The Mine That Knew Too Much and Decided Too Little
How PeopleSoft ABM and Metify Predictive Planning Forced Fresnillo's Operations Team to Make the Hardest Call in Underground Mining — and Why I Was in the Room
By Pedro San Martín — Asher & Company
I've been in a lot of boardrooms. But very few conversations have stayed with me the way a specific afternoon in Zacatecas did.
It was 2003. I was part of the PwC team deployed to Industrias Peñoles — one of Mexico's largest mining conglomerates and the majority shareholder of what would eventually become Fresnillo plc, today the world's largest primary silver producer. We were implementing PeopleSoft Activity-Based Management across the group's mining operations, integrated with Metify — a predictive planning engine that, years later, would become the intellectual and methodological foundation of what I now do at Asher & Company.
The room had a CFO, a COO, three mine managers, and a whiteboard covered in numbers that didn't reconcile. Silver spot was under pressure. The peso was moving. And Mina Proaño — the flagship underground silver mine in Fresnillo, Zacatecas, operating since 1554, one of the deepest and most complex polymetallic mines in the world — was, for the first time in years, mining ore that cost more to extract than it was worth.
The question on the whiteboard was not a finance question. It was not a tax question. It was not a strategy question. It was a physics question dressed up in accounting clothes: given what silver is doing to us today, which parts of this mountain are worth digging, and which parts should we leave alone until the market recovers?
PeopleSoft ABM gave us the cost anatomy. Metify gave us the foresight. Together, they gave us something the operations team had never had before: a causal map of profitability, zone by zone, drift by drift, tonne by tonne.
What we found was uncomfortable. What the team decided to do about it was remarkable.
The Battlefield Context
To understand why this case matters, you need to understand how underground silver mining actually works — and why conventional accounting is structurally inadequate for the decisions mine operators make every single day.
Mina Proaño had announced a cost per tonne of US$48 and a cash cost of US$5.6 per ounce in its earlier operational years. By the early 2000s, as the mine deepened and ore grades in certain zones declined, those numbers had moved materially. The mine was not one homogeneous operation. It was a portfolio of micro-operations: different veins, different depths, different ore grades, different extraction methods, different haulage distances, different equipment requirements. Each zone had its own economic fingerprint.
The problem was that Peñoles' cost reporting system didn't see any of that. It saw a single mine with a single cost per tonne. When that aggregate number started moving in the wrong direction, nobody could tell you why — whether it was a bad vein, a bad week, an equipment breakdown, a labor inefficiency, or simply the inevitable physics of mining deeper into lower-grade rock.
This is the fundamental intelligence failure in mining cost management: the aggregate always lies. It lies by averaging. It lies by concealing. It lies most dangerously when the average looks acceptable while individual zones are hemorrhaging value.
PwC was brought in to fix the intelligence failure. PeopleSoft ABM was the tool. Metify was the forward-looking layer that turned the cost map into a decision engine. And the silver price, which was doing something uncomfortable to the group's margins, was the forcing function that made everyone in that room willing to look at numbers they'd previously preferred not to see.
I. The Pressure Point
The silver price. Silver spot was trading in a range that put Fresnillo's consolidated all-in sustaining economics under serious stress. Fresnillo PLC's AISC was running at approximately $25.40 per ounce — a number that required careful zone-by-zone management to keep the operation viable at prevailing spot prices. The buffer between cost and market had compressed to the point where any operational inefficiency, any zone running above its economic threshold, translated directly into value destruction.
The cost per tonne. By 1H23, Fresnillo's cost per tonne had increased 5.1% to US$91.7, driven by underlying cost inflation — but the root of that trajectory was visible much earlier, in the deepening of the mine and the natural grade decline in certain historical extraction zones. The COO's problem was not that he didn't know costs were rising. It was that he didn't know where they were rising fastest, and he didn't have a tool that could tell him which zones to defend, which to defer, and which to abandon entirely at a given silver price.
The mine plan. Fresnillo's underground mine plan had been designed in a period of higher silver prices and more favorable ore grades. As both variables moved against the operation, the plan had not been updated to reflect the new economic reality. Resources — crews, equipment, explosives, haulage — were being allocated according to a plan built on assumptions that were, in some zones, off by 30–40%.
The CFO turned to me: "We need to know which parts of this mine are making us money and which parts are costing us money we don't have. And we need to know what happens to each zone if silver drops another dollar."
That question is what PeopleSoft ABM and Metify were built to answer.
