How the evaluation works

The backbone is Making Strategies Work (MSW), the strategy-implementation method taught in the Edinburgh Business School MBA: a mission, an honest scan of what helps and what blocks you, a strategy that answers those barriers, the few factors that must go right, and the activities that deliver them — each measured, each connected by an explicit cause-and-effect chain. Where MSW doesn't reach, mainly early validation and unit economics, we add well-established tools.

The frameworks, in plain language

• Jobs-to-be-done and the value proposition canvas — what the customer is trying to get done, what hurts, and whether your offer relieves those specific pains.
• Lean canvas and riskiest assumptions — the whole business on one page, and the beliefs most likely to kill it, ranked by impact and evidence.
• Bottom-up market sizing — customers × price × frequency, checked against a top-down figure.
• Porter's five forces, PESTEL, and MSW barriers and enablers — whether the industry allows profit, and what specifically helps or blocks you, each with an indicator to watch.
• VRIO, positioning, trade-offs and activity fit — whether your advantage is valuable, rare, hard to copy and organised to exploit, and what you will deliberately not do.
• Unit economics — lifetime value, fully-loaded acquisition cost, payback, break-even and three scenarios.
• Critical success factors and critical activities — each starts with “We must…”, passes a necessary-and-sufficient test, and has leading and lagging indicators.
• OKRs, RACIS and a control dashboard — who is responsible, who supports, and which numbers trigger a change of course.
• Pre-mortem and causal-chain risk — imagine it failed, explain why, then test every link for failure to deliver and unintended effects.

Evidence tags

Every material number is tagged S (sourced from research we performed), U (you told us), E (our estimate, with the derivation) or A (an assumption to test). We never invent statistics, company names or links; unknowns stay unknown.

The score and the verdict

Eight dimensions are scored 1–5 and weighted 15/10/10/15/15/10/15/10. Points per dimension = (score − 1) ÷ 4 × weight, so a uniformly “plausible but unproven” idea lands at 50. The total maps to a verdict: 75 and above Go, 60–74 Go with conditions, 45–59 Validate first, 30–44 Pivot, below 30 No-go.

Three rules override the score: a fatal flaw forces Pivot or No-go; low evidence confidence caps the verdict at Validate first; and an idea that fails the first gate (no painful problem for a reachable customer) can do no better than Pivot. The score is recomputed in code from these rules — the language model's arithmetic is never trusted.

What it is not

It is not legal, tax or investment advice, and it does not guarantee success. It is a disciplined first pass that tells you what to test next.