Sound Recording Industries (U.S.) — NAICS 5122

A Histometrics rollup primer for public-market and private investors. This is an industry group (four-digit) in the North American Industry Classification System (NAICS), the U.S. government's standard scheme for sorting businesses. It rolls up four child industries. Its value here is the contrast between them.

---

1. Overview

"Sound Recording Industries" is the slice of the U.S. economy that turns performances into ownable, licensable assets and sells the services around them. NAICS groups four very different businesses under code 5122:

• Record labels that own the master recording (the specific captured performance);
• Music publishers that own the composition (the underlying song — melody and lyrics);
• Recording studios that rent rooms and engineering time but own neither copyright; and
• an "other" residual — background-music services, production-music libraries, syndicated radio, event and audiobook recording-for-hire.

The single most important fact about this group is that its four children are not variations on one business — they are two intellectual-property (IP) businesses and two service businesses stapled together by subject matter. The two IP children (labels and publishers) own copyrights that earn like annuities and scale without adding staff; together they are 88.9% of the group's receipts but only ~37% of its firms. The two service children (studios and "other") sell labor and rooms; together they are ~11% of receipts but ~63% of firms.[2] For an investor, that split — not the "music" label on the box — is what determines where the money, the margins, and the buyers are.

A structural quirk worth stating once: every one of the four children is itself a single-child level — each five-digit industry (51223, 51224, 51225, 51229) contains exactly one six-digit national industry, so at the child level there is nothing to aggregate. The genuine rollup analysis therefore happens here, at the four-digit group, where four distinct siblings finally sit side by side. That is what this page does.

---

2. What's inside — the four children and how they differ

The four children answer four different investment questions. The table below is the core of this page; everything after it is commentary.

Receipts shares are computed from our federal file and cross-foot exactly to the group total of $22.33 billion.[2] Firm shares use each child's own firm count; those child counts sum to 4,022, slightly above the group's de-duplicated 4,015 because a few firms (notably the global majors) operate across more than one child and are counted once at the group level.[2]

How to read the four differences that matter:

• Relative size. By revenue the group is really a labels-and-publishers business (89%) with a studio-and-services tail (11%). By firm count it is the opposite — a studio-and-services business (63% of firms) with a small number of large IP owners on top. The same industry looks concentrated or fragmented depending on whether you weight by dollars or by companies.

• Economics. Revenue per firm ranges from about $16.7 million (labels) and $9.7 million (publishers) down to $1.9 million ("other") and $0.77 million (studios).[2] Revenue per employee tells the same story — roughly $1.3M (publishers) and $1.1M (labels) versus ~$0.34M ("other") and ~$0.24M (studios) — because the IP children monetize an asset they own, while the service children sell staff time.

• Ownership mix. The IP children are where public equity and institutional capital live (the majors are listed; PE buys catalogs). The studio child is almost entirely small independents and sole proprietors. The "other" child is private-equity-and-conglomerate territory. Only labels and publishers offer a real public market.

• Concentration. Labels are an oligopoly (top four firms = 86.5% of the child's receipts); "other" is moderately concentrated (57.6%); studios are fragmented (25.7%); publishers sit in between (66.4%).[2] The group as a whole lands at a top-four share of 62.7% and a Herfindahl-Hirschman Index (HHI, a standard concentration gauge that rises toward 10,000 as a market concentrates) of 1,394 — below the 1,800 "highly concentrated" line in the 2023 federal merger guidelines.[2][4]

Full scope, exclusions, and boundary cases for each child are in the four child primers (51223, 51224, 51225, 51229). The group-level scope note that carries up: NAICS puts the streaming platforms that pay this group (e.g., NAICS 516210), the manufacturing of physical media (334610), and independent songwriters and performers who self-release (711510) outside code 5122. The platforms are customers; the self-releasing creators are a large adjacent economy the group's figures do not capture.[1]