II. The PACE Causal Model
III. The Data
Fresnillo Mine Operational Context (Public Data, Fresnillo PLC Reports)
| Metric | Value | Source |
|---|---|---|
| Mine location | Fresnillo, Zacatecas, México | Fresnillo PLC |
| Operating since | 1554 | Historical record |
| Owner/operator | Industrias Peñoles / Fresnillo PLC (75% Peñoles) | Fresnillo PLC |
| Attributable silver production FY23 | 56.3 moz (group total) | Fresnillo PLC AR 2023 |
| Average realised silver price 2023 | US$23.6/oz | Fresnillo PLC 1H23 report |
| Average realised silver price 2024 | US$28.8/oz | Fresnillo PLC FY24 prelim |
| Group AISC (silver) | US$25.4/oz | Industry benchmark |
| Cost per tonne — Fresnillo mine 1H23 | US$91.7/t | Fresnillo PLC 1H23 report |
| Cost per tonne YoY change 1H23 | +5.1% | Fresnillo PLC 1H23 report |
| Adjusted production costs 1H23 | US$773.9M (group) | Fresnillo PLC 1H23 report |
| Group gross profit 2023 | US$503.2M | Fresnillo PLC FY24 prelim |
| Group gross profit 2024 | US$1,246.3M | Fresnillo PLC FY24 prelim |
PeopleSoft ABM Activity Cost Breakdown: Underground Mining Operations
| Activity pool | Cost driver | Cost weight % | Notes |
|---|---|---|---|
| Drilling & blasting | Metres drilled per drift | 18% | Varies sharply by rock type |
| Loading & hauling | Tonnes moved · haulage distance | 24% | Distance penalty in deep zones |
| Ground support & timbering | Linear metres of development | 14% | Higher in structurally complex zones |
| Ventilation | Active metres of workings | 11% | Exponential with depth |
| Equipment maintenance | Equipment hours per zone | 9% | Age and utilization dependent |
| Personnel (mining crews) | Shifts per active zone | 16% | Fixed cost — hard to flex |
| Technical services | # of active fronts | 5% | Overhead distribution |
| Environmental & water management | Active zone count | 3% | Fixed per active working area |
| Total cost-to-mine | 100% |
Zone-Level Economic Profile Post-ABM Allocation
| Zone | Silver grade g/t | Fully loaded cost/t | Revenue/t at $23.6/oz | Economic margin/t | ABM verdict |
|---|---|---|---|---|---|
| San Carlos vein system | 218 | $78 | $165 | +$87 | Defend — increase allocation |
| Valdecañas vein extension | 196 | $84 | $148 | +$64 | Defend |
| Proaño deep zone (–600m) | 142 | $118 | $107 | −$11 | Marginal — defer at <$25 data-preserve-html-node="true"/oz |
| Development drifts | n/a | $134 | $0 | −$134 | Sunk cost — halt or sequence |
| Beneficiation plant | n/a | $41/t processed | Recovery 88% | Efficient | Operational anchor |
Metify Predictive Scenarios: Silver Price vs. Viable Zone Count
| Silver price scenario | Zones viable | Recommended fronts | Cost/t reduction | Cash flow impact |
|---|---|---|---|---|
| $18/oz (stress) | 1 of 4 | 6 | −22% | −$31M vs. base |
| $21/oz (bear) | 2 of 4 | 9 | −14% | −$18M vs. base |
| $23.6/oz (base 2023) | 2 of 4 marginal | 11 | 0% | Base |
| $26/oz (recovery) | 3 of 4 | 14 | +8% | +$22M vs. base |
| $28/oz (bull) | 4 of 4 | 16 | +15% | +$41M vs. base |
IV. The Three Decisions
Option A — Suspend Proaño Deep, Redeploy to San Carlos
Halt active extraction in the Proaño deep zone (–600m), generating negative economic margin at prevailing silver prices. Redeploy six crews, associated equipment, and ventilation capacity to the San Carlos vein system — 218 g/t grade, $78/t fully loaded cost, $87/t positive margin at $23.6/oz. Maintain minimum ground support to preserve the zone for reactivation on price recovery.
PACE stage: Execute — immediate (0–3 months)
Expected benefits:
- AISC improvement of approximately $4–6/oz through grade mix improvement
- Cost per tonne reduction of 12–16% at mine level
- Free cash flow improvement estimated at $18–24M annually at base silver price
- Operational simplification — fewer active fronts reduces coordination overhead and safety exposure
Risks:
- Proaño deep contains significant silver reserves — suspension reduces optionality if prices recover faster than expected
- Redeployment requires retraining crews on different extraction methods — 4–6 week transition
- Ground support maintenance costs continue during suspension
- Union engagement required under Mexican mining law
Trade-off: Optimizes today's economics at the cost of tomorrow's reserve access. Right if silver stays below $25/oz. Wrong if silver recovers to $28/oz within 18 months.
Option B — Halt All Development Drifts, Protect Operating Cash Flow
Suspend all non-essential capital development — drifts consuming $134/t with zero production contribution — and redirect capital toward sustaining highest-margin production zones. Reduce capex by approximately 25–30% and defer future access development until silver prices justify the investment.
PACE stage: Execute — immediate (0–6 months)
Expected benefits:
- Immediate capex reduction of $22–28M annually
- Cost per tonne improvement as development overhead is eliminated from the active cost base
- Strengthens balance sheet — reduces net debt / EBITDA ratio
- Signals operational discipline to Peñoles and external investors
Risks:
- Development halts create future access constraints — restarting from a disadvantaged position when prices recover
- Certain development is required for structural integrity and safety — cannot be fully deferred
- Creates a production cliff in 3–5 years when current zones are exhausted
- Metify's analysis showed this option maximized 12-month free cash flow while minimizing 5-year NPV
Trade-off: Looks best on next quarter's financial statements. Looks worst on a five-year reserve life analysis. The wrong choice for an organization that can see the long-term consequences.
