Signed in as:
filler@godaddy.com
Signed in as:
filler@godaddy.com
Industrial Patterns publishes reference editions describing operating performance, industry structure, and consolidation behavior in US industries. The methodology is designed to be visible, auditable, and stable across editions. Every figure in every edition is computed from publicly available federal data using consistent definitions, with the full methodology appendix included in each edition. This page summarizes the methodology for readers who want to assess it before purchasing an edition.
Industrial Patterns is organized as reference modules. Operating Benchmarks (Module 1) is the annual reference on operating performance; Industry Structure Reference (Module 2) is the census-anchored reference on the structural condition of the same industries. A third module, Add-On Density Atlases, is forthcoming. The two active modules share a single peer set, a common doctrine, and a common publisher; they differ in source architecture and edition cadence. This page describes the shared framework first, then each module in turn.
The shared framework
Peer set
Both modules use the same fixed peer set: 46 six-digit NAICS codes across 14 subsegments and five structural segments — nonmetallic mineral manufacturing, wood product manufacturing, plastics product manufacturing, fabricated and architectural metals, and construction-materials distribution. The peer set is reported on the NAICS 2022 classification spine. Three exclusions are notable: construction firms (NAICS 23) are excluded as a different buyer set with different operating dynamics; retail (NAICS 444) is excluded as too diverse for meaningful intra-segment metrics; mining adjacency (NAICS 21231) is excluded for substantially different operating dynamics.
Distributional benchmarking
Every quantitative pattern in every edition is presented as a distribution across the peer set — typically as 10th, 50th (median), and 90th percentiles. Means appear as reference points but are never visually dominant. The reasoning is straightforward: the median observation is more representative of typical experience than the average, and the spread between p10 and p90 carries more information than either does alone.
This is the principal way the methodology differs from industry reports that present single-point industry averages or forecast trajectories. A reader evaluating a target against Industrial Patterns data sees where that target sits in the cross-industry distribution, not whether it is above or below an average. The interpretation is distributional throughout, in both modules.
Industry-level only — limits of inference
All figures aggregate across firms within a NAICS code or operational segment. Firm-level inference from these aggregates is not supported. Each edition documents the peer-set scope explicitly — the NAICS codes included, the firms or activities excluded, and the source vintages used for each metric. A single firm's figures should be read against its industry's distribution, never treated as a target. The reference describes the population; it does not predict any individual firm's behavior within it.
Descriptive language only
Industrial Patterns describes what is observed. It does not prescribe what should be done. Phrases like "tends to," "is associated with," and "commonly observed" are used to describe regularities. Phrases like "should," "best practice," "causes," and "drives" are not used. Where a reader wants to draw causal inferences, the reader does so with their own judgment against their own information. Each headline page in an edition includes an explicit interpretation boundary describing what that page cannot diagnose.
Restraint posture
The reference does not rank industries, score them on any composite, or characterize any structure as attractive. It does not forecast, value, or recommend. Where the data supports observation, the observation is reported. Where it supports interpretation, the interpretation is reported with its boundaries. Where it does not support causal inference, no causal claim is made. The doctrine is the product's commitment to its readers: every figure can be audited, every observation can be questioned, and every interpretation has a stated limit.
Operating Benchmarks — Module 1
Operating Benchmarks is the annual reference describing how operating performance behaves across scale, volatility, and capital intensity for mid-tier firms in the peer set. Each edition applies a fixed analytical framework organized into four metric families. The framework remains stable across editions; new metric families are not added unless an existing family fails to explain a material class of observed behavior.
Family 1 — Margin Behavior
Question: How do industry operating margins behave under changes in output volume and over time? Primary metric: value-added margin (VA/GO), the share of industry gross output not consumed by intermediate inputs. Source: BEA Industry Accounts. Distribution across the industry peer set, plus trailing time-series.