---

3. Size (this level's rollup figures)

These are Histometrics' ground-truth federal figures for NAICS 5122, from the U.S. Census Bureau's County Business Patterns (CBP, 2023) and 2022 Economic Census, as recorded in our federal file for this group.[2]

How the group breaks down across its four children (same federal programs; receipts are 2022 Economic Census, the rest 2023 CBP):

Two features jump out. First, this is a small-headcount, high-dollar group: ~26,000 employees nationwide book ~$22 billion of receipts, because most of the value is IP that scales without labor. Average pay is about $128,000 per employee, but it splits sharply by child — roughly $177K at labels and $112K at publishers versus ~$73K at studios and ~$79K at "other" — the signature of an IP-heavy top and a service-heavy bottom.[2] Second, concentration is real but moderate at the group level because the same three global majors — Universal, Sony, and Warner — sit atop both the label child and the publisher child, so the group's top-four share (62.7%) largely reflects the majors' cross-segment dominance, diluted by the fragmented studio tail beneath them.[2][16]

Undercount caveat (matters more here than in most groups). The federal series above count only employer establishments; they exclude the self-employed, nonemployer sole proprietors, and firms without an Employer Identification Number (EIN).[3] That exclusion is not uniform across the four children — it distorts the mix:

• Studios and "other" are badly undercounted. Private industry research counts roughly 22,000 U.S. audio-production studios and ~$1.7 billion of revenue, against the ~2,131 employer establishments in the federal studio figure — implying that on a headcount basis most of that cottage economy is invisible here.[8] Freelance audiobook narration, one-person production-music work, and for-hire event recording sit largely outside the "other" count too.[3]
• Self-releasing creators sit in a different code entirely. Independent artists and self-publishing songwriters are classified in NAICS 711510, not 5122, even though they now account for a large and rising share of new releases. The group's figures capture the corporate sound-recording industry, not everyone making or owning recordings.[1]
• End-market spending is a different number. Census receipts (revenue booked by U.S. establishments) are not the same as trade-body market revenue. For context, the Recording Industry Association of America (RIAA, the U.S. labels' trade group) reported U.S. recorded-music revenue at a record $11.5 billion wholesale in 2025,[5] and the National Music Publishers' Association (NMPA) reported U.S. publishing revenue of $7.3 billion in 2025.[7] These are built differently from Census receipts and should not be used interchangeably.

Read the $22.3 billion as the measured corporate core, understating the studio and "other" tails most and the two IP children least.

---

4. Investable universe (where value concentrates across the children)

Value concentrates exactly where the receipts do — in the two IP children — and the public market reaches only a sliver of the group.

• The two IP children are where public equity exists — and it is thin. The only clean listed exposure to the whole group runs through the three global majors, each of which owns both a top record label and a top publisher: Universal Music Group (Euronext Amsterdam: UMG; U.S. over-the-counter: UNVGY), the largest; Warner Music Group (Nasdaq: WMG), the cleanest listed pure-play major; and Sony Group (NYSE: SONY), where a fast-growing major is buried inside a games/film/electronics conglomerate. The one small U.S.-listed near-pure-play rights owner is Reservoir Media (Nasdaq: RSVR).[15][16][21] Buying any of these buys labels and publishing together.
• The largest private owners define the top of the IP children but are not listed equities: independents Concord and BMG (Bertelsmann), plus a fast-growing tier of private-equity (PE) catalog buyers (e.g., Blackstone-backed vehicles) that finance purchases with asset-backed securities (ABS) — bonds repaid by the recordings' or songs' royalty income.[15][17][20]
• The two service children have no public pure-play at all. Studios are private, owner-operated, and reachable only by owning/operating one or playing the commercial real estate around them. The "other" child is private-equity-and-conglomerate territory — background-music leader Mood Media (Vector Capital), production-music firm Epidemic Sound (EQT/Blackstone), the APM Music library (a Sony/Universal joint venture), and audiobook roll-up RBMedia (H.I.G./Francisco Partners) — with the nearest listed proxy being commercial-music provider Stingray Group (Toronto Stock Exchange: RAY.A/RAY.B), plus podcast/audio names such as PodcastOne (Nasdaq: PODC), iHeartMedia (Nasdaq: IHRT), and Sirius XM (Nasdaq: SIRI).[22][23][24][25]

The one-line takeaway: the public market lets you buy the top of the two IP children (as diversified majors or one small-cap) and almost nothing else. The studio and services value is private by nature. Company-by-company tickers, valuations, and the private-vehicle landscape are in the four child primers.

