GradeCircle
GradeCircle
Pricing
Model a Decision

What decision is waiting on better evidence?

Answer five questions and see the scope, price band and timeline for modelling it — before you give us your details.

Scope my decisionSee a sample decision room

Monthly briefing

One email a month on decision modelling. No noise.

By subscribing you agree to our privacy policy.

GradeCircle
GradeCircle

Decision models and AI implementation, from a team that ships its own products.

Solutions

  • Free Tools
  • Sample Decision Room
  • Methodology
  • Applications
  • Strategic Services
  • Consulting
  • Pricing
  • Portfolio
  • Investors

Applications

  • Healthcare Solutions
  • Legal Services Platform
  • Financial Applications
  • Government Solutions
  • Tech Solutions
  • View All Applications →

Company

  • About
  • Blog
  • Changelog
  • Industries
  • Careers
  • Press
  • Contact

Legal & Resources

  • The App
  • FAQ
  • Help & Support
  • Privacy Policy
  • Terms of Service
  • Cookie Policy
  • Security

© 2026 GradeCircle. All rights reserved.

🔒 256-bit TLS·SOC 2 Aligned·NDA Protected

Methodology

Every number should survive a challenge.

A model is only useful if the people it has to convince can take it apart. This is how ours are built, what we check before handing one over, and — just as importantly — what we do not yet claim.

How the calculation works

Calculations are deterministic. Every figure is produced by a formula evaluator that accepts arithmetic and a fixed whitelist of maths functions, applied to named inputs. The same inputs always produce the same outputs. A language model never computes one of your numbers — it can help draft an explanation of a result, but it cannot create the result.

Free calculator results are recomputed on our servers. The figures shown in the browser are recalculated independently before we send you a breakdown, so what you receive does not depend on anything editable in the page.

Sensitivity is measured, not asserted. We perturb every input across its plausible range — several hundred samples — and rank inputs by how strongly each moves the headline result. That ranking is what tells you which assumption to argue about first.

You can check all of this. The public sample decision room runs the same engines, shows its assumption ledger, and lets you move the deciding assumption until the recommendation flips.

The model validation checklist

Before a model is delivered, an analyst who did not build it reviews it against this list. It is published so you can hold us to it — and so you can run the same checks against a model built by anyone else.

Structure

  • The model answers the decision as written, not a nearby question.
  • Every alternative under consideration is represented, including doing nothing.
  • Each output traces to a formula built from named inputs — no constants standing in for logic.
  • Units are consistent, and every rate states its period.

Assumptions

  • Every material assumption has a value, a source, an owner and a date.
  • Observed facts, customer assumptions and external estimates are labelled distinctly.
  • Any figure the model itself produced is marked as an output, not an input.
  • Assumptions with no evidence behind them are flagged, not quietly averaged.

Scenarios and sensitivity

  • Base, upside and downside all run and produce different answers.
  • Sensitivity analysis names the assumption that most changes the result.
  • The break-even or flip point is stated where one exists.
  • Scenarios that would reverse the recommendation are identified explicitly.

Honesty

  • Exclusions are written down — tax, working capital, discounting, terminal value where they do not apply.
  • No precision is implied beyond what the inputs support.
  • “Do not proceed” is available as an outcome and stated when the numbers support it.
  • A reviewer other than the model's author has run the checks above and signed off.

What we do not claim yet

A methodology page that overstates the method is worse than none. These are on the roadmap and are not shipped:

  • Assumption provenance inside the product. Sources, owners and dates are delivered as a written assumption sheet today, not as tracked fields you can query or version.
  • Validation-state badges on every model surface. Review status is recorded in the engagement, not displayed as a state machine in the app.
  • Outcome tracking. Comparing what actually happened against what the model predicted is done manually at the 30-day review; there is no automated variance dashboard.
  • Continuous monitoring and threshold alerts. This is the Decision Room subscription, which is not yet available for sale.

Your data

An NDA is signed before any work begins. Your data is not used to train models and is not shared across engagements. Controls and the current subprocessor list are on the security page. You own every deliverable outright.

Have a decision that needs this?

Five questions and you will see the scope, price band and timeline — before giving us your details. Decision Model Sprint from $8K–$20K.

Scope my decisionSee a worked example