the advisor, awaiting confirmation”, with the basis for each one.
In the finished sample, one inference was still open at print: the cost to win a new seat, $250, inferred because customers who say they always refer friends usually cost less to win than the $400 median.
One number, one place
The inferred cost per seat moves the value by $0. The valuation reads the gross margin, and the cost per seat is already inside it. The cost line matters to the plan, where it is a lever a team can pull, so it’s asked, inferred and tracked; it just isn’t counted twice.
- Three kinds of fill. An inference is worked out from your other answers and waits for confirmation. An advisor estimate is made from size and business type without your data. A default is the industry median or a simple derivation, such as staff estimated from revenue.
- Weights. Confirmed: full. Inference: half. Estimate and default: zero.
- Where they print. Appendix: estimates, inferences awaiting confirmation, and assumptions, each in its own table with the basis or the typical range.
See it on a real report, or see it on yours.Download a sample reportSchedule demo