Computer Systems Design and Related Services (U.S.) — NAICS 54151

A rollup investor's primer covering the whole industry group and how its four child industries differ. Written for both public-market and private investors. NAICS = North American Industry Classification System, the standard code set U.S. statistical agencies use to group businesses; IT = information technology; AI = artificial intelligence, computer systems performing tasks that normally require human judgment.

1. Overview

NAICS 54151 — Computer Systems Design and Related Services — is the umbrella over the four codes that describe the U.S. "IT services" economy: firms that get paid for the skilled labor of building, designing, running, and advising on other organizations' computer systems. It is one of the largest professional-services groups in the country: roughly $600.6 billion in annual receipts, about 2.05 million paid employees, and roughly 145,800 employer locations run by around 125,600 firms.[2][3] The four children, from largest to smallest, are custom software development (541511), systems design and integration (541512), running clients' IT operations (541513), and an "everything else" residual of consulting, disaster recovery, and installation (541519).[1]

The common thread is that all four are people businesses, not product businesses. None of them sells a packaged good; they sell expertise, billed by the hour, by the project, or by a recurring service fee. That makes the whole group a fairly direct read on how much corporate America and the U.S. government are spending to build and rebuild their technology — high-quality, high-skill, but cyclical demand tied to discretionary IT budgets. Packaged-software publishers (NAICS 513210) and cloud/hosting providers (NAICS 518210) sit outside this group; they sell products and capacity, not labor.

Why a rollup view helps. No single company, index fund, or contract sits cleanly inside "54151." The interesting decisions are relative: which of the four children is growing versus shrinking, where value concentrates, and how you actually buy exposure — which differs sharply by child. This primer leads with that contrast (Section 2), then covers the group as a whole.

Two doors in. Public-market investors reach the group through a deep but imperfect bench of listed companies — global consulting majors, offshore-delivery majors traded as ADRs (American Depositary Receipts, foreign shares traded in New York), government-focused integrators, and mid-cap "digital engineering" specialists — none a pure play on any one code. Private investors meet a far larger opportunity set: the group is overwhelmingly private and fragmented, and it is one of the most actively consolidated sectors in private equity (PE, firms that buy companies using investor capital and debt, aiming to improve and resell them) today.

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

The group splits into two big children that together are ~87% of it, and two much smaller ones with very different economics. The table contrasts them on the dimensions that matter to an investor; figures are receipts from the 2022 Economic Census and employment from 2023 County Business Patterns (CBP), so shares are approximate.[2][3][4][5][6][7]

(CR4 = share of receipts held by the four largest firms; HHI = Herfindahl-Hirschman Index, a 0–10,000 concentration gauge where anything under ~1,500 is "unconcentrated"; MSP = managed service provider, a firm that runs IT for others under contract.)

How to read the contrast:

• Size is bimodal. Custom programming (541511) and systems design (541512) are near-twins in scale and together make up ~87% of receipts and ~87% of employment. Facilities management and "other" are rounding errors by comparison — but they carry distinct economics and distinct investment routes, which is why they matter out of proportion to their size.
• The line between the two big children is genuinely blurry. A firm that both designs a system and writes the custom code for it self-classifies by primary activity, so revenue slides between 541511 and 541512. Read them as one ~$520 billion "build and integrate" economy split by an accounting convention, not a hard wall.
• Everything here is extraordinarily fragmented — and the group is more fragmented than any child. Every child's HHI sits far below the 1,500 "unconcentrated" line; the group HHI is just 49.9, lower than any single child.[2] The reason is analytically useful: different firms lead each sub-code (a government integrator tops 541519, a digital-engineering specialist tops 541511), so pooling them dilutes any one leader's share. There is no "champion" of U.S. IT services.
• Direction of travel differs. The two big children ride the same tailwind (enterprise AI build-out, cloud, modernization) and the same threat (AI compressing billable hours). Facilities management (541513) is a genuine two-speed story — its legacy data-center-outsourcing core is in slow secular decline even as the AI data-center boom creates fresh operating work. "Other" (541519) is the most exposed to the federal-budget cycle.
• Ownership and access differ most of all. Systems design (541512) has by far the richest public menu; custom programming (541511) has almost no large pure play despite being the biggest child; facilities management is reachable through a few near-pure operators plus adjacent data-center real estate; "other" is dominated at the top by PE-owned federal contractors. That is the single most important practical difference for an investor choosing where to put money.