---

5. How the money works

The group runs on two engines that map directly onto the IP-versus-service split:

• Copyright income (the labels-and-publishers engine). A master recording (owned by labels, 512250) and a composition (owned by publishers, 512230) are separate copyrights, licensed and paid separately. When a song is streamed, the digital service keeps roughly 30% and pays about 70% to rights holders; of that, the recording side takes roughly 80% and the song side about 20%.[15] Publishers earn across four streams — performance (public performance/broadcast/streaming, collected by performing-rights organizations, or PROs), mechanical (reproduction), synchronization ("sync," placements in film/TV/ads/games), and print — while labels earn from streaming, downloads, physical formats, sync, and public-performance licensing. Because these flows are recurring and predictable, both catalogs of masters and catalogs of songs trade like financial assets, priced on a multiple of net annual royalty income.[15][19]
• Service fees (the studios-and-"other" engine). Studios (512240) sell room-hours × utilization × rate and, critically, capture no copyright — the master flows to the label, the composition to the publisher. Their economics resemble a boutique hotel: high fixed cost (real estate, buildout), perishable capacity, very profitable incremental sessions, and cash-bleeding empty rooms. The "other" child (512290) is a blend — recurring, high-margin licensing (background-music subscriptions per location; production-music sync licenses) that behaves like the IP engine, plus lower-margin for-hire recording (events, narration) that behaves like the studio engine.

The investable prize across all four is the owned-asset half — catalogs, subscription bases, and libraries that earn again from work done once — not the labor-driven service half. Deal structures, catalog-valuation multiples, advances/recoupment, and break-even math are in the child primers.

---

6. Demand drivers

One force dominates the whole group: paid streaming. It has driven U.S. recorded-music revenue up for nine straight years and is the primary engine for both IP children.[5][6] Around it:

• Sync and content volume — the flood of film, TV, streaming, gaming, advertising, and creator-economy (YouTube/TikTok/Instagram) content lifts publishers' sync fees, labels' licensing, and production-music libraries in the "other" child.[7][10]
• Administrative rate increases — Copyright Royalty Board (CRB) and PRO rate-court rulings raise royalties industry-wide, benefiting publishers and labels at once.[12][13]
• Superfan and physical fandom — U.S. vinyl revenue topped $1 billion in 2025, supporting labels.[5]
• Adjacent audio booms — double-digit growth in audiobooks (U.S. sales ~$2.22 billion, +13% in 2024) and podcast advertising (~$2.86 billion, +17.6% in 2025) feeds the "other" child and studio demand.[9][10]
• Interest rates — the swing factor for asset values: catalogs of masters and songs are valued like bonds, so their prices move inversely to rates.[15]

The clearest headwinds are child-specific and pull in opposite directions: cheap home-recording gear and urban real-estate cost shrink the commercial studio middle, even as licensing demand grows above it.[8]

---

7. Regulation

NAICS is a statistical classification, not an operating license — there is no permit to run a label, publisher, studio, or background-music business. What binds the whole group is U.S. copyright and music-licensing law, and it applies most heavily to the two IP children:

• Two copyrights, licensed separately. The Music Modernization Act (MMA, 2018) created the Mechanical Licensing Collective (MLC) and its blanket streaming license for compositions; SoundExchange collects the statutory digital public-performance royalty owed to recording owners.[11]
• Administrative rate-setting. The Copyright Royalty Board (a federal-judge panel) sets statutory mechanical rates and is actively re-setting rates for the late-2020s and early-2030s; performance royalties for the dominant PROs (ASCAP, BMI) run under Department of Justice antitrust consent decrees and federal rate courts.[12][13]
• The AM/FM gap. The U.S. still grants no general performance right for over-the-air radio to recording owners — a gap the American Music Fairness Act (AMFA) would close, though it is not yet law.[15]
• Lighter-touch corners. Studios and for-hire "other" work face contract/work-for-hire terms, union agreements, and (for meeting/call recording) consent rules — real obligations but no licensing barrier to entry.[15]
• The open question everywhere: generative artificial intelligence (AI) and digital replicas — an unsettled area the U.S. Copyright Office is still addressing, with direct bearing on catalog values in both IP children.[14]

Full detail is in each child primer.

---

8. Consolidation

Concentration is child-specific, and the group figure averages four very different structures:

• Labels (512250): an oligopoly. The top four firms take 86.5% of the child's U.S. receipts; globally in 2024 Universal (~31.7%), Sony (~21.7%), and Warner (~15.3%) held about two-thirds of recorded-music revenue.[2][16]
• Publishers (512230): concentrated but with a long tail. Top four = 66.4%, HHI ~1,380 — below the "highly concentrated" line because a tail of 700-plus firms holds the index down.[2]
• "Other" (512290): moderately concentrated — top four 57.6%.[2]
• Studios (512240): fragmented — top four just 25.7%, and even that understates fragmentation once nonemployers are added.[2]

At the group level this blends to a top-four share of 62.7% and an HHI of 1,394 — moderate, and driven by the majors straddling both IP children.[2] (A NAICS group is not automatically an antitrust market; read this as a directional signal, not a legal finding.[4]) The defining trend across the IP children is a multi-year catalog-buying wave — a 2020–2022 frenzy, a 2023–2024 cooling as rates rose, and a 2025–2026 revival — increasingly financed with ABS. Recent company-level landmarks span the group: BMG + Concord (combining the two largest independents), Universal's Virgin Music Group acquiring Downtown Music, Primary Wave's acquisition of Kobalt, and Concord's takeover of the Hipgnosis catalog fund.[17][18][19][20] In the service children, consolidation runs through PE assembling recurring-revenue assets (Mood Media in background music; RBMedia in audiobooks), while the studio child sees attrition, not roll-ups, as mid-tier rooms close.[23][25]

---

9. Risks

The group's risks are its children's risks, weighted by where the money is:

• Interest-rate and catalog-valuation risk (hits the two IP children hardest). Catalogs of masters and songs are valued like bonds; rising rates compress the multiples and can impair ABS-financed vehicles.[15]
• Generative AI (the shared wildcard). AI-generated music and narration could dilute the royalty pool and devalue catalogs, or (for library owners) widen margins — with the legal status of training on copyrighted work unresolved. Majors have both sued AI startups and begun licensing.[14][15]
• Platform concentration and pricing. A few streaming services control access to listeners (Warner reports its three largest digital accounts are ~45% of recorded-music revenue); disputes over "bundle" pricing have erased hundreds of millions in royalties.[15]
• Regulatory / rate risk. An adverse CRB or rate-court ruling hits the IP children at once.[12][13]
• Service-child structural pressure. Studios face technology substitution (home gear), urban real-estate inflation, thin utilization-driven margins, and — critically — no copyright upside; the "other" child faces commoditization of for-hire work and platform dependence.[8][15]
• Measurement and access risk. Most of the group's dollars are IP the public market lets you touch only as diversified majors; most of the group's firms are private micro-businesses. A listed company may carry a recognizable audio brand yet have little economic connection to a given child — look through to the specific sub-segment and owner.

