Inventory Intelligence Platform
Capture in the Field · Enrich with AI · Act with Confidence
Eagle Materials delivered record revenue and EPS in fiscal 2025 — but Q3 fiscal 2026 results show a 300 bps year-over-year compression in gross margin, with higher maintenance costs specifically called out. The next compounding advantage isn't another acquisition or the next capacity expansion. It's the master data foundation that lets every dollar of MRO, every part on every shelf, across all fourteen production plants and the broader 70+ facility network, work as hard as the cement leaving Buda.
FY26 Q4 & Full-Year Earnings · May 19, 2026 — next milestone for the executive team
Eagle Materials operates a high-margin, asset-intensive, fourteen-plant network (within a 70+ facility footprint across 21 states) on a maintenance-and-MRO base that — even at industry-average performance — represents roughly $80 million of annual spend. Industry data shows catalog-and-data discipline alone separates top-quartile heavy-materials operators (2.1–2.8% maintenance-to-revenue) from average ones (3.8–4.6%). Ariva closes that gap — and the timing matters: Q3 FY2026 results show a 300 bps year-over-year compression in gross margin, with higher maintenance costs specifically called out in the press release.
Quantified across eight discrete value drivers anchored to Catalyst & Cortex outputs, applied to EXP's revenue base, plant count, and SKU footprint. Geo-corrected: long-lead-time critical-parts driver scoped to actually recoverable spend (commodity emergencies excluded).
Driven primarily by a one-time working-capital release from de-duplicated inventory (~$3M) and Year-1 spend reduction. The model supports faster payback at lower commercial sizing — adjust subscription and implementation sliders in the calculator to test.
Net of implementation and subscription, across the conservative–aggressive range. Materially accretive to ROE — the metric on which EXP's executive compensation is structured. Strategic upside (capex commissioning, M&A, insurance, ESG, cyber) is on top and not in the model.
Field counting with photo and timestamp traceability removes a recurring audit-cycle headache, and provides true-manufacturer and country-of-origin data that Oracle's item master cannot supply natively.
The Company has invested roughly $1.4 billion over the last five years and is in the middle of three concurrent capacity expansions. Every one of them is about to land more equipment, more SKUs, and more spare parts into a master data system that wasn't built to absorb that scale. Now is the right time to put the foundation in place — not after the assets are running.
+50% capacity expansion for the Denver and Salt Lake City markets, completion targeted late 2026. New kiln, new mill, new spare-parts universe.
$330M, +25% capacity at the Oklahoma facility. New SKUs flowing into the American Gypsum item master through commissioning.
Two pure-play aggregates acquisitions in Kentucky and Western Pennsylvania, with aggregates capacity expanded approximately 50%. Each comes with its own catalog, its own suppliers, its own duplicates.
Gross margin compressed 300 bps year-over-year (31.9% → 28.9%) on higher maintenance costs and weather-driven volume losses. Net leverage now 1.8× from 1.2×. Margin defense is the new strategic priority — Ariva is a direct contributor.
Portland Limestone Cement now 75% of cement sales. The Terra CO2 SCM agreement, expansion of slag cement at Houston — every blended-cement product introduces new raw-material SKUs that need accurate identification and country-of-origin coding.
The FY25 10-K cites $3.2M of incremental IT spend specifically on Oracle ERP upgrades — including item-master work — with more in FY26. Ariva's master data layer integrates cleanly with the upgrades already funded.
IIJA-funded infrastructure project demand brings Buy America certification requirements. Country-of-origin traceability is no longer a back-office task — it's a sales-enabling capability.
The May 2024 put option ($550M per 50% interest) expired August 1, 2025 unexercised. The 50/50 JV with Heidelberg remains intact, but the optionality for future consolidation persists — and clean master data makes any future integration faster and cheaper.
Independent industry research on cement and aggregates plants — published 2026 — quantifies the gap between top-quartile and average operators with surprising precision. None of these numbers are Eagle-specific. They describe what is normal in this industry without a data-management foundation.
