Mining investment diligence: from data room to defensible decision
Mining investment diligence: from data room to defensible decision
A mining investment decision is not short of documents. It is short of coherence.
A typical opportunity arrives as a data room: a technical report, drill tables, resource estimates, maps, ownership records, permits, a financial model, management presentations, and internal notes. The evidence is all there. The clarity is not.
Someone still has to work out which claims hold up, which sources disagree, what the answer is really resting on, and what new evidence would change the decision. That work is often divided across files, spreadsheets, email, generic data rooms, and individual memory.
The structural problem
More documents do not automatically produce a better decision. Evidence becomes useful only when it is connected to claims, assumptions, reviews, decisions, and later revisions.
The current process has a recurring failure mode:
- Two reports disagree on tonnage, and nobody has flagged it.
- The financial model assumes 88 percent metallurgical recovery, but the lab data only supports 71 percent on an older test.
- The investor presentation calls the resource "high confidence," but much of it is classified as inferred.
- When someone in the investment committee asks where a number came from, the answer is not in the deck.
The result is a decision that may be correct but cannot be defended, traced, or revisited without rebuilding the record from scratch.
A better workflow
A credible mining diligence process connects evidence to claims at every step:
- Bring everything in. Every source is kept in one place with its origin and version intact. Nothing gets lost or quietly changed.
- Tie each claim to its evidence. Every material claim (grade, tonnage, recovery, ownership, access) links to the exact page and table it came from. The machine drafts the claims; people decide what is true.
- Check it the same way every time. The evidence goes through a fixed diligence framework before anyone decides. What is supported, what conflicts, and what is still missing are kept apart, not blended into one score.
- Make the call, on the record. A named person decides: reject, monitor, or advance. The decision is saved with the reasoning behind it.
- Keep watching. When new evidence arrives, you see what it affects and can reopen the decision without rebuilding the record.
A concrete example
Consider a copper project called Red Valley. The project owner claims a large resource with significant exploration upside and access to nearby infrastructure.
The technical report states an average copper grade of 0.42 percent, sourced from page 147, Table 14-8. That claim is supported and reviewed.
But the metallurgical recovery claim is disputed. Two recent laboratory tests reported recoveries between 86 and 89 percent. An older test reported 71 percent. No variability testing has been done across the full deposit. The financial model assumes 88 percent. The status is unresolved.
Three months later, new metallurgical results arrive showing average recovery of only 74 percent. That single piece of new evidence affects the recovery claim, the financial model, the metallurgy finding, the assumption about recovery staying above 82 percent, and the previous decision to advance.
The decision changes from "advance into limited diligence" to "reject at the current valuation." The old decision remains in history with its full reasoning trail. The new one is reasoned, not rebuilt from scratch.
The value is not a faster report
The real value is giving the organization a reliable memory of what it believed, why it believed it, who approved it, and what later caused the decision to change.
A report is one output. The durable asset is the structured, versioned relationship among opportunities, evidence, claims, assumptions, reviews, and decisions. That is what SubsurfaceOS is building.