Office Furniture (including Fixtures) Manufacturing (United States)

NAICS 2022 code 3372 — an investor's rollup primer

NAICS (the North American Industry Classification System) is the standard the U.S. government uses to sort businesses into industries. This is a NAICS industry group (four-digit). It is a single-child level: it contains exactly one industry beneath it, NAICS 33721, and nothing else. For that reason the two codes describe the same universe, and this page is intentionally short — it gives you this level's own ground-truth federal figures and then points you to the child primer, 33721 — Office Furniture (including Fixtures) Manufacturing, for the full story.

---

1. Overview

At the four-digit level, "Office Furniture (including Fixtures) Manufacturing" (NAICS 3372) is one of the smaller, more specialized industry groups in the government's furniture-manufacturing family. The important thing to understand up front is what the name hides. Despite reading like "the business of making desks and chairs," by the government's own numbers office furniture proper is a minority of this group — roughly two-thirds of the revenue is custom architectural woodwork/millwork (built-in casework, paneled boardrooms, reception desks) and store fixtures, shelving, and lockers (retail gondola aisles, display cases, school lockers). Actual office furniture is only about a third.[1][2]

And the imbalance has been widening for two decades. Census benchmarks show the two halves moving in opposite directions: wood office furniture had 569 establishments and 24,301 workers in 2002 against 299 and 13,196 today, while custom millwork went from 1,552 companies and $4.15 billion of shipments in 2002 to 2,227 firms and roughly $10.0 billion — growing and fragmenting further at the same time. The group's center of gravity has been drifting toward the craft and fixtures trades for a generation.[2]

Because 3372 equals its one child, everything true of 33721 is true here: this is a small, mature, deeply cyclical corner of U.S. manufacturing, split across three different demand cycles (the office cycle, the non-residential construction cycle, and the retail capital-spending cycle) and two ownership worlds (a publicly investable office-furniture oligopoly versus thousands of private millwork and fixture shops). All of that detail lives in the child primer; this page establishes the level and its numbers.

---

2. What's inside — and why this level equals its one child

NAICS 3372 contains a single industry:

There are no sibling industries to weigh against each other, so the group's totals are 33721's totals — they are the same rows in the census tables. The real internal variety sits one level further down, inside 33721, which itself splits into four national industries (six-digit codes). Their weights are the opposite of what the group's name implies: custom architectural woodwork/millwork (337212) is the largest at ~35% of revenue and ~43% of jobs, followed by showcase/partition/shelving/locker (337215) at ~32%/~27%; the two office-furniture children — non-wood (337214) at ~21%/~17% and wood (337211) at ~12%/~13% — together account for only about a third.[1][2]

Those four also differ enormously in structure, which is why the group's own averages mislead. Millwork is among the most atomized manufacturing industries in America (~2,227 firms, top-four share 6.7%, HHI 24.4) and fixtures is not far behind (~710 firms, HHI 150.5), while the two office-furniture children are genuinely concentrated (~174 and ~316 firms, top-four shares of 57.7% and 54.3%, HHIs of 1,075 and 944.9) — a roughly 44-fold spread in concentration inside one five-digit code. They ride different cycles and have very different ownership — but that is the child primer's subject. Read 33721 for the four-way breakdown, the investable names, and the segment-by-segment analysis.[1][2]

Note on the family tree: 3372 sits under the three-digit subsector 337 (Furniture and Related Product Manufacturing) alongside household/institutional furniture (3371) and other furniture-related products (3379). Those are siblings of 3372, not part of it — so household cabinets, mattresses, and the like are counted elsewhere. Several close neighbors also sit outside this group entirely: wood kitchen cabinets and laminate countertops (337110), institutional school/lab/library furniture (337127), stock moldings and factory windows and doors (321918/321911), on-site finish carpentry and partition installation (construction codes 238350/238390), and the independent dealers who sell, deliver, and install contract furniture (wholesale trade, 423210) — so the dealer channel's large service markup is not in the $28 billion below.[2]

---

3. How big it is (this level's rollup)

These are our ground-truth federal figures for NAICS 3372 (identical to 33721, since the group is a single child):

Average pay works out to roughly $60,000 per worker per year (annual payroll ÷ employment),[3] and the four sub-industries cluster tightly around that — about $55,000 in wood office furniture, $58,700 in fixtures, and $62,000 in millwork.[2] The rollup below this level reconciles cleanly on the physical counts: the four children's establishments and employment tie out to 3,648 and 104,723 exactly, and receipts add to ~$28.4 billion. Only the firm count differs (the children sum to 3,427 against 3,409 here) — because a company operating in two of these industries is counted once at this level and twice below.[1][2]