Top-quartile cement plants spend 2.1–2.8% of revenue on maintenance. Average plants spend 3.8–4.6%. Bottom quartile exceeds 6%. The single largest separator is data infrastructure.
Typical of unmanaged catalogs. The same bearing, same fastener, same lube — six descriptions, six purchase histories, six suppliers, no leverage.
Charged every time a production-critical part isn't on the shelf when needed. Average plants pay this premium on 18-month cycles for parts that exist in another plant's warehouse.
The cumulative premium paid versus a consolidated 40–80 preferred supplier base typically exceeds 12–18% of spend in that category. Visibility is the precondition for consolidation.
Of total maintenance-warehouse value. Spare parts for assets that were decommissioned years ago. Frozen working capital that field counting plus item master linkage exposes immediately.
The fragmented supplier base prevents any single transaction from being large enough to warrant negotiation. Catalog standardization is the precondition for category-level leverage.
Sources: iFactory AI 2026 Cement Plant Maintenance Cost Benchmarks; Oxmaint 2026 Cement Industry Benchmark Report; Oxmaint MRO Procurement Optimization for Cement Plants (March 2026).
Ariva is the inventory intelligence platform purpose-built for asset-intensive manufacturers. The architecture is deliberately layered — capture in the field, enrich with AI, decide with knowledge — because the value of each layer depends on the integrity of the one beneath it.
When Catalyst extracts "S.K.F." from a bearing label at Sugar Creek, Cortex recognizes it as the SKF parent entity. When 500 different Emerson valves get classified by Catalyst, Cortex auto-maps Emerson as a valve manufacturer. The cross-tenant knowledge graph is the structural asset no competitor can copy quickly — it compounds across every Ariva deployment.
Software platforms can scrub records they're given. They can't see what's actually on the shelf in Laramie or Bernalillo. The field-data foundation — physical eyes on every part, with photographic and chain-of-custody audit trail — is what gives the rest of the platform a defensible answer to "are you sure?" That defensibility is what makes the numbers below underwritable by Finance, not just believable to Operations.
"A reference deployment in a comparable multi-site industrial network delivered a 64% increase in inventory data accuracy, identified original manufacturers for critical parts, and eliminated duplicate inventory across locations — enabling consolidated procurement and reducing supplier dependency." — ARIVA · ALLSERV Reference Case · Global Multi-Terminal Operator
All inputs are adjustable. Defaults are seeded from Eagle's public financial profile (FY2024 10-K), the 14-site / 100K-SKU scope, and industry-standard MRO/MDM benchmarks. Toggle scenarios to test sensitivity.
Every line of the model maps to a specific Catalyst output, Cortex graph capability, or ARIVA field artifact — anchored against a published industry benchmark. Where the Ariva GTM project has tighter empirical data than the industry range, those numbers should replace these — flagged below.
| Value driver | Specific Catalyst / Cortex output | Benchmark applied | Why it matters at Eagle |
|---|---|---|---|
| MRO spend reduction | Catalyst separates true manufacturer from OEM repackagers — buy from the upstream OEM, not the distributor markup · Cortex entity resolution collapses "SKF / S.K.F. / subsidiary brands" into one master supplier | 5–15% on managed MRO spend over 18 months (industry MDM range) TBD · Replace with Ariva GTM # |
EXP's fragmented supplier base across 14 sites is the largest single category for leverage. Even mid-range delivery here funds the entire program. |
| Inventory working capital release | Catalyst noun-modifier classification + ARIVA field count make cross-site matching deterministic — the same part in Laramie and Louisville is now provably the same part | 20–40% of duplicate-flagged inventory liberated (one-time) | EXP carries inventory on its balance sheet. Releasing $3–5M one-time has direct impact on free cash flow and ROIC in the year of release. |
| Inventory carrying cost | Catalyst lifecycle flags (Active / Superseded / Discontinued) drive right-sized stocking and disposition of obsolete inventory | 20–28% of inventory value annually (industry standard) | Recurring annual savings on capital cost, storage, insurance, shrinkage, obsolescence — typically the most under-counted lever. |
