How Incertive Works Under the Hood

Step 1: How Incertive Interprets Your Plan

When you describe your plan in plain language, Incertive performs a structured analysis of the text. It identifies the key components: what you are trying to achieve (objectives), what resources you are committing (budget, time, people), what you expect to happen (assumptions), and what needs to go right (dependencies).

The platform extracts quantitative information - cost estimates, timelines, revenue targets, headcount - and qualitative information - market assumptions, competitive positioning, technical approach. Both types of information feed into the uncertainty identification process. More detailed plan descriptions produce more precise analyses, but the platform is designed to work with imperfect and incomplete information because real-world planning always involves gaps.

This is important: Incertive interprets your plan as stated. It does not verify factual claims or validate market assumptions against external data. If you state that your market size is $10 billion, the platform works with that figure (while flagging market size as an uncertainty). The accuracy of the analysis depends on the accuracy and completeness of your description. See the platform overview for the full user workflow.

Step 2: How Uncertainties Are Identified

After interpreting your plan, Incertive identifies the variables that carry significant uncertainty. It examines every assumption - stated or implied - and evaluates whether it is a fixed fact or an uncertain estimate. A signed contract with a fixed price is a fact. An estimated customer acquisition cost based on industry benchmarks is an uncertainty.

The platform organizes uncertainties into six domains: market, technical, operational, financial, regulatory, and team. This framework ensures comprehensive coverage. For each uncertainty, the platform estimates a probability distribution - the range of possible values and their relative likelihoods. These distributions are based on the type of variable, the industry context, and patterns observed across similar plans.

Critically, the platform also performs gap analysis - it checks whether important categories of risk are unaddressed. If your plan involves entering a regulated industry but does not mention regulatory requirements, the system flags this gap. This is designed to catch the risks that teams miss not because they are unlikely but because they are outside the team's area of focus.

Step 3: How the Simulation Runs

With the uncertainties identified and quantified, Incertive runs a Monte Carlo simulation. In plain language, this means the platform creates a model of your plan and then runs it thousands of times, each time with different values for the uncertain variables.

In each iteration, the simulation randomly samples a value for every uncertainty from its probability distribution. One iteration might combine an optimistic customer adoption rate with a pessimistic development timeline. Another might combine moderate assumptions across the board. Yet another might test the scenario where three things go wrong simultaneously. The simulation does not just test individual risks - it tests every possible combination.

Typically, 10,000 iterations are run. This is enough to produce statistically stable results - running more iterations would not materially change the output. Each iteration produces a complete outcome: did the plan succeed? What was the cost? What was the revenue? How long did it take? Aggregating these 10,000 outcomes produces the probability distribution, success probability, and sensitivity data that form the analysis results.

It is worth noting what Monte Carlo simulation is and is not. It is a well-established computational technique first developed by Stanislaw Ulam and John von Neumann during the Manhattan Project and used across engineering, finance, science, and defense for decades. It is not a prediction engine. It does not tell you what will happen. It tells you the range of what could happen and how likely each outcome is, given the uncertainty ranges in your inputs.

Step 4: What the Outputs Mean

Every analysis produces several outputs, each designed to answer a specific question. The success probability is the percentage of iterations in which the plan achieved its stated goals. This is the headline number, but it is most useful in context - alongside the sensitivity analysis, the probability distribution, and the plan variants.

The probability distribution and S-curve show the full range of possible outcomes - not just whether the plan succeeds or fails, but the range of costs, timelines, and revenues you might experience. The tornado diagram shows which variables have the most influence on the outcome, helping you prioritize your risk management efforts. The plan variants offer concrete alternative approaches, each with their own probability of success.

These outputs should be read together, not in isolation. A 65% success probability might be acceptable if the tornado diagram shows that the primary risk is addressable, or unacceptable if the distribution shows severe tail risk. The outputs are designed to give you a complete, multi-dimensional view of your plan's risk profile - not a single number that hides the complexity.

Limitations and What This Tool Is Not

Transparency about limitations is essential for trust. Incertive is a powerful analytical tool, but it has boundaries that you should understand before relying on it for important decisions.

It is decision support, not a decision. Incertive provides probability-based analysis to inform your judgment. It does not replace the need for domain expertise, strategic thinking, and contextual understanding that only you have. A 73% probability is an input to your decision, not the decision itself.

It cannot account for truly novel risks. The uncertainty identification process draws on patterns from business planning. Black swan events - entirely unprecedented risks with no historical parallel - may not be captured. The probability distributions are based on observable patterns, not crystal balls.

Outputs are only as good as inputs. If critical information is omitted from the plan description, the analysis will be incomplete. If assumptions are stated as facts ("our CAC is $35" when it is actually uncertain), the platform will not treat them as uncertainties unless you flag them. Garbage in, garbage out - this applies to all analytical tools.

Probabilities are estimates, not measurements. A 73% probability is not measured with the precision of a thermometer reading. It is the output of a model, and all models are simplifications of reality. The probability should be interpreted as "approximately in this range" rather than "exactly this number." The value is in the relative comparison (73% vs. 52%) and the order of magnitude, not in the last decimal place.

It does not guarantee outcomes. A plan with a 90% probability of success will fail about 10% of the time. A plan with a 30% probability will succeed about 30% of the time. The probability tells you the odds, not the outcome. Over many decisions, accurate probabilities lead to better overall results - but any individual decision can go against the probability. Learn more about our approach to data protection on the security page.

How to Interpret Probabilities Correctly

Think of probabilities like weather forecasts. When the forecast says 70% chance of rain, it does not mean it will definitely rain. It means that in similar atmospheric conditions, it rains about 70% of the time. You might carry an umbrella, but you would not cancel your wedding. Business probabilities work the same way - they inform preparation, not paralysis.

The right threshold for proceeding depends on the stakes. For a low-cost, easily reversible decision, even a 40% probability might be worth pursuing - the upside could justify the risk. For a high-stakes, irreversible decision - like a major capital investment - you might want 80% or higher. There is no universal threshold. The probability gives you the information; your judgment determines what to do with it.

The most common mistake is treating probabilities as binary. People hear "73%" and think "go." They hear "45%" and think "no-go." But 73% means roughly three successes for every one failure. Whether that ratio is acceptable depends on the cost of failure, the value of success, and whether you have alternatives with better odds. Always consider the probability in the context of the full analysis - the distribution, the sensitivity, and the plan variants.

How Calibration Improves Over Time

One of the most important features of Incertive's methodology is its ability to improve through use. Calibration tracking compares predicted probabilities against actual outcomes across all users and all plan types. When the system predicts a 70% success rate and the actual rate is 60%, that discrepancy feeds back into the models to improve future predictions.

At the individual and organizational level, calibration tracking helps you understand your own biases. You might discover that you are well-calibrated on financial estimates but consistently overconfident on timelines. This specific feedback is far more useful than a general admonition to "be more realistic" - it tells you exactly where to adjust.

This creates a virtuous cycle: better calibration leads to more accurate probabilities, which leads to better decisions, which leads to better outcomes. It does not happen overnight - calibration requires dozens of tracked decisions to become statistically meaningful. But for organizations that commit to the process, the improvement in decision quality compounds over time. This is the methodology behind success probability and the broader Monte Carlo simulation approach that powers Incertive.

Frequently Asked Questions

See It in Action

The best way to understand the methodology is to use it. Describe a plan, see the analysis, and evaluate the results for yourself.