Read the concentration number with suspicion. The group's Herfindahl-Hirschman Index (HHI — a standard concentration gauge that runs to 10,000) is just 109.7, with a top-4 firm share (CR4) of 17.5%, CR8 of 24.2%, CR20 of 33.8%, and CR50 of 45.2%.[1] Taken at face value that reads as "highly competitive, no dominant player" — but it is misleading, for the same reason at this level as at 33721: the score is dragged down by the ~2,900 tiny millwork and fixture shops that swamp the firm count, while the office-furniture businesses inside are genuinely concentrated (HHIs of 944.9 and 1,075, top-four shares of 54.3% and 57.7%), and the recent mega-mergers pushed effective concentration higher still. Aggregating an oligopoly with an atomized craft trade produces a number that describes neither. Never quote 3372's headline HHI as a description of the office-furniture business.[1][2]

Measurement caveats specific to this rollup:

• The pieces are not all measured the same way. The fixtures child carries two federal revenue figures — ~$9.2 billion from the 2022 Economic Census (the vintage inside the $28.38 billion above) and $8.807 billion from the 2023 Annual Integrated Economic Survey — and two employment readings, 28,674 in County Business Patterns against roughly 37,900 payroll jobs in BLS's December 2022 series.[2]
• The level is getting harder to track. BLS discontinued the dedicated wood-office-furniture producer-price indexes with the October 2025 release, and folded the fixtures industry into a broader NAICS 3379 grouping that no longer breaks it out. You can still separate price from volume in millwork and racking; you no longer can in wood office furniture.[2]

Undercount and boundary caveats (they run both ways):

• Millwork is genuinely undercounted. County Business Patterns counts only employer establishments, so the many one-person and informal custom-woodworking sole proprietors ("nonemployers") are excluded — we have no published nonemployer count, so we won't put a number on it, but the direction is clear. Much finish woodwork is also installed on site and booked under construction codes, not manufacturing, and some custom work is done in-house by firms classified elsewhere.[2]
• The internal boundary is blurry too. A firm building branded store environments in wood may land in millwork rather than fixtures; many "store-fixture" companies straddle both codes.[2]
• These are U.S. production figures, not consumption. Shipments exclude imports, which are large in furniture and lighter fixtures (Vietnam and China supply a big share of what U.S. offices and stores actually buy). Heavy steel goods like gondola shelving and lockers are freight-protected and stay mostly domestic. Domestic demand exceeds the $28 billion factory line.[2]
• This is not the "office furniture market" you read about. BIFMA and S&P Global size the U.S. business-and-institutional furniture market at $15.27 billion (2023), $15.39 billion (2024, preliminary), and $16.07 billion (2025, preliminary) — a differently drawn series spanning several NAICS codes. Don't compare it to the $28.38 billion here. The nearest independent check on this level is IBISWorld's roughly $30 billion, the same neighborhood as the census figure.[2]

Net effect: the group modestly understates the true economic footprint of these trades. The full undercount discussion is in the child primer.

---

4. The investable universe (where value concentrates)

Because 3372 = 33721, the investable map is identical, and its defining feature is an inversion: public-market value sits in the office-furniture portion, while the two biggest pieces of the group — millwork and store fixtures, together ~two-thirds of revenue — are almost entirely private with no listed pure-play.[1][2]

• Public exposure, cleanly (only through the office-furniture piece): HNI Corporation (NYSE: HNI), the U.S. office-furniture leader at ~$2.8 billion of FY2025 standalone net sales and ~$5.8 billion pro forma after acquiring Steelcase in December 2025; and MillerKnoll (NASDAQ: MLKN, $3.67 billion FY2025 revenue), the Herman Miller + Knoll premium-design pairing. Virco (NASDAQ: VIRC) adds small-cap, school-tilted exposure.[2]
• Public exposure to millwork and fixtures — more threads than you'd think, none of them clean. This is a correction to the simple story: listed exposure exists through UFP Industries (NASDAQ: UFPI, which now owns idX, long cited as a private millwork platform), LSI Industries (NASDAQ: LYTS, Display Solutions), Nucor (NYSE: NUE, pallet rack via Warehouse Systems), Stanley Black & Decker (NYSE: SWK, industrial storage), Berkshire Hathaway (NYSE: BRK, Marmon Retail Solutions), and Global Industrial (NYSE: GIC, a shelving/locker distributor), plus foreign-listed Quadient and InPost on parcel lockers. In every one of those cases the relevant revenue is buried inside a much larger parent and not separately disclosed. You can get exposure; you cannot get a clean read.[2]
• Private exposure (where most of the group actually lives): large family- or employee-owned furniture makers (Haworth, at $2.7 billion of global sales in 2025; KI; Teknion), thousands of family millwork shops plus private-equity roll-up platforms, and the big private fixtures/shelving names (Lozier, Madix, Streater, Spacesaver).[2]