| Procurement labor productivity | Catalyst structured JSON (parameters, UOM, part-number decode) eliminates manual lookup · Cortex smart alternates surface replacements at point-of-need | 20–35% productivity gain on requisitioning labor | Centralized procurement at Dallas HQ + plant-level requisitioning means productivity gain compounds in two places. |
| Audit & compliance | ARIVA field photography + Catalyst's Hierarchy of Truth = defensible audit artifact for every item · ERP-discrepancy flags pre-empt audit findings | $50–200K per site annually (~$75K base) | Big-4 inventory audit hours on 14 sites collapse when records are visually defensible. SOX implications for inventory accuracy. |
| Trade compliance / COO | Catalyst reads GTIN/barcode and decodes COO from dataplate images — the raw material for tariff classification & Buy America certification | 0.5–1.5% of MRO spend (USMCA, Section 232, Buy America) | IIJA-funded infrastructure projects require Buy America certification on cement. COO isn't optional — it's a sales prerequisite for parts of the cement segment. |
| Maverick spend capture | Catalyst alternates with confidence scores let Procurement enforce "buy this from contract, not that from spot" — the catalog becomes the policy | 40% conversion × 10% savings on maverick base | Reduces leakage from out-of-system buying common at distributed-plant operators. |
| Long-lead-time critical-parts sourcing | Catalyst lifecycle status (Active / Superseded / Discontinued) surfaces high-value critical parts (kiln gear reducers, specialty refractory, custom bearings) weeks before stockout · Cortex M&A lineage produces the modern replacement path (e.g. old Dodge bearing → current ABB part) before the supplier discontinues | 20–30% reduction on the 3–5% of MRO spend that is long-lead-time critical parts (excludes commodity emergency premiums, which are not recoverable across EXP geography) | Commodity-item cross-site sharing isn't viable across 500+ mile distances. The recoverable value is concentrated on long-lead critical equipment where downtime cost dwarfs procurement premium and where Catalyst/Cortex lifecycle data is the actionable signal. |
One of the most common ways an MRO business case overstates value is by assuming every plant in the network can support every other plant. EXP's reality is more constrained — and the model above reflects that. Most plants are 500+ miles from the nearest sibling. Physical cross-site parts sharing only works inside three real clusters, and is only viable for genuine production-critical emergencies, not routine procurement.
| Cluster | Plants | Spread |
|---|---|---|
| NM Twin | Albuquerque + Bernalillo gypsum | ~17 mi |
| Oklahoma | Tulsa cement + Duke gypsum | ~200 mi |
| Midwest cement | LaSalle, Sugar Creek, Louisville, Fairborn (+ Chicago slag) | 200–500 mi |
| Rocky Mtn | Laramie cement + Gypsum CO | ~250 mi |
| Texas | Buda JV + Centex aggregates | ~30 mi |
| Isolates | Fernley NV, Georgetown SC | 500+ mi to nearest |
Implication for the model: emergency-premium avoidance has been sized down from the industry baseline (50% reduction on 8% of MRO) to a more defensible 30–50% reduction on 5–8% of MRO. Most of the value in that driver comes from knowing earlier (Catalyst lifecycle flags, Cortex M&A lineage) — not from physically shipping parts across the country.
Process equipment is different — kilns aren't board lines. But the MRO consumable layer that drives most of the case is heavily shared:
| Common across plant types (~40–60% of MRO SKUs) | |
| ✓ | Bearings, lubricants, gaskets, fasteners |
| ✓ | Electrical (motors, VFDs, contactors, sensors) |
| ✓ | Conveyor belting & components |
| ✓ | Hydraulic / pneumatic / compressed air |
| ✓ | Filters, HVAC, safety/PPE, mobile equipment |
| Process-specific (~40–60%) | |
| ✗ | Cement only: refractory, kiln components, mill liners, cooler grates, preheater cyclones, baghouse media |
| ✗ | Gypsum only: board-line formers, slurry mixers, drying lines, paperboard handling, calciner parts |
Implication for the model: Catalyst's de-duplication value is highest within each segment (cement-to-cement, gypsum-to-gypsum). The commodity-industrial layer adds incremental cross-segment value. The 20% duplicate rate default in the calculator is a reasonable blended estimate.
