Modeling Uncertainty

Uncertainty isn’t the problem. Leaving it implicit is.

Every new-business strategy is a hypothesis — a structured set of assumptions about a future that hasn’t happened yet. Most strategy work buries that uncertainty inside confident-sounding point estimates and vague words such as likely or high-confidence.

BRI’s approach does the opposite. It makes uncertainty explicit, so it can be measured, prioritized, and deliberately reduced — turning a strategy from a narrative you defend into a hypothesis you can evaluate and improve.

Every element of a strategy is one of three things

Before you can model uncertainty, you have to name it. In BRI’s Strategy Hypothesis Model, every element of a strategy is one of three types — a distinction most frameworks never formalize, and the basis for disciplined evidence prioritization.

Mixing these up — treating an assertion as a choice, or hiding an uncertainty behind a point estimate — is one of the most common ways strategy work goes quietly wrong.

Quantify uncertainty as a range, not a point

A point estimate hides everything that matters about an uncertain input. A range makes it visible. For each uncertain value, you describe the shape of its probable distribution:

• Expected — the single value you’d bet on if you had to choose one. It need not sit at the midpoint of the range: a market-share input might have an expected value of 8% even though it ranges from 3% to 25%, because the downside is more constrained than the upside.
• Min, the 10th percentile — a low outcome you’d be surprised to fall short of; only about a 1-in-10 chance the true value lands below it. If you expect 8% share, your min might be 3% — disappointing, but not the worst imaginable.
• Max, the 90th percentile — a high outcome you’d be surprised to beat; only about a 1-in-10 chance the true value lands above it — say, 25% share if things break your way.
• Never-exceed limits — hard logical bounds the value simply cannot cross, which keep the model honest as it samples across the range. Market share can’t fall below 0% or rise above 100%, and a price can’t go negative. Where there is no natural ceiling — units sold, say — set the upper limit deliberately high so it never distorts the result.

Defined this way, every uncertain input becomes something the model can reason about — and something evidence can tighten.

Raise the evidence. Tighten the range.

Not every assumption deserves equal attention. BRI scores each one on two axes — its potential impact on the outcome, and your current confidence in it. The high-impact, low-confidence assumptions are your Critical Assumptions: the handful worth testing first, because that’s where reducing uncertainty changes the decision the most.

Confidence is a function of evidence. BRI tracks the support behind each assumption on a ladder that runs from a rough guess to hard fact:

• Low Evidence — a rough first guess with no real support behind it yet.
• Some Evidence — a judgment grounded in experience or analogy, but not yet in data.
• Medium — extrapolated from your own or comparable past performance or strong proxy data.
• High Evidence — primary research aimed at the specific assumption: surveys, pricing tests, pilots.
• Very High Evidence— observed results from the market itself, the strongest evidence there is.

As evidence climbs the ladder, uncertainty ranges narrow and the evaluation sharpens. Investment scales with demonstrated evidence — not the calendar — so you commit more only as the bet becomes less risky.

How Growth Forge® Software turns this into decisions

Most teams attempt this across a fragmented patchwork of PowerPoint templates and Excel models — where ranges, assumptions, and evidence live in separate files and rarely connect. Growth Forge® Software builds the whole discipline into one platform:

• Strategy Hypothesis Model — define your strategy across six dimensions, with every element typed as a Choice, Assertion, or Uncertainty and each uncertain input expressed as a range.
• Assumptions Manager — organizes and prioritizes assumptions by uncertainty and impact, surfaces your Critical Assumptions, and flags inconsistencies between them.
• Evidence Gathering Manager — defines and tracks the experiments and research that support or refute each key assumption, moving it up the evidence ladder.
• Strategy Evaluation — evaluates the full hypothesis against stage-specific criteria, organized by desirability, feasibility, and viability, so decisions rest on evidence rather than presentation quality.
• Evidence-based investment — ranged inputs flow into the financial model and simulation, so you see a distribution of outcomes rather than a single forecast, and scale investment as uncertainty falls.

What’s the difference between an assumption and an uncertainty?

Every uncertainty is an assumption, but not every assumption is quantitative. BRI splits assumptions into Strategy Choices (within your control), Assertions (about what others will do), and Uncertainties (quantitative assumptions expressed as ranges). Naming which is which is what makes evidence-gathering disciplined.

Why express inputs as ranges instead of best estimates?

A single best estimate hides how wrong it could be. A range captures the spread of plausible values, lets you see which inputs actually move the outcome, and gives evidence something concrete to narrow over time.

What do the 10% and 90% points on the range mean?

They are the 10th and 90th percentiles of the value’s probable distribution. There’s only about a 1-in-10 chance the true value lands below the min, and about a 1-in-10 chance it lands above the max — so roughly 80% of the probable outcomes fall between them.

How do you decide which assumptions to test first?

Score each assumption on impact and confidence. The high-impact, low-confidence ones — your Critical Assumptions — come first, because that’s where reducing uncertainty changes the decision the most.

How does Growth Forge® Software model uncertainty?

Every element of the Strategy Hypothesis Model is typed as a Choice, Assertion, or Uncertainty, with uncertain inputs entered as ranges. The Assumptions Manager prioritizes them by uncertainty and impact, the Evidence Gathering Manager tracks the support behind each one, and the ranged inputs flow into the financial model and simulation so you evaluate a distribution of outcomes rather than a single forecast.