What scheem examines in every model
A coherent story on a slide is only a starting point for validation. Customer behaviour, delivery capacity, unit costs, and real-world observations supply the proof. scheem keeps three questions distinct:
Does the logic fit?
Do the offer, customer segments, channels, delivery activities, resources, partners, and economics make sense together as one connected system?
What must be true?
Which critical assumptions and capacity constraints does the model depend on — and what breaks across the rest of the business if one of them fails?
What does evidence support?
What has been tested in the real world, what remains uncertain, and what single observation would change the next strategic decision?
Three principles of the engine
1. Structural & quantitative integrity
A business model is a system, not nine independent boxes. Changing your pricing affects your customer segments; scaling customer volume changes your required delivery hours. scheem checks logical consistency and explicit capacity arithmetic across all 9 blocks — when one part changes, the engine checks whether the rest still holds together.
2. Findings are not verdicts (Evidence over assumption)
AI suggestions remain additive proposals, and recognising a familiar business pattern does not establish that your particular business will work. Only the Validation Object Model may emit a model verdict (Holds, Holds conditionally, Does not yet hold, or Refuted) based on explicit deal-breaker assumptions, tests, and appraised evidence.
3. Atomic specificity
Vague statements like “various customer segments” or “multiple revenue streams” are untestable. scheem structures concise, specific items — typically 3–8 words each with an atomic identifier (vp01, cs01) — that can be individually validated or invalidated.
Creative intelligence paired with deterministic rules
When you run Check in scheem, the engine combines relational AI with explicit deterministic rules across your saved model and workshop data:
Contradiction & dependency detection
scheem identifies conflicting assumptions between blocks (such as enterprise buyer segments paired only with self-serve acquisition channels) and links every finding to the exact block items that produced it.
Deterministic operational & economic checks
Where structured inputs support it, scheem checks workload hours, resource capacity, and unit-economics constraints using repeatable calculations — clearly labelled as Calculated from your inputs.
Flywheel & causal loop mapping
scheem traces reinforcing loops across your business model — cycles where outputs from one block strengthen inputs to another — and highlights single-point fragilities where a loop could stall.
Publication claim gates
When you export a Strategic Dossier or Pitch Deck, scheem verifies that no document claims validation unless its deal-breaker assumptions are backed by recorded test outcomes and appraised evidence.
Established frameworks, each reading the same model
There are no separate worksheets to fill in. Every framework below reads your saved canvas and workshop data, so a change in one block reaches every reading that depends on it. Each row says where the framework lives in scheem.
Nine blocks of short, specific items. It is the base every other reading on this page works from.
Sets pain relievers and gain creators against each segment’s pains and gains, then assesses the fit.
Describes each segment by the jobs it needs done, plus its pains and gains. AI suggestions stay proposals until you apply them.
Scores each resource on value, rarity, imitability and organisation, so you can see which ones could carry an advantage.
Derives strengths, weaknesses, opportunities and threats from your blocks, alongside rivalry, buyer and supplier power, substitutes and new entrants.
Places the model’s risks by probability and impact, so the few that matter stand out.
Follows a customer from awareness onwards through your channels and relationships, then turns that into a launch plan.
Sizes the market for the segments you have defined, from the total down to what you could realistically serve.
Works from your revenue streams and cost structure, calculating what the inputs allow before AI adds a narrative and names the gaps. The growth model simulates the flywheel of the pattern your model most resembles.
Turns an assumption into a hypothesis, a test, a metric and a pass criterion you commit to before you run the test.
Traces how one block drives another and where a reinforcing loop depends on a single link.
Simulates twelve months of your numbers thousands of times under lean, base and premium variants, then compares survival and margin across them.
A framework gives you a reading, not a verdict. None of them can say whether your model holds. Only validation can, from your tests and appraised evidence.
Build. Test. Validate. Scale.
scheem is designed around a continuous strategic loop: Build → Test → Validate → Scale. For step-by-step product instructions — from initial canvas intake to board-ready dossier export — see the User Guide, or read our deep-dive articles on the Editorial Blog.