Compare
Not another confident number.
A forecast is a disciplined act: it makes a team put numbers on what it believes, and numbers can be checked. The danger arrives when the forecast becomes a single figure on a slide. The assumptions that produced it fall out of view, and a guess acquires the authority of arithmetic.
Typical forecast
- offers one figure where a range belongs
- buries the assumptions underneath
- mistakes precision for certainty
- rarely says what would reverse it
- is seldom checked against what happened
Argumont
- keeps each assumption visible and rated
- sets the plan against your own history
- runs the scenarios that would break it
- states what would change the Verdict
- can return No Strong Winner Yet
What forecasting does well
A written forecast turns a mood into a claim. Once a team has said that 1,200 customers will subscribe in the first year, it can be wrong in a measurable way, and that is where learning begins. Comparing the forecast with what later happened is how forecasters improve.
The inside view and the outside view
In 1979 Daniel Kahneman and Amos Tversky named the planning fallacy: people forecast their own projects from the inside, picturing how this plan will unfold, and so underestimate time and cost. The remedy they and later researchers proposed is the outside view, which starts from how similar projects actually turned out and only then adjusts for what is special about this one.
Most forecasts skip that step. The base rate often sits in the company's own history, in the last launch or the last migration, and nobody looks it up.
Precision is not certainty
A forecast of $417,600 looks better informed than one of about $400,000, but the extra digits come from multiplication, not from knowledge. State a range, name the two or three assumptions that move the result most, and write down in advance what you would need to see to change course. Then keep the forecast, so the comparison with reality can actually happen.
A forecast in a Decision Room
The room treats each assumption as a claim. Inspector marks which ones a source or the company's own records support and which are hopes, and the engine works the sums from your figures exactly. The Champion for the plan defends the base case, the Challenger aims at the assumption the plan leans on hardest, and the Red Teamer runs the scenarios that would break it.
The Verdict's confidence describes how strong the record is, never the probability that the plan succeeds. When the record cannot separate the options, the room says No Strong Winner Yet, and the Human Decision is made knowing that.
Four percent or one and a half
A software company forecasts 1,200 subscribers for a new paid tier in its first year, from 30,000 free users converting at 4%. At $29 a month that is $417,600 a year in recurring revenue, against $180,000 a year to build and run the tier. Inspector marks the 4% as an assumption with no source, while the company's last paid add-on converted 1.5% of users, which its own records support.
At 1.5% the tier reaches 450 subscribers and $156,600, short of its cost. The tier pays only above about 518 subscribers, a conversion rate near 1.7%. The Verdict favors a two-month pre-sale test before any build, and names that threshold as the line.
Who owns the bet
A forecast informs a bet, and the bet belongs to the people who will live with it. The room's Verdict and the Human Decision sit side by side, and the Outcome can be added once the year is out, which is how a team learns whether its forecasts deserve trust. The room offers decision support, never investment advice.
Bring Argumont a Decision worth challenging.
Start with a paid 60 day pilot on your own real Decisions.