3. How big it is

Ground-truth federal figures for the whole group. As with the children, these are not one same-year series — receipts and concentration come from the 2022 Economic Census, while employment, payroll, and establishment counts come from 2023 County Business Patterns:[2][3]

A few things these imply. Average pay is about $122,000 per employee (2023 payroll ÷ 2023 employment) — high, reflecting a workforce of developers, engineers, consultants, and cleared specialists.[3] Average revenue per firm is roughly $4.8 million, but that average is misleading: the vast majority of the 125,600 firms fall well under the Small Business Administration's (SBA) ~$34 million threshold, and a small number of giants earn a big slice of the dollars.[2][8] Payroll runs ~42% of receipts — before benefits, subcontractors, and owner pay — confirming the whole group is labor-led.

The children reconcile cleanly to this total. The four CBP figures sum exactly to the group's establishments (145,813), employment (2,050,626), and payroll (~$250.8B), and their receipts sum to ~$600.6 billion — a good internal check that the ground-truth data is consistent.[3][4][5][6][7] The one figure that does not simply add up is firm count: the group's 125,645 firms is slightly below the sum of the four children (~126,100), because a company that operates establishments in more than one sub-code is counted once here but can appear in several child totals.[2]

Undercount caveat — read the $600.6 billion as a floor. County Business Patterns and the Economic Census count only employer establishments with payroll.[9] Three exclusions all push the same way, and several are large precisely where small or individual ownership dominates:

1. Nonemployer firms are largely excluded — the many one-person development, install, and freelance-consulting operations tracked separately in the Census Bureau's Nonemployer Statistics. In custom programming especially, this uncounted freelance layer is substantial.[9]
2. Captive in-house IT never appears here. Enormous amounts of custom software and systems work are built and run inside banks, retailers, insurers, and manufacturers, whose primary industry is finance or retail, not IT services.
3. Government IT staff sit outside the code — federal, state, and local employees running or building public systems are not private-industry establishments.

A fourth distortion is geographic: much work performed for U.S. clients is delivered from offshore centers in India, Latin America, and Eastern Europe, and never lands in U.S.-establishment receipts. This is why third-party "IT services" or "systems integration" markets are sized at roughly $550–760 billion and up — they bundle sibling codes, captive spend, and global delivery.[10][11] Our ground-truth file carries no nonemployer total, no in-house/government total, no public-versus-private ownership split, and no industry-wide margin, utilization, or contract-duration figures — so none is invented here.

4. The investable universe — where value concentrates across the children

There is no large-cap pure play and no dedicated index fund for this group. Public-company revenue should never be read as market share in any of these codes — the big listed firms span all four sub-codes plus adjacent ones, and earn much of their revenue outside the United States. Scale figures below are approximate, from the cited periods, and meant to convey size, not code-level share. Tickers and multiples are confined to this section and Section 10.

The clearest pattern: value on public markets concentrates in the two big children (541511/541512) and in the federal-facing slice of 541519 — not in proportion to the codes' statistical size. Facilities management (541513), a small child, punches above its weight in listed names because a few companies are near-pure operators.

Tier 1 — global diversified majors (systems design + custom build + managed services; the scale benchmarks): Accenture (NYSE: ACN, ~$69.7B revenue, ~786,000 staff),[16] IBM (NYSE: IBM, large consulting arm), Cognizant (Nasdaq: CTSH, ~$20B), and the offshore-delivery majors reached mostly as ADRs — Infosys (NYSE: INFY), Wipro (NYSE: WIT), plus India-listed Tata Consultancy Services and HCLTech, and Europe's Capgemini and DXC.

