Why Definitions Matter
Two companies can report the same metric name with completely different calculation methods. We track this per company, per year — and flag when it changes.
The Problem
AISC is not one number
All-In Sustaining Cost (AISC) is the most widely tracked cost metric in gold mining. The World Gold Council published guidance defining it in 2013 and updated it in 2018. But “guidance” is not a standard. Companies interpret it differently, and the same acronym can mean fundamentally different things depending on who is reporting it.
The same three letters — AISC — reported by different companies, can mean:
A company that nets out significant by-product credits will show a much lower AISC than a pure gold producer. Comparing them directly without knowing this is misleading. A company that excludes corporate G&A will appear more efficient than one that includes it. The numbers look comparable. They are not.
How We Track It
definition_name per company, per year
Every datapoint has a definition_name field. For cost KPIs, this is a short label describing how the company calculates the metric. It is checked for every year, not assumed to be constant.
| KPI | definition_name examples | Why it matters |
|---|---|---|
| AISC (gold) | "AISC net of by-product credits" / "AISC excl corporate G&A" | By-product credits can reduce AISC by $100-300/oz |
| C1 cash cost (copper) | "C1 net of by-product credits" / "C1 including TC/RC" | Treatment and refining charges add $0.15-0.30/lb |
| Iron ore produced | "wet basis" / "dry basis" | Wet tonnes are 5-10% higher than dry tonnes for the same ore |
| Uranium cost | "includes purchased uranium" / "mine-site only" | Companies that buy and process ore have different cost structures |
| PGM production | "4E basis" / "6E basis" | 4E = Pt+Pd+Rh+Au; 6E adds Ir+Ru. 6E totals are always higher |
Break Detection
What happens when definitions change
When the definition_name changes between consecutive years for the same company and KPI, the website shows a break. A break means: the year-over-year percentage change may be unreliable because the underlying calculation methodology changed, not just the actual operational performance.
For example: if a gold miner reported AISC of $1,200/oz in 2022 using “AISC net of by-product credits” and $1,300/oz in 2023 using “AISC excl corporate G&A,” the +8.3% year-over-year change is misleading. The cost might have actually gone down if measured on the same basis. But we cannot know that because the definitions differ.
Break Types
Hard breaks vs soft breaks
Hard break
The comparison is invalid
A hard break occurs when the change in methodology is so fundamental that the year-over-year comparison is meaningless. Examples: switching from attributable to consolidated reporting basis (AngloGold Ashanti in 2022), or changing from wet tonnes to dry tonnes for iron ore measurement. The same physical operation produces a completely different number.
Soft break
The comparison is imprecise but directional
A soft break occurs when the methodology change affects the magnitude but not the direction of the comparison. Examples: switching from “AISC net of by-product credits” to “AISC excl corporate G&A.” The absolute number shifts, but a large year-over-year change is still directionally meaningful.
The workspace shows both types inline. Breaks are not hidden — they appear next to the affected values so users can judge for themselves whether the comparison is meaningful.
False Trend Prevention
Why this prevents bad analysis
Without definition tracking, a data consumer might conclude:
Without definition tracking
“Company X's AISC increased 15% from 2022 to 2023 — a major cost blow-out.”
With definition tracking
“Company X changed from AISC net of by-product credits to AISC including corporate G&A in 2023. The definition change accounts for most of the apparent increase. On a like-for-like basis, costs were approximately flat.”
This is not a theoretical problem. It happens regularly in mining data. Companies change their cost reporting methodology, sometimes without even highlighting it in the annual report. The only way to catch it is to compare the definition text year over year, which is exactly what ProveMines does.
Backfilling
Consistent definitions across the full history
If a company used the same AISC methodology for all 6 years (2020–2025), we apply the same definition_name to all 6 years. Not just the year we first found it documented.
This matters because the workspace detects a break whenever definition_name changes — including when it changes from null to something. If years 2020–2023 have no definition and 2024 does, the workspace would show a spurious break in 2024. Backfilling eliminates that.
Currently 21% of records have definition_name set (669 of 3,220). The majority of nulls are correct — production volumes do not need methodology definitions. A number like “350,000 ounces of gold produced” means the same thing regardless of how you calculate it. Cost metrics are where definitions matter, and these are progressively being backfilled.
Beyond AISC
Definitions matter across many KPIs
AISC is the most visible example, but definition ambiguity exists across the mining data landscape:
Iron ore production
Can be measured on a wet basis or dry basis. Wet tonnes are 5-10% higher than dry tonnes for the same ore. Fortescue (FMG) historically reports in wet metric tonnes while most others report dry. This is not an error — it is a measurement methodology difference.
PGM production
Platinum group metals can be measured on a 4E basis (platinum, palladium, rhodium, gold) or 6E basis (adding iridium and ruthenium). A company's 6E total will always be higher than its 4E total for the same operation.
Lithium production
Can be measured in tonnes of lithium carbonate equivalent (LCE), tonnes of spodumene concentrate (SC), or tonnes of lithium hydroxide. These are fundamentally different products with different conversion factors.
Realized price
Some companies report realized price gross of treatment and refining charges (TC/RC), others report net. For copper, TC/RC can account for $0.15-0.30/lb difference.
See definitions in context
Open the workspace and click any cost metric to see how the company defines it — and whether that definition changed.
Open Workspace