What Entrepreneurs Can Learn from Superforecasting
Running a business means making a stream of forecasts — will this feature drive signups, will this hire work out, will the cash last until the next round or the next quarter of revenue — and most founders make them the same way they make a pitch: with confident, rounded-off language that sounds decisive but hides how little they actually know. Superforecasting offers a different way to hold uncertainty, one that doesn't require pretending to be sure. It treats forecasting as a skill you can practice and improve, not a trait some founders have and others don't.
The whole book, compressed
This edition works through 4 core lessons — take them one at a time, not all at once.
What's different for your conditions
Tap whichever ones sound like you. Nothing here is sent or saved — it's just for your own noticing.
0 of 5 selected
The core lessons, applied to you
Turn "will this work?" into smaller, answerable questions
A vague question like "will this product succeed?" can't really be answered. It gets tractable once you split it into pieces: who buys in the first month, what makes them stay past week two, what does a competitor response look like.
A founder tells the board "we're confident Q3 will be strong" based on a good gut feeling from recent customer calls. When Q3 underperforms, nobody can say which part of that feeling was wrong — was it the sales cycle, the pricing, the onboarding — because the confidence was never broken into parts that could be checked.
Before your next planning meeting, take your single biggest uncertain claim about the business and write down three to five sub-questions that, if you knew the answers, would tell you whether the big claim holds.
Trade slogans for numbers you can update
"Probably" and "should be fine" feel natural but can't be checked against reality later. Putting a percentage on a forecast, and nudging it as new information comes in, is what actually sharpens judgment over time.
A founder says the new hire will "definitely be ramped up by month two." Two months in, ramp is slow, and because the original claim was never a number, there's no clean way to say how wrong the estimate was or what to adjust next time.
Put a specific percentage on your next big internal forecast (e.g., "70% chance we hit $50K MRR by the end of the quarter"), write it down with the date, and revise it in writing whenever a new data point arrives.
Pull in views that don't already agree with you
Founders tend to surround themselves with people who already believe in the vision. That's useful for morale and fatal for forecasting, because it filters out exactly the perspectives that would catch a blind spot.
A founder plans a launch timeline using input only from the engineering team, who are optimistic about their own velocity. A support or sales perspective would have flagged the onboarding bottleneck that ends up delaying the real launch by six weeks.
Before locking in a major forecast, deliberately ask one person who has expressed doubt or comes from a different function to poke holes in it, and actually log what they say.
Keep score on your own calls
None of this pays off unless you look back. Keeping a simple, honest record of what you predicted and what actually happened — without turning it into self-criticism — is what turns forecasting from a gut habit into a skill that compounds. Over enough cycles, this record shows you exactly where your judgment tends to run hot or cold, which is far more useful than a general sense of confidence.
Breaking questions into parts and seeking outside views both take time you often don't have before a decision is due; the practical answer is to scale the rigor to the stakes — a two-minute version for small calls, the full version for anything that could sink the business.
Name one thing, honestly
Which recurring business forecast do you currently make with a confident phrase instead of a number, and what would change if you started writing down an actual percentage?
Good. Now look back at the "Try this" line under whichever lesson feels most connected to what you just wrote, and treat that as this week's actual assignment — not all 4 at once.
The practices from this edition
0 of 3 selected
Run this for seven days
Before locking in a major forecast, deliberately ask one person who has expressed doubt or comes from a different function to poke holes in it, and actually log what they say.