Solve Problems Nobody Else Noticed

Proof required

Submit your problem research file: notes from at least 3 real conversations (bullet points per conversation with specific things said, not paraphrases) and a written assumption inventory (at least 8 assumptions, each labelled evidence-backed or hypothesis-only, with one sentence of justification). The problem statement must be a real problem you intend to work on, not a classroom exercise.

What gets checked

• 3 real conversations documented with specific content — 'spoke to 3 people who agreed with me' is not research; the notes must capture what the people said, including things that surprised you
• Assumption inventory contains at least 8 assumptions with evidence/hypothesis labels — assumptions that are all 'evidence-backed' suggest the inventory was not honest about the hypothesis-only ones
• At least 3 of the assumptions are labelled hypothesis-only with an honest acknowledgement of what evidence would be needed to test them

Common mistakes

• Conducting 3 conversations only with people who already share your view of the problem — research that only confirms prior beliefs is not research
• An assumption inventory that describes the solution space rather than the problem space — assumptions about what users want are not the same as assumptions about what the problem is
• Skipping the evidence/hypothesis labelling and treating all assumptions as facts — the labelling is the core practice of honest problem framing

Resources

What a verifier looks for

• Ask the submitter to describe one thing a research conversation revealed that surprised them or challenged a prior assumption — genuine research always surfaces something unexpected
• Check the assumption inventory for 'hypothesis-only' labels — an inventory where all assumptions are 'evidence-backed' has not been applied honestly
• Confirm the problem is real and the research conversations are with real people who experience it — ask the submitter to describe one of the people they spoke to
• Ask what they would do differently in the next research conversation based on what they learned — this tests whether the research produced genuine updates

Proof required

Submit your 3 solution concepts (one paragraph each, making the core idea and its logic clear) plus test documentation for each: what the test was, who you tested with, what they said or did, and which of your M1 assumptions was most affected by the test result.

What gets checked

• 3 concepts are meaningfully different — not variations on the same mechanism; a reviewer should be able to describe the fundamental difference between them in one sentence
• Each concept was tested with at least 1 real person who experiences the problem — testing with the people who designed it does not qualify
• Each test documentation references a specific M1 assumption — the test was designed to answer a specific question, not to get general feedback

Common mistakes

• Generating 3 concepts that are all variations on the same core idea (e.g. 3 different apps rather than one app, one physical intervention, and one behaviour change)
• Tests that ask users to evaluate the idea rather than to interact with it — 'do you like this idea?' is a much weaker test than 'here is a prototype — try to use it to solve your problem'
• Not connecting test findings back to the M1 assumption inventory — without this connection, the test is feedback-gathering, not assumption-testing

Resources

What a verifier looks for

• Ask the submitter to describe the most important thing they learned from the 3 tests — genuine testing always produces a surprise or a significant update to prior beliefs
• Check that the 3 concepts are genuinely different — ask the submitter to name the fundamental mechanism behind each one in one sentence; if the mechanism is the same, the concepts are not meaningfully distinct
• Confirm tests were done with people who experience the problem, not with colleagues who evaluated the idea abstractly — ask who specifically was tested and what their relationship to the problem is
• Ask which M1 assumption was most overturned by the testing and what the submitter now believes instead — this is the core evidence of innovative thinking

Proof required

Submit your pitch document (slides or structured written brief, minimum 600 words) covering problem evidence, 3 concepts, selection rationale, and test findings, plus the reviewer's written assessment: their challenge questions, the riskiest untested assumption they identified, and their sign-off confirming their role and that this was a live session.

What gets checked

• Pitch document covers all 4 sections — problem evidence, 3 concepts, selection rationale, test findings — not just the favoured concept
• Reviewer identifies at least one riskiest untested assumption that was NOT in the submitter's own assumption inventory from M1
• Reviewer sign-off confirms innovation evaluation or product development background: investor, intrapreneur, product director, or founder with evidence of real products brought to market

Common mistakes

• Presenting only the favoured concept and glossing over the ones that failed testing — the M2 testing failure data is the most credible evidence of rigour
• A reviewer who focuses on the idea's viability rather than the submitter's process — the process (framing, divergence, testing) is what this outcome is developing, not the idea itself
• Pitch document that describes what the tests showed without naming what specific assumption was addressed by each test

Resources

What a verifier looks for

• Ask the submitter what they would do next — a genuine pitch session produces a clear next action, not just a feedback summary
• Identify the riskiest untested assumption yourself before reading the submitter's notes — then compare with the reviewer's identification to see if they agree
• Confirm the reviewer's background involves evaluating or building real innovations — not just being knowledgeable about the domain
• Ask the submitter which of their M1 assumptions was most wrong by the end of M2 — genuine innovation work always updates prior beliefs significantly