Family 2 — Labor Elasticity
Question: How does industry labor scale with output and pay? Primary metric: employment-to-wagebill elasticity, computed from BLS Quarterly Census of Employment and Wages (QCEW). Supplementary metrics include real wage per employee (CPI-U deflated) and employment dispersion by firm-size band.
Family 3 — Capital Efficiency
Question: How effectively does invested capital convert into operating output? Primary metric: assets-to-receipts intensity from IRS Statistics of Income (SOI). Sub-metrics include depreciation intensity and asset turnover. SOI data is tax-basis; this caveat appears in the methodology appendix of every edition.
Family 4 — Scale Penalty
Question: What operating costs and frictions emerge purely as a function of organizational scale and coordination complexity? Primary metric: refined overhead intensity, constructed bottom-up from SOI line items (compensation of officers, salaries and wages, repairs, rents paid, taxes and licenses, advertising, pension and employee benefits, and other published deductions). The construction is mechanically coherent and survives SOI's data-suppression patterns at the industry level.
Federal data sources
Operating Benchmarks draws on five federal data sources for structural metrics, plus six additional federal series for the current-conditions appendix. All sources are publicly available; their provenance is verifiable independent of Industrial Patterns.
Structural sources (five):
Current-conditions sources (six): Federal Reserve G.17 Industrial Production; U.S. Census Building Permits Survey; AIA Architecture Billings Index; BLS Job Openings and Labor Turnover Survey (JOLTS); U.S. Census Manufacturers' Shipments, Inventories, and Orders (M3); plus BLS PPI as a faster-publishing cross-check.
Two deflators, two purposes
Dollar-denominated figures in Operating Benchmarks are deflated in two distinct ways depending on the metric frame. Industry-output figures — revenue, gross output, value added, asset aggregates, SOI receipts — are deflated using the BEA Industry Chain-Type Price Index, base year 2017. This reflects industry-output composition and is the defensible choice for industry-level real-dollar metrics. Real wages per employee are deflated using CPI-U, which measures worker purchasing power in the broader consumer market — the policy-relevant frame for wage analysis. The two-deflator construction is standard in BLS productivity work and BEA industry analysis.
Edition lock dates and source vintages
Each Operating Benchmarks edition has an explicit lock date — set to the natural ship period when the edition's structural data would first have been complete, December of the year following the edition's nominal year. The 2022 edition locks at 31 December 2023; the 2023 edition at 31 December 2024; the 2024 edition at 31 December 2025. Once an edition is locked, its data is immutable. Subsequent data revisions appear in the next edition; they do not amend prior editions retroactively. Slow-moving structural anchors (IRS SOI and the Economic Census) carry forward across editions until their next release; this lag is disclosed in the front matter of each edition.
Refined overhead reconstruction (Family 4)
SOI Total Deductions is suppressed at minor-industry granularity for several peer industries and, in any case, is not the correct top-down base because income tax sits outside it. Refined overhead is therefore built bottom-up from published component deduction lines, omitting COGS, depreciation, interest, and income tax. For the industries where Total Deductions IS published, a reconciliation diagnostic confirms the bottom-up figure agrees with the published total within a small tolerance — typically within roughly a third of a percent. The diagnostic is retained in the build for audit.
Honest naming and reconciliation
Where the methodology computes a quantity whose conventional name does not match what the computation actually measures, Operating Benchmarks uses the technically accurate name. Family 1's primary metric is the value-added margin (VA/GO), labeled as such throughout — not the firm-level gross margin from accounting statements, which is a different concept. Family 2's primary metric is the elasticity of headcount with respect to total wage spending, labeled accurately as employment-to-wagebill elasticity, not as the doctrinally-cleaner but uncomputable employment-to-revenue elasticity that QCEW alone cannot support. Family 4's primary metric is refined overhead intensity, reconstructed bottom-up rather than read from a published total.