The ROI model is deliberately conservative — only quantifies value drivers with defensible industry benchmarks. The capabilities below are real and material at EXP, but harder to size precisely without internal data. They are upside on the case, not downside risk.
Mountain Cement expansion (~$300M, completion late 2026) + Duke OK Gypsum modernization ($330M, startup H2 2027). Clean master data at commissioning means new parts land with correct OEM, MPN, COO from day one — avoiding 6–12 months of post-go-live cleanup. Even a 1–2% efficiency on those projects is $6–12M in one-time value.
Eagle has acquired Kentucky aggregates (Aug 2024, $24.9M) and Bullskin Stone & Lime (Jan 2025, $152.5M) recently, and the Texas Lehigh JV remains 50/50 (option expired but consolidation optionality persists). A standard Ariva-formatted item master collapses post-close diligence and integration timelines materially — $200–500K per transaction in pre-close diligence labor alone.
ARIVA field photography + Catalyst chain-of-custody is effectively a pre-built property-loss claim file. In a major loss event (kiln fire, weather damage), the claim documentation is already complete. Property and business-interruption underwriters favorably price networks with verified asset documentation — typical premium-reduction range 1–3%.
EXP's Portland Limestone Cement is already 75% of cement sales, and the Terra CO2 SCM agreement is in flight. Catalyst's true-OEM and COO data, combined with Cortex's manufacturer profile, makes carbon-weighted supplier selection an executable strategy — not a marketing claim. Direct input to PLC's sustainability disclosures.
New SEC cybersecurity disclosure rules and OT-security frameworks (NIST 800-82) require granular inventories of plant-floor controllers, sensors, and network gear with true-OEM identification. Catalyst's image-decoded MPN/manufacturer data is the exact artifact required — and it's a byproduct of the MRO program, not a separate workstream.
Cortex's lifecycle tracking + M&A graph give Eagle 12–18 months of forward visibility on parts supply transitions. Plan ahead for capex on equipment that uses parts whose OEM is about to discontinue them — avoid the 5x cost surprise of a forced unplanned retrofit.
When a bearing fails prematurely, Catalyst's true-manufacturer identity makes it impossible for distributors to deflect warranty/quality claims. Eagle goes to the upstream OEM with chain-of-custody photographic evidence — converts disputed claims to recoverable losses.
For the parts that ARE worth strategic central stocking (specialty refractory, kiln gear reducers, long-lead-time custom components), Catalyst's technical attributes + Cortex's lifecycle data inform a more rigorous central-stocking decision. The right answer to "should we stock this?" is data-driven, not heuristic.
Part-level data with true OEM and manufacture date supports accurate capital asset tagging, more defensible depreciation schedules, and improved property-tax positioning at the plant level. The downstream Finance benefit compounds across the network.
The proposed sequence prioritizes the highest-density spend and the cleanest data wins first, so the savings curve outpaces the cost curve from quarter one.
The risks below are honest. The mitigations are concrete. None of these are unique to Eagle; all of them have been handled at comparable multi-site operators.
Plant-level procurement habits are durable. Catalog discipline collapses without local champions.
Item master is a high-impact system; merge errors propagate to BOMs, planning, and finance.
Common failure mode: the catalog is clean at go-live and dirty again 18 months later as new items are added.
The May 2024 put option creates a 15-month window in which JV ownership may change. Scope inclusion timing matters.
Implementation cost is incurred upfront; full run-rate value lands in Year 2–3.
Reliance on a single platform partner for an item-master-grade capability.
Total Phase 1 commitment is bounded. Decision-grade ROI evidence in 120 days. Full multi-site rollout decision in Q1 of the following fiscal year, with at-scale value capture before the Mountain Cement and Duke OK expansions land their first new SKUs into Oracle.