---

10. How to invest & outlook

Where the routes actually are, by child:

• Labels + publishers (89% of receipts). This is the investable core. Public: the diversified majors UMG, WMG, and SONY (each buys labels-and-publishing together), plus small-cap Reservoir Media (RSVR) as the nearest near-pure-play rights owner. Private (where most institutional capital goes): music-royalty and catalog funds, music ABS (a credit route dominated by recent catalog deals), and direct catalog ownership (illiquid, specialist). Investors here watch subscriber and subscription-revenue growth, catalog-versus-frontline mix, operating margin, per-subscriber pricing, and — for any deal — ownership quality and purchase price.[15][21]
• Studios (7%). No listed pure-play. Exposure is operational — own/operate a studio as a small business, or play studio-anchored commercial real estate — or a demand-side proxy through podcast/audio platforms.[8]
• "Other" (4%). No listed pure-play. Nearest listed proxy is Stingray (RAY.A/RAY.B) for commercial music; podcast/audio via PODC/IHRT/SIRI; the real specialist assets (Mood Media, Epidemic Sound, APM, RBMedia) are private and reached through PE/growth-equity funds or direct deals in subscription bases and production-music catalogs bought as yield assets.[22][23][24][25]

Outlook (judgment). The base case is streaming-led growth in the two IP children, extended by rising sync demand and better royalty collection, keeping labels and publishers the group's growth-and-value engine; the main downside risks there are the path of interest rates, streaming/rate-setting disputes, and how AI copyright questions resolve. The studio child stays roughly flat to slightly down (private research puts the professional core near $1.7 billion), with real growth concentrated in podcasting and immersive/spatial audio; the "other" child grows in its licensing and subscription niches while its for-hire work stays commoditized. Net for a public investor: this group is best owned as one of the diversified majors plus, optionally, a small-cap rights owner — a bet on copyright annuities, not on rooms or services. Net for private capital: the prize is the owned-asset half — catalogs, subscription bases, and libraries — which is exactly why it is priced richly and held privately.

Because each of the four children is itself a single-child level, the full deep-dive for any one of them lives in its own primer (51223 publishers, 51224 studios, 51225 labels, 51229 other) — start there for company tables, deal detail, and diligence checklists.

---

Sources

Group-level figures are from Histometrics' ground-truth federal file for NAICS 5122 (U.S. Census Bureau, County Business Patterns 2023 and 2022 Economic Census). Remaining citations are drawn from the four child primers (51223, 51224, 51225, 51229).

