A Lodestake report is built to be checked, not admired. Read it in the order below and you will know in ten minutes where the idea is solid and where it is soft.
Start with Unknowns
The unknowns section lists what the pipeline could not establish: a market size the sources disagree on, missing pricing evidence, a competitor nobody checked. In our benchmark a typical report listed 7. Each one is a test you could run. If an unknown sits under a belief your idea depends on, that is your first job.
Then the verdict columns
Reasons to build and reasons not to, side by side. Each reason should carry an evidence ID. A reason with no citation is an argument rather than a finding; it may be right, but weigh it as an argument.
Kept, removed and unknown
- Kept sentences state numbers that appear in measured evidence they cite. The citation becomes a link. Kept does not mean true; it means traceable.
- Removed sentences stated a number no cited source carried. They are taken out and counted, so you can see where the draft got ahead of its sources.
- Unknown marks data that could not be found or matched, such as a public dataset with no category for your market. Unknown is never shown as zero.
Scores are summaries with reasons
Each score (problem, opportunity, feasibility, why now) carries a sentence explaining it. Read the sentence, not just the number. A low “why now” with no evidence means no forcing event was found, not that none exists.
Open the sources that matter
The evidence ledger lists every item with its source and link. You do not need to open all of them; in our benchmark a typical report cited 78. Open the ones behind the beliefs your decision rests on.
Turn it into next week
Pick the cheapest unknown under your most important belief and test it directly. Then update the plan. A report is a snapshot; your evidence keeps growing.