One statistic captures how far below the radar the two largest trades sit: across the roughly 50 largest tracked North American architectural-woodwork and retail-fixture manufacturers, combined 2025 sales were only about $3.5 billion — against the roughly two-thirds of a $28.38 billion group those trades represent. The overwhelming majority of the revenue is in shops nobody ranks.[2]

There is no dedicated exchange-traded fund (ETF) for this level and no public pure-play in its two largest segments. The company-by-company detail — scale, brands, and which segment each touches — is in the child primer.

---

5. How the money works

One economic engine runs the whole group: cyclical, fixed-cost manufacturing where profit comes from keeping expensive plant busy and defending the spread between input cost and selling price.[2]

• Orders and backlog are the pulse. Most output is project-based (a corporate relocation, a building fit-out, a retail rollout), so orders lead revenue by weeks to months; book-to-bill and backlog are the leading indicators. MillerKnoll carried roughly $0.76 billion of unfilled orders at fiscal year-end 2025.[2]
• Capacity utilization drives margins. Heavy fixed-cost factories mean full plants expand margins and idle plants destroy them — the operating leverage behind the group's cyclicality, and it works in reverse fast, since showrooms and dealer-support infrastructure don't shrink as quickly as orders.
• Margins are thin and sit in a surprisingly narrow band across very different businesses. MillerKnoll's North America Contract segment ran a 35.7% gross and 6.2% GAAP operating margin in fiscal 2025; LSI's Display Solutions segment — the closest listed read on fixtures and millwork — ran 18% gross and ~8% operating; custom woodwork shops typically run low-to-mid double-digit gross and single-digit net margins. (The child primers cite conflicting figures for HNI's fiscal-2025 workplace-furnishings GAAP operating margin, 8.5% versus 9.5%, with 10.5% non-GAAP — treat the precise number with care.)[2]
• Input costs split by material and are tariff-exposed. Wood segments live on hardwood, plywood, MDF (medium-density fiberboard) and laminate; metal segments live on steel and aluminum, where Section 232 tariffs flow straight into cost. Contract pricing lags, so input spikes squeeze margins for a quarter or two.[2]
• Stock vs. custom is the margin story. Commodity product competes with imports at thin margins; custom, design-led work is stickier and higher-margin.
• Freight quietly draws the map, but unevenly. Shelving, racking, and lockers are heavy and low-value-per-pound, which favors regional plants and shields the heaviest goods from imports; the office-furniture segments get no comparable protection at the value end.[2]
• A large profit pool sits outside these codes. The independent dealer channel (wholesale, NAICS 423210) captures the design/delivery/installation markup — MillerKnoll put 53.7% of its fiscal-2025 sales through independent dealers, which is also a credit exposure, since dealers often hold the end-customer receivable.[2]

The scorecard: orders/backlog, capacity utilization, gross and operating margin, input-cost pass-through, and dealer health. The full mechanics — including millwork's slow working capital and 5–10% retainage — are in the child primer.

---

6. What drives demand

The group's distinctive feature is that it answers to three separate demand cycles, so there is no single "demand driver" for 3372:[2]

• The office cycle — corporate capital spending, white-collar hiring, and the swing factor of the decade, return-to-office versus hybrid work. Office attendance had recovered to roughly 72.6% of pre-pandemic levels by 2025, and the 2025–26 tone has tilted toward tighter mandates. The sources genuinely disagree on the real-estate backdrop: NAIOP put U.S. office vacancy at 11.8% in its second-quarter 2025 forecast while CBRE cited ~20.7% in the third quarter of 2025 — a gap driven by methodology and coverage, not by one of them being wrong. Either way, new office construction is near its lowest since the 2008–09 financial crisis, partly offset by "flight to quality" refurbishment.[2]
• The non-residential construction and fit-out cycle — offices, hospitals, schools, labs, and data centers buying architectural woodwork. U.S. non-residential construction spending reached $1.226 trillion in 2024, up 7.0%, but the aggregate hides sharp divergence (education +8.5% and healthcare +5.1% against commercial −10.6% and lodging −5.6%). The leading indicator is the AIA (American Institute of Architects) Architecture Billings Index (ABI), which architects register 9–12 months ahead of construction spend. Update from the prior read: the ABI edged toward breakeven in early 2026 (49.8 in March) but slipped back to 47.3 in June 2026, leaving architecture firms 41 months without a majority seeing billings growth. Institutional work remains the firmest segment.[2]
• The retail capital-spending cycle — store openings and remodels, plus e-commerce-driven demand for warehouse racking and parcel lockers. E-commerce was 16.9% of U.S. retail sales in the first quarter of 2026, growing 9.8% year over year against 3.9% for total retail — a shift that pressures traditional store fixtures while creating demand in racking and lockers.[2]