1. U.S. Census Bureau. "2022 NAICS Definitions — code 512 and children" (scope, exclusions; 516210 streaming, 334610 media manufacturing, 711510 independent artists). https://www.census.gov/naics/?details=512&input=512&year=2022
2. Histometrics ground-truth federal file, NAICS 5122 and its four children — County Business Patterns 2023 (establishments, employment, annual/Q1 payroll) and 2022 Economic Census (firms, receipts, CR4/CR8/CR20/CR50, HHI). U.S. Census Bureau. https://www.census.gov/data/datasets/2023/econ/cbp/2023-cbp.html and https://data.census.gov/table/ECNSIZE2022
3. U.S. Census Bureau. "County Business Patterns Methodology" (excludes self-employed and nonemployer businesses). https://www.census.gov/programs-surveys/cbp/technical-documentation/methodology.html
4. U.S. Department of Justice / Federal Trade Commission. "2023 Merger Guidelines — Guideline 1" (HHI thresholds). https://www.justice.gov/atr/merger-guidelines/applying-merger-guidelines/guideline-1
5. Recording Industry Association of America. "US Recorded Music Annual Revenue Achieves New High of $11.5 Billion in 2025." 2026. https://www.riaa.com/riaa-reports-us-recorded-music-annual-revenue-achieves-new-high-of-11-5-billion-in-2025/
6. Recording Industry Association of America. "2024 Year-End Music Industry Revenue Report" (U.S. recorded-music revenue $17.7B). 2025. https://www.riaa.com/reports/2024-year-end-music-industry-revenue-report-riaa/
7. National Music Publishers' Association. "US music publishing revenues hit $7.3B in 2025." 2025. https://www.nmpa.org/us-music-publishing-revenues-hit-7-3b-in-2025-nmpa-reveals-at-annual-meeting/
8. IBISWorld. "Audio Production Studios in the US — Industry Analysis (NAICS 512240)" (~22,000 studios, ~$1.7B revenue, podcast growth). 2026. https://www.ibisworld.com/united-states/industry/audio-production-studios/1254/
9. Audio Publishers Association / Publishing Perspectives. "US Audiobook Sales Reach $2.22 Billion" (+13% YoY, 2024). 2025. https://publishingperspectives.com/2025/06/audio-publishers-association-us-audiobook-sales-reach-2-22-billion/
10. Interactive Advertising Bureau & PricewaterhouseCoopers. "IAB/PwC Internet Advertising Revenue Report: Full Year 2025" (U.S. podcast ad revenue ~$2.86B, +17.6%). 2026. https://www.iab.com/insights/internet-advertising-revenue-report-full-year-2025/
11. U.S. Copyright Office. "The Music Modernization Act" (MLC, blanket mechanical license, SoundExchange). https://www.copyright.gov/music-modernization/
12. Copyright Royalty Board. Announcements and rate determinations. https://www.crb.gov/announcements/
13. U.S. Copyright Office. "Issues Related to Performing Rights Organizations" (ASCAP/BMI consent decrees, rate courts). https://www.copyright.gov/policy/pro-issues/
14. U.S. Copyright Office. "Copyright and Artificial Intelligence" (AI training, outputs, digital replicas). https://www.copyright.gov/ai/
15. NAICS 512250 and 512230 child primers, this series — full label and publisher economics, royalty splits, catalog valuation, platform-concentration and rate-setting detail, and complete numbered sources.
16. Music Business Worldwide / MIDiA Research. "Sony vs. Universal vs. Warner music revenues; global major-label market shares." 2025. https://www.musicbusinessworldwide.com/data/sony-vs-universal-vs-warner-annual-music-publishing-revenues-in-usd-2021-2024/
17. BMG. "BMG and Concord Combine to Create World's Leading Independent Music Company." 2026. https://www.bmg.com/news/bmg-and-concord-combine-to-create-worlds-leading-independent-music-company
18. Music Business Worldwide. "Done Deal: Primary Wave's Acquisition of Kobalt Has Closed." 2026. https://www.musicbusinessworldwide.com/done-deal-primary-waves-acquisition-of-kobalt-has-closed/
19. Universal Music Group. "Virgin Music Group Receives European Commission Approval for Downtown Acquisition." 2026. https://www.universalmusic.com/virgin-music-group-receives-european-commission-approval-for-downtown-acquisition/
20. Variety. "Hipgnosis Songs Fund Agrees to $1.4 Billion Takeover by Concord." 2024. https://variety.com/2024/music/news/hipgnosis-songs-takeover-by-concord-1235974846/
21. Reservoir Media, Inc. Form 10-K for the fiscal year ended March 31, 2025 (SEC EDGAR). https://www.sec.gov/Archives/edgar/data/1824403/000141057825001379/rsvr-20250331x10k.htm
22. Stingray Group Inc. "Annual Report 2025" (broadcasting & commercial-music segment). 2025. https://corporate.stingray.com/wp-content/uploads/2025/06/stingray-annual-report-F2025-vf.pdf
23. Mood Media. Company materials (500,000+ subscriber locations, 100+ countries). 2025. https://us.moodmedia.com/
24. Wikipedia. "APM Music" (Sony/Universal joint venture; ~1.1M tracks). 2025. https://en.wikipedia.org/wiki/APM_Music
25. Publishers Weekly. "With a Twist, Sale of RBmedia Completed" (H.I.G./Francisco Partners acquire RBMedia for >$1B). 2023. https://www.publishersweekly.com/pw/by-topic/industry-news/industry-deals/article/93205-with-a-twist-sale-of-rbmedia-completed.html