Option C — Re-sequence the Mine Plan Using Metify, Grade-First Extraction
Use Metify's predictive planning engine to dynamically re-sequence the mine plan around grade-first extraction logic. Active fronts prioritized strictly by economic margin per tonne at current silver price. Development continues only where forward models show price recovery making access investment positive NPV. Resources reallocated monthly based on price signals and zone performance.
PACE stage: Validate → Execute (3–6 month implementation)
Expected benefits:
- Cost per tonne reduction of 14–22% vs. existing plan — without headcount cuts
- Reserve life protected — no zones permanently suspended, only dynamically sequenced
- Creates a repeatable operating model that adjusts automatically to price signals
- Builds the operational intelligence infrastructure that underpins future ABM cycles
- Positions Fresnillo as a technically sophisticated operator ahead of its LSE listing in 2008
Risks:
- Most complex to implement — sustained change management across mine planning, operations, and finance
- Monthly re-sequencing creates planning instability that crew supervisors resist
- Metify models are only as good as the grade and cost data feeding them — data quality investment is mandatory
- 3–6 month transition period means the mine continues on its current suboptimal sequence while implementing
Trade-off: The only response that treats the mine as the dynamic, price-sensitive portfolio it actually is. Options A and B are point-in-time decisions. Option C is a permanent capability.
V. Lessons Learned
1. In mining, the aggregate always lies. Zone-level economics is the only economics that matters. Fresnillo's mine-level cost per tonne looked uncomfortable but manageable. Zone-level cost per tonne, as revealed by PeopleSoft ABM, showed two of four principal zones in negative economic territory. The aggregate was masking a structural problem that would only worsen as silver prices continued to move. The most dangerous number in mining cost management is the average.
2. The mine plan is a financial document, not just a geological one. The sequence in which ore is extracted determines the cost structure of the operation as much as the grade of the ore itself. A mine plan built for $28/oz silver is actively value-destructive at $23/oz — not because the mine got worse, but because the sequence was never designed to flex with price. Metify's predictive planning engine was the first tool that allowed Fresnillo's operations team to treat the mine plan as a variable rather than a fixed constraint.
3. Development costs are the most dangerous hidden subsidy in underground mining. The development drifts at Fresnillo were consuming $134/t in fully loaded costs against $0 in production revenue. What was invisible was the causal attribution: which development costs supported which future production zones, and whether those future zones were economically viable at prevailing or projected silver prices. ABM made the subsidy explicit. Once you see $134/t attached to a zone with no near-term production horizon, the deferral conversation becomes much easier to have.
4. Price scenarios are not forecasts — they are operational triggers. The most powerful output from Metify was not a prediction of where silver prices would go. It was a set of clearly defined thresholds: at $18/oz, do this. At $21/oz, do this. At $26/oz, unlock this. This reframed the entire management conversation from "what will silver do?" to "what will we do when silver reaches each threshold?" — a question the operations team could prepare for in advance. That preparation is what separates a reactive mine from a resilient one.
5. Fixed costs don't flex — but their allocation can be managed. The most important insight from the ABM analysis was not that the mine had too many costs. It was that fixed costs — crews, ventilation, equipment — were being spread across too many active fronts simultaneously, driving up cost per productive tonne in every zone. Concentrating those fixed costs on fewer, higher-grade, lower-depth zones was not a cost cut. It was a cost reallocation that improved economic output without eliminating a single peso of expenditure.
6. What we built at Peñoles became the foundation of what Asher & Company does today. The Metify predictive planning engine that we deployed at Fresnillo in those years was not just a software tool. It was a methodology — a way of thinking about operations as portfolios of activities with distinct economic profiles, managed dynamically against external price signals. Twenty years later, that methodology is the core of how I approach every EPM engagement at Asher & Company. The mine taught me something that no MBA program teaches: the most valuable thing a management information system can do is not report on the past. It is make the future navigable.
7. The hardest conversations in management are not about numbers — they are about identity. The Fresnillo mine had been extracting ore from the same veins in the same sequence for generations. Telling a mine manager that his deepest, most technically challenging zone was, at current silver prices, destroying value rather than creating it, was not a data conversation. It was an identity conversation. PeopleSoft ABM gave us the numbers. What made the difference was having those numbers in a format that the operations team could interrogate, challenge, and ultimately own. The technology was necessary. The human work of translating data into decisions was irreplaceable.
Pedro San Martín is Managing Partner at Asher & Company, a specialized advisory firm in Enterprise Performance Management, profitability architecture, and strategic finance across the Americas and Europe. The methodology behind Asher & Company's practice was developed during implementations including the PeopleSoft ABM and Metify Predictive Planning engagement at Industrias Peñoles / Fresnillo in the early 2000s, when Pedro was part of the PwC advisory team.
asher.company | Battlefield Lessons — real decisions, real consequences, no textbook answers.