A cross-cutting replacement cycle (furniture and fixtures wear out and are refreshed every ~7–15 years) and steady government/institutional procurement put a floor under all three. Details in the child primer.

---

7. Regulation

Light-touch overall, and mostly standards- and trade-driven rather than heavily licensed:[2]

• Formaldehyde emissions on composite wood panels and the products made from them, under TSCA Title VI (Title VI of the Toxic Substances Control Act), enforced by the U.S. Environmental Protection Agency (EPA) and harmonized with California's CARB (California Air Resources Board) Phase 2 standard. This binds three of the four segments, not just the two wood ones.
• Voluntary but near-universal product standards — ANSI/BIFMA (American National Standards Institute / Business + Institutional Furniture Manufacturers Association) for furniture; AWI Standards and the QCP (Quality Certification Program) of the Architectural Woodwork Institute for millwork, effectively mandatory on institutional work and a genuine competitive moat; and ANSI/RMI MH16.1 for storage racking.
• Trade policy, which now cuts finer than tariffs alone — Section 232 tariffs on imported steel and aluminum (raised to 50% in 2025) and Section 301 tariffs on Chinese goods reshape input costs and import competition; on top of that sit product-specific antidumping and countervailing duties on Chinese steel racks and boltless steel shelving, with further cases covering India, Malaysia, Taiwan, Thailand, and Vietnam. Duty scope and country-of-origin shifts can move domestic price competition quickly.
• Government procurement through GSA (General Services Administration) schedules, the Trade Agreements Act, and Buy American content rules.
• Factory and sourcing rules — OSHA on woodworking dust and metal fabrication, EPA coating NESHAPs, and the Lacey Act's timber declarations (APHIS Phase VII from December 2024; electronic filing from January 2026). Fuller treatment in the child primer.

---

8. Consolidation

This page used to describe two halves consolidating in opposite styles. The revised child supports a sharper version: three distinct consolidation styles, and the difference is where the investment thesis lives.[2]

• The office-furniture segments — dramatic, top-down mega-mergers that collapsed the historic "big five" into a two-company race in five years: Herman Miller + Knoll → MillerKnoll (2021, initially valued at ~$1.8 billion); HNI + Kimball International (2023); HNI + Steelcase (closed December 10, 2025), creating a ~$5.8 billion leader with a targeted ~$120 million of eventual synergies. Two of those price tags are reported differently, and the difference is definitional rather than factual — Kimball at ~$485 million as announced versus ~$504 million of total accounting consideration, and Steelcase at ~$1.9 billion in cash and stock per HNI's SEC filing versus ~$2.2 billion of enterprise value in press accounts.
• The millwork segment — slow, bottom-up private-equity roll-up of profitable regional shops, with no national brand and no wave. It is more fragmented than in 2002 (top-four share 6.7% today against 10.1% then), which is precisely the opportunity.
• The fixtures segment — the update: consolidation here has been led as much by strategic industrial buyers as by financial sponsors. Nucor built Warehouse Systems on its own steel (Hannibal Industries, $370 million in 2021; Elite Storage Solutions, $75 million in 2022); Ali Group bought Metro/InterMetro (2015); UFP Industries absorbed idX; LSI added EMI Industries; and Lozier bought Leggett & Platt's store-fixtures operations (2014). Expect deal-by-deal gathering rather than a wave — and, if you are bidding, expect a strategic across the table.

The barrier in all three is the same: skilled craft and fabrication labor. You can buy shops and brands; you cannot easily buy master woodworkers. Named deals and platforms are in the child primer.

