50 mining companies, scored for investor findability
Fifty listed mining companies scored on all fifteen dimensions. The leaders cluster at 7.4 Brand / 6.9 Signal — strong, but not unreachable.
Evaluated
50 organizations- Upper-third average
- 7.4 / 6.9
- Brand, rebuilt site
- 3.9 → 9.1
- Investors 18–24 using social for investing
- 82%
- Likely to act on GenAI information
- 88%
Brand and Signal across the leading cluster.
Same company, same scale, content rebuilt around evidence.
Canadian Securities Administrators, 2024 Investor Index.
Broadridge, 2025 Canadian Investor Study.
Summary
Mining investor discovery has moved online. This study scored fifty listed mining companies on publicly observable evidence: page identity, investor language, project architecture, machine-readable context and third-party proof. The field is tightly bunched. Upper-third companies average 7.4 Brand and 6.9 Signal; the middle third sits in competent sameness around 7.0; the lower third is held back by thin, blocked or unverifiable evidence rather than by the quality of the underlying business. The study also carries a before-and-after: one company's site was rebuilt around content as evidence and moved from 3.9 Brand / 4.0 Signal to 9.1 / 8.9 on the same scale.
What was measured
- All fifteen dimensions — seven Brand, eight Signal — on the published scale.
- Corporate and investor-facing properties, including project and news architecture.
- Crawl files, metadata and structured data on the homepage and key templates.
- Entity consistency: company name, ticker, assets, location and exchange.
- Third-party corroboration: filings, exchange profiles, market databases and media.
Results
Rebuilt site against three upper-third peers.
| Dimension | Before | After rebuild | K92 | Pan American | Equinox |
|---|---|---|---|---|---|
| Brand Entity Presence | 3.5 | 8.6 | 7.5 | 7.3 | 7.0 |
| Page Identity | 3.0 | 9.2 | 7.8 | 7.8 | 7.8 |
| Audience Language | 3.5 | 9.4 | 7.6 | 7.6 | 7.6 |
| Visual Brand | 4.0 | 9.0 | 6.6 | 6.6 | 6.6 |
| Usefulness | 5.0 | 9.1 | 7.5 | 7.5 | 7.5 |
| Brand Voice | 3.0 | 9.3 | 6.8 | 6.8 | 6.8 |
| Content Vitality | 4.5 | 8.7 | 7.3 | 7.3 | 7.3 |
| Average | 3.8 | 9.0 | 7.3 | 7.3 | 7.2 |
| Dimension | Before | After rebuild | K92 | Pan American | Equinox |
|---|---|---|---|---|---|
| AI & Search Accessibility | 4.0 | 9.0 | 7.3 | 7.3 | 7.3 |
| Engineering Strength | 4.0 | 9.0 | 8.4 | 8.4 | 8.2 |
| Offering Architecture | 4.0 | 9.3 | 7.4 | 7.4 | 7.4 |
| Machine Content & Context | 2.5 | 8.8 | 6.9 | 6.9 | 6.9 |
| Channel Reach | 3.5 | 8.7 | 7.1 | 5.8 | 7.1 |
| Third-Party Verification | 5.0 | 8.8 | 7.0 | 6.4 | 5.9 |
| User Action Pathways | 4.5 | 9.1 | 6.9 | 6.9 | 6.9 |
| Response Infrastructure | 5.0 | 8.9 | 6.9 | 6.9 | 6.9 |
| Average | 4.1 | 9.0 | 7.2 | 7.0 | 7.1 |
| Group | Brand average | Signal average | Read |
|---|---|---|---|
| Upper third | 7.4+ | 6.8+ | Strong, but not unreachable |
| Middle third | 6.9–7.1 | 6.5–6.8 | Competent sameness |
| Lower third | 3.9–6.8 | 4.0–6.4 | Thin or blocked evidence |
Brand rank across the fifty-company set.
| Rank | Company | Ticker | Brand | Signal |
|---|---|---|---|---|
| 1 | Rebuilt site (case subject) | — | 9.1 | 8.9 |
| 2 | K92 Mining | KNTNF | 7.4 | 7.1 |
| 2 | Equinox Gold | EQX | 7.4 | 6.9 |
| 2 | First Majestic Silver | AG | 7.4 | 6.9 |
| 2 | Pan American Silver | PAAS | 7.4 | 6.9 |
| 2 | Southern Cross Gold | SXGCF | 7.4 | 6.9 |
| 2 | Eldorado Gold | EGO | 7.4 | 6.7 |
| 8 | Capstone Copper | CSCCF | 7.3 | 6.8 |
| 8 | First Quantum Minerals | FQVLF | 7.3 | 6.8 |
| 12 | Perpetua Resources | PPTA | 7.2 | 6.9 |
| 19 | Teck Resources | TECK | 7.1 | 6.7 |
| 19 | Ivanhoe Mines | IVPAF | 7.1 | 6.8 |
Findings
- Metadata still separates the field. The rebuilt site scores 9.0 on AI & Search Accessibility against 4.0 for its previous version, driven by clearer page identity, crawlable structure and contextual copy.
- Structured data is broadly missing. Neither version of the rebuilt site exposed JSON-LD in the automated homepage pass, which caps Machine Content & Context and entity confidence across the sector.
- The leaders are good, not unreachable. Pan American Silver anchors the upper third at 7.4 Brand / 6.9 Signal because its investor content is broad, current and navigable — not because it does anything exotic.
- Lower-third scores are usually evidence failures. Blocked fetches, unresolved pages and thin content force conservative treatment, which is itself the finding: if an automated agent cannot read it, neither can an answer engine.
- Third-party verification is infrastructure, not PR. Investors cross-check filings, market data and news, and AI systems weight the same corroboration.
What the sector should fix first
- Give every project, report, release and investor action its own clear, linkable page.
- Publish structured data on the homepage and project templates; entity confidence depends on it.
- Keep entity facts identical everywhere: company name, ticker, assets, location and exchange.
- Make the investment thesis readable in plain language before it is defensible in technical language.
- Treat filings, exchange profiles and market databases as part of the digital estate, not as someone else's problem.
Limitations
- Publicly observable evidence only. Private analytics, internal process and unpublished work are outside scope.
- Point-in-time measurement. Scores reflect the evidence visible on the evaluation date.
- Scores are comparable within this study and this methodology version only.
- Where automated fetches were blocked or unresolved, scores were treated conservatively and flagged in the scorecard.
Sources, inclusion and corrections
- 50 Mining Co AI Findability Benchmark scorecard workbook, August 2026.
- Direct site fetches, sitemap/robots/schema checks and search-result checks across the peer set.
- Canadian Securities Administrators, 2024 Investor Index; Broadridge, 2025 Canadian Investor Study; Scotia Global Asset Management / Environics, 2026.
Inclusion in a Brand & Signal benchmark is not an endorsement, and the organizations evaluated are not clients, sponsors or partners of the benchmark. Evaluation is unsolicited and based on publicly observable evidence.
Any organization may request a correction. We review the evidence, and where a factual error is confirmed we republish with a dated change note. Corrections address factual accuracy only — scores cannot be purchased, negotiated or removed, and no commercial relationship changes a result.