Most consequential decisions inside organisations are still made on point-estimate spreadsheets — a single base-case NPV, a single forecast, a single answer. Reality is uncertain. The single number hides the downside that could sink the business, and it can't answer the questions that actually matter: what if the price falls 20%? How sure are we? Is it worth paying for a study before we commit?
ModelChoice replaces the point estimate with a structured decision tree, automatic backward-induction, and a full library of analyses — sensitivity, risk profile, robustness, value of information, multi-criteria scoring, Monte Carlo — that turn a single Excel workbook into a complete decision-analysis workbench.
You build the decision tree directly on a worksheet. A starter node appears with one click, and you add decision options, chance branches, probabilities and outcome values by editing real Excel cells. Every value is a real cell — link it to a financial model, an assumptions sheet, an external data source, anywhere in the workbook.
As you build, ModelChoice continuously solves the tree using backward induction (rollback). The optimal strategy is highlighted; the expected value of every node updates in real time as you adjust inputs. There is no "run" button, no recalc step — it works the way Excel formulas work.
When the tree is ready, the Analysis ribbon group gives you sixteen distinct analytical reports from a single click. Each one writes a fresh report sheet — charts, tables, summary statistics, written narrative — ready to copy into a board paper. Inputs change? Regenerate any report in seconds.
Book a live online demo and we'll walk you through building a decision tree, running the analyses, and generating a board-ready Decision Brief — focused on your questions.
ModelChoice is a real Microsoft Excel add-in — not a separate application, not a parallel file format. Trees live next to the data that feeds them, in the same workbook, under the same version control. Trees serialise cleanly to JSON. Models from PrecisionTree import with one click.
A single tree drives sixteen different analyses. Sensitivity tornadoes. Full outcome distributions. Robustness stress-testing with flip-distance scoring. Force-to-Outcome reverse goal-seek. EVPI and EVII. MCDA with AHP. Monte Carlo via ModelRisk. Every angle on the same model, no extra modelling work.
A decision tree in ModelChoice is the same thing you'd sketch on a whiteboard — only drawn for you, kept aligned, and re-solved every time you change an assumption. Most analysts are productive in their first hour. Sixty industry example models ship with the product, ready to open and adapt.
Beyond expected value: utility-function rollback with certainty equivalents and risk premiums; weighted multi-criteria scoring across financial and non-financial criteria; AHP weight elicitation with consistency-ratio checking; continuous-distribution Monte Carlo through ModelRisk. The analyses you'd otherwise need three tools for, in one add-in.
The full product, every analysis, every example model, no feature restrictions.
Twenty-four highlights from the product — the complete decision-analysis toolkit, native to Microsoft Excel.
Complete decision-analysis toolkit, native to Microsoft Excel
Build decision trees directly on the worksheet — no external editor
Continuous rollback: optimal strategy updates as you type
Risk Profile: the full distribution of outcomes, not just the average
Cumulative Risk Profile with percentile statistics per strategy
One-way sensitivity with tornado and spider diagrams
Two-way sensitivity with strategy-region charts
Break the Decision — robustness scoring 0–100 with flip distances
Probabilistic-reversal Monte Carlo on the optimal strategy
Force to Outcome — reverse goal-seek to flip the optimal choice
Policy Suggestion: every strategy ranked by EV with optimality gaps
Strategy Table: full enumeration with audit-ready decision paths
EVPI — the maximum value of resolving uncertainty completely
EVII — the value of a real (imperfect) study before you commit
Multi-Criteria Decision Analysis: financial + non-financial criteria
AHP pairwise weight elicitation with consistency-ratio validation
Utility-function rollback: exponential, logarithmic, square-root, linear
Certainty equivalents and risk premiums for risk-averse stakeholders
Continuous-distribution Monte Carlo via ModelRisk integration
One-click import from PrecisionTree
JSON tree export and import for version control and team sharing
Single-page Decision Brief — board-ready executive summary
UI and documentation in seven languages
Sixty industry example models — energy, pharma, finance, MCDA, more
Seven outcome-focused scenarios — the questions a decision-maker actually asks, and the analyses that answer them.
The classical oil-and-gas decision. Build the tree once: cost of drilling, dry-hole probability, payoff sizes, survey accuracy. Rollback gives the optimal strategy. EVII tells you whether the survey is worth its cost. Sensitivity ranks every input by impact. The robustness report tells you exactly how much each assumption has to move before the answer changes.
Decision trees are the standard language of pharma R&D portfolio management. Encode Phase II go/no-go choices, success probabilities and payoffs in a single tree; rollback prioritises the optimal pipeline. EVPI caps what additional research can be worth. Monte Carlo through ModelRisk handles continuous uncertainty on cost and timing.
Multi-criteria decisions where price is only one factor. ModelChoice's MCDA mode blends financial payoff with weighted non-financial criteria (quality, delivery reliability, ESG score, technology fit). AHP weight elicitation derives defensible weights from a stakeholder workshop with built-in consistency checking.
A board paper that says "the deal is worth £40M" invites the question "how sure are you?" The Break the Decision report quantifies the answer: for every input, the precise threshold at which the optimal choice changes, and a 0–100 robustness score. Probabilistic-reversal Monte Carlo gives the probability that the recommendation is wrong given joint uncertainty.
A staged-capacity decision is a perfect fit for a tree with multiple decision points and continuous chance variables. Couple ModelChoice with ModelRisk to model demand as a continuous distribution; the simulation runs thousands of paths and ranks the staged-expansion strategies by full distribution of outcomes, not just expected NPV.
Pilots, market studies, technical trials all have a cost. Their value depends on how often they would change the decision. EVPI gives you the ceiling — the most the study could ever be worth. EVII gives you the realistic value given the study's known accuracy. Together they answer the go / no-go on commissioning the work — in hours, not weeks.
The Decision Brief report is a single-page executive summary: rollback result, sensitivity headlines, risk profile, robustness score, value-of-information findings, all formatted to print or paste straight into a steering-committee deck. Generated in seconds; rerun in seconds when an input changes.
The maximum price for ModelChoice is a very competitive € 950 per user per year. Volume discounts and multi-product bundles (ModelChoice + ModelRisk + Tamara) are available — calculate your own quote below.