---

9. Risks

Because 3372 = 33721, the risk set is identical:[2]

• Cyclicality — three demand cycles over a fixed-cost base that turns thin margins negative fast in a downturn.
• The secular office question — durable hybrid/remote work and shrinking corporate footprints pressure office-furniture seat counts; the sharper question is how much occupied space is being consolidated or reconfigured, and which categories get the spending.
• The retail e-commerce shift — a long-term migration online caps traditional store-fixture growth, partly offset by warehouse racking and parcel lockers, and further pressured by used racking and automated storage that replaces static shelving.
• Input-cost and tariff whipsaw — lumber/panels and steel/aluminum, with tariffs cutting both ways, and fixtures' price protection resting on trade remedies that can be narrowed or circumvented.
• Skilled-labor scarcity, most acute in custom millwork, where BLS projects the woodworking occupation to shrink 2% through 2034 even as ~21,400 openings a year need filling.
• Fixed-price estimating and working-capital risk in millwork — blown labor estimates come straight out of profit, and slow milestone billing plus 5–10% retainage strains cash.
• Product-liability exposure in fixtures, where pallet rack carries the highest-consequence risk in the group (seismic calculations, anchoring, installation quality, field modifications).
• Dealer failure and customer concentration — a weak dealer can stop paying and take local relationships with it, and losing a national retail program can idle a plant overnight.
• Integration risk as HNI digests Kimball and Steelcase, including overlapping dealer networks and brands.
• Thinning public data — with the wood-office-furniture PPI discontinued and the fixtures employment series folded into a broader grouping, this level is measurably harder to track than it was two years ago.

---

10. How to invest and the outlook

Match the vehicle to the segment — they are not interchangeable.[1][2]

• Public-market investors effectively get only the office-furniture third of this group cleanly. The cleanest listed exposures are HNI (NYSE: HNI) — the scaled, dividend-paying leader — and MillerKnoll (NASDAQ: MLKN) — the premium-design pairing; both are diversified, so you are buying a broad workplace-furnishings cyclical, not a clean segment bet, and with Steelcase inside HNI the public menu is narrower than it was a year ago. Threads into the other two-thirds exist (UFPI, LYTS, NUE, SWK, BRK, GIC, plus foreign QDT and INPST), but in every case the relevant revenue is undisclosed inside a bigger company. There is no ETF and no public pure-play in the two largest segments — anyone marketing "the millwork stock" or "the fixtures stock" is selling a diversified or foreign proxy.
• Private-market investors get the millwork-and-fixtures two-thirds — the larger, more direct opportunity: buy a profitable regional shop, or back a private-equity roll-up platform consolidating fragmented millwork or fixture shops, noting that in fixtures the competing bidder is often a strategic industrial buyer rather than another sponsor. The dealer, installation, and workplace-services layer is almost entirely private and is itself an investable surface.

Near-term to watch: on the office side, the stickiness of return-to-office mandates against ~72.6% of pre-pandemic attendance, and whether HNI realizes its ~$120 million synergy target without integration missteps; on the construction/millwork side, whether the ABI can hold above the low-47s after its March 2026 brush with breakeven, with institutional work the firmest segment; on the fixtures side, retail remodel budgets plus the faster-growing warehouse-racking and smart-locker pockets; and, across all three, lumber/panel and steel/aluminum prices and tariff policy — the single line most likely to make or break a given year's margins.[2]

Bottom line. NAICS 3372 is a single-child industry group: it is the same universe as NAICS 33721, a small, mature, deeply cyclical group that is not one business but four adjacent trades split across three demand cycles and two ownership worlds. The office-furniture side is consolidated, publicly investable, and shrinking — roughly half its 2002 employment; the millwork and fixtures side is the larger, private, still-fragmenting majority that has been growing while the furniture half shrank. Read the headline HHI with suspicion, and go to the 33721 child primer for the segment-by-segment detail, the full investable universe, and the complete sources.

---

Sources

1. U.S. Census Bureau — 2022 Economic Census, Concentration/Comparative Statistics, NAICS 33721 and its four national industries (receipts, firms, CR4/CR8/CR20/CR50, HHI) (2022). https://www.census.gov/programs-surveys/economic-census.html
2. Child primer 33721 — Office Furniture (including Fixtures) Manufacturing (four-way segment breakdown, 2002 benchmarks, investable universe, economics, demand cycles, regulation, consolidation, risks) (2026), and its underlying sources (U.S. Census / County Business Patterns and construction spending; BLS and FRED producer price indexes; SBA size standards; AIA Architecture Billings Index and Consensus Construction Forecast; CBRE and NAIOP office data; BIFMA standards and market size; AWI-QCP; EPA TSCA Title VI; USDA APHIS Lacey Act; USITC trade remedies; IBISWorld; Woodworking Network FDMC-300; SEC filings and releases for HNI, MillerKnoll, UFP Industries, LSI Industries, Nucor, Griffon; Haworth).
3. U.S. Census Bureau — County Business Patterns 2023, NAICS 3372 (establishments, employment, annual and Q1 payroll) (2023). https://www.census.gov/programs-surveys/cbp.html