Tier 2 — mid-cap "digital engineering" specialists (closest thing to a custom-programming pure play, 541511-leaning): EPAM Systems (NYSE: EPAM, ~$4.7–5.5B),[17] Globant (NYSE: GLOB, ~$2.4B),[18] Endava (NYSE: DAVA, ADR),[19] and small-cap Grid Dynamics (Nasdaq: GDYN, ~$0.4B).[20] These are the most concentrated public bets on the group's core — and sit directly in the path of both the AI-demand tailwind and the AI-disruption risk.

Tier 3 — near-pure managed-infrastructure operators (541513-leaning): Kyndryl (NYSE: KD, ~$15B, spun off from IBM in 2021), DXC Technology (NYSE: DXC, ~$12.9B), and Unisys (NYSE: UIS, ~$1.95B).[21][22][23]

Tier 4 — government-focused integrators (heavy in 541512/541519 for federal clients): Leidos (NYSE: LDOS, ~$17B), Booz Allen Hamilton (NYSE: BAH, ~$12B), CACI International (NYSE: CACI, ~$8.6B), SAIC (Nasdaq: SAIC, ~$7.5B), and General Dynamics (NYSE: GD) via its GDIT unit.[24] Government-contracting ("govcon") revenue comes with unusual visibility — multi-year backlog and book-to-bill ratios (Section 5).

Tier 5 — value-added resellers / IT staffers (541519 flavor): CDW (Nasdaq: CDW, ~$21B), Insight Enterprises (Nasdaq: NSIT), ePlus (Nasdaq: PLUS), and staffing names such as ASGN and Kforce — product pass-through and staff augmentation dilute margins here.

Adjacent, not core — data-center owners. Some investors approach the AI-infrastructure theme through the companies that own the buildings — colocation REITs (real estate investment trusts) like Equinix (Nasdaq: EQIX), Digital Realty (NYSE: DLR), and Iron Mountain (NYSE: IRM). These are classified in real estate / 518210, not 54151; they are a capital-and-power bet, not a labor-services one.[6]

Major private and PE-owned owners dominate by count and increasingly by size. The group is where PE goes to consolidate IT services. Recent take-privates reached even listed firms: EQT owns Perficient (2024), Apax Partners owns Thoughtworks (2024), KKR owns Ensono, Carlyle owns ManTech (2022), Bain Capital owns Guidehouse (2023), and Clayton, Dubilier & Rice owns Presidio.[25][26][27][28][29][30] Large founder-led private operators include World Wide Technology, SHI International, Slalom, and Softtek; Peraton (Veritas Capital) is a major private federal integrator; and the technology arms of Deloitte, PwC, EY, Capgemini, and CGI are partner-owned or listed abroad. Beneath all of them sit tens of thousands of small, owner-operated development shops, MSPs, and IT consultancies — the fragmented long tail that the statistics only partly capture.

Takeaway: public exposure means buying a differentiated proxy (a specialist, a major, a govcon prime, or a reseller) and knowing which children it actually touches. The "purest" businesses in the group — small custom-dev shops, boutique MSPs, disaster-recovery specialists — are almost entirely private.

5. How the money works

Because all four children sell labor, the economics rhyme across the whole group. Simplified, revenue for any of these firms is billable headcount × utilization × rate, and profit is that revenue minus mostly-labor delivery cost minus overhead. There is little to depreciate, so cash flow tends to be strong and the balance sheet light — the main assets are receivables, people, and (for acquirers) goodwill.

The shared levers:

• Billable utilization — the share of a worker's paid hours billed to clients, typically 70–80%. Idle "bench" time is pure cost; a few points of utilization is the difference between a good quarter and a bad one. Sector-wide utilization slipped from ~73% (2021) toward ~69% (2024) as demand softened.
• The bill-rate / pay-rate spread and the delivery pyramid. Firms charge a client rate well above what they pay the engineer, and leverage that spread across a pyramid of junior-to-senior staff delivered from lower-cost geographies — "labor arbitrage." Erosion of the onshore-vs-offshore wage gap squeezes margins.
• Contract type sets the risk. Time-and-materials (T&M, bill hours as incurred) is low-risk to the firm; fixed-price keeps efficiency gains but eats overruns; managed services (recurring fees to operate a system) is steadier, more valuable, and commands higher valuation multiples. The mix drives both margin and predictability.