Industry Structure Reference — Module 2
Industry Structure Reference is the descriptive reference on industry structure for the peer set, anchored to the U.S. Economic Census. The module describes industry structure across seven dimensions rather than the four metric families of Operating Benchmarks. Each dimension is presented as a distribution across the peer set, not a single average.
Seven structural dimensions
Sources and vintages
Population, size, and geography draw on the U.S. Census Economic Census, County Business Patterns (CBP), and Statistics of US Businesses (SUSB). Concentration and legal form draw on the Economic Census. Entry and exit dynamics draw on the Census Business Dynamics Statistics (BDS). Labour composition draws on the Economic Census. The Economic Census is the slow-moving structural anchor and defines the edition; the annually-published sources are drawn at the same anchor for coherence.
Census-anchored edition cadence
The Industry Structure Reference is anchored to the U.S. Economic Census, which is conducted every five years (years ending in 2 and 7). One edition is published per census vintage and serves as the current reference across the five-year cycle, until the next census releases. The First Edition is built on the most recent Economic Census (2022); the next edition publishes with the 2027 Economic Census release (approximately late 2028). This cadence differs from Operating Benchmarks (annual) by design — industry structure changes slowly, and an annual reference would report mostly noise around small year-to-year changes.
Dual-vintage NAICS crosswalk
The reporting spine is NAICS 2022 (the Economic Census vintage). County Business Patterns, SUSB, and BDS publish on the prior NAICS vintage; each is filtered in its native vintage and mapped onto the 2022 spine through a frozen crosswalk, verified code-by-code. Within the Building Materials peer set the only material classification change is in engineered wood, where two prior-vintage codes combine into one on the 2022 spine. Establishment dynamics resolve only to four-digit NAICS, which the Business Dynamics Statistics does not publish below; that granularity seam is disclosed where the figures appear.
Suppression, coverage, and missing data
Every withheld value is shown, never imputed. The Economic Census suppresses small cells (concentration and legal-form detail are not published for every code); County Business Patterns noise-infuses employment and payroll but publishes exact establishment counts; small state-industry cells are withheld at fine geography. Where a published sum falls short of a national total because small cells are withheld, a coverage ratio records the gap rather than filling it.
Cross-module reconciliation
The two modules describe the same industries from complementary angles. Population counts in Industry Structure Reference §1 — firms, establishments, employment, and receipts — are drawn from the Economic Census, the same source Operating Benchmarks uses for its industry-structure section. The two modules consequently reconcile on peer-set headcounts: a reader comparing the segment-composition table in Operating Benchmarks against §1.1 in Industry Structure Reference sees the same firm and establishment counts, by construction.
The size distributions in Industry Structure Reference §2 use Statistics of US Businesses (firms by enterprise employment) and County Business Patterns (establishments by employment); those programs cover the population on slightly different bases, so the size sections are reported as shares of their own universe rather than as counts, and should not be expected to re-sum exactly to the §1 totals. This is documented in the methodology appendix of each Industry Structure Reference edition.
A reader purchasing both modules — typically via the bundle — gets a coherent two-layer view of the same peer set: Operating Benchmarks describes how the industry behaves; Industry Structure Reference describes how it is structured. Each module stands alone; together they answer different questions about the same population.
Where to find more
Each Operating Benchmarks edition contains a complete methodology appendix sufficient for an independent reader to audit any computed value. Industry Structure Reference editions carry the same standard. Free samples of the current Operating Benchmarks edition (Building Materials 2024) and the Industry Structure Reference First Edition (Building Materials) are available on the Editions page. The full editions are available for purchase individually or as a bundle. Both are published by Green Shoot Research.
Inquiries about editions or methodology may be directed through the contact page.
Industrial Patterns is published by Green Shoot Research, a division of Green Shoot Capital Corp. Materials are provided for informational and research purposes only and do not constitute investment, legal, tax, accounting, or operational advice.
© 2026 Green Shoot Capital Corp. Calgary, Alberta, Canada. All rights reserved.
This site uses standard web technologies to operate and to process transactions.