Where the children diverge economically:

• Custom programming (541511) and systems design (541512) are the highest-value, highest-paid work — revenue per employee ~$275K and ~$318K respectively, average pay ~$123K and ~$122K. Margins live on rate discipline, utilization, and a well-shaped pyramid. Well-run private firms target gross margins above ~50% and EBITDA (earnings before interest, taxes, depreciation, and amortization) margins above ~20%.
• Facilities management (541513) is thinner and more annuity-like: multi-year (3–7 year) managed-services contracts, revenue per employee ~$230K, average pay ~$102K (operations labor, not developers), and mid-teens-or-lower EBITDA margins. The metrics that matter are signings/bookings, backlog, and book-to-bill (new bookings ÷ revenue; above 1.0 means backlog is growing) — Kyndryl's record ~$18.2B of FY2025 signings is why investors tolerated a revenue decline that year.[21] Contracts carry SLA (service-level agreement) penalties, so reliability directly protects margin.
• "Other" (541519) blends consulting margins with lower-margin hardware resale and staff augmentation; the federal primes add cost-plus government contracts (low-risk, capped-upside) and live on backlog and recompete wins.

For private owners, the group is valued on EV/EBITDA (enterprise value to EBITDA). Small MSPs and shops trade around 4–8x EBITDA, while scaled, recurring-revenue platforms fetch 11–15x or more — and that spread is the entire engine of the PE roll-up game (Section 8).

6. What drives demand

Demand across all four children is discretionary corporate and government technology spending, so it is cyclical — it expands in confident periods and gets cut fast in downturns. The durable long-run drivers:

• AI adoption — the current wave and the group's biggest near-term demand story. Enterprises and agencies are moving from pilots to production; the money flows to integration — connecting AI to data, workflows, and governance. Accenture alone reported cumulative "advanced AI" bookings around $11.5 billion by early fiscal 2026.[16] (This same wave is also the group's biggest disruption risk — see Section 9.)
• Cloud migration and legacy modernization — a multi-year workstream replacing aging in-house systems, spanning all four children.
• Enterprise software rollouts — deploying and integrating packaged platforms (ERP — enterprise resource planning — plus CRM and workflow suites) is steady, ecosystem-driven work for 541512.
• The AI data-center boom — a specific tailwind for facilities management. Hyperscaler capital spending is on track to approach ~$1 trillion a year by 2030, and all that physical compute needs operating, securing, and managing; the "services for data center" market is forecast to grow from ~$116B (2025) to ~$321B (2030).[12][13]
• Cybersecurity, privacy, and resilience — rising threats and compliance mandates create durable, often recurring demand.
• Government IT and defense budgets — federal, state, and local agencies are large, steady buyers and the single biggest demand engine for the top tier of 541512/541519; government demand is politically cyclical (appropriations, continuing resolutions, shutdowns) rather than economically cyclical.

As a labor-market proxy — not a group forecast — the U.S. Bureau of Labor Statistics projects employment of software developers, QA analysts, and testers to grow 15% from 2024 to 2034, citing AI, automation, cybersecurity, and software embedded in more products.[15] The strongest demand should be for work that is hardest to automate: architecture, regulated-sector implementation, security-sensitive systems, and complex integration.

7. Regulation

The commercial core of all four children is lightly regulated and unlicensed — there is no license to write software, design a system, or run a data center, and no sector-specific regulator. Regulation is driven by the customer, the data, the contract, and the delivery location, and it concentrates in two places common to the whole group:

• Government contracting. Selling to federal agencies means operating under the FAR/DFARS (Federal Acquisition Regulation and its Defense supplement), often requiring security clearances, cloud authorization via FedRAMP (Federal Risk and Authorization Management Program), and — phasing into defense contracts since November 2025 — CMMC (Cybersecurity Maturity Model Certification), which grades how contractors protect sensitive information. Non-compliance can bar a firm from bidding, and false compliance claims carry False Claims Act liability. The SBA size standards (~$34M receipts; $37M for 541513) govern who qualifies for small-business set-asides.[8][34]
• Data privacy and security. Firms must build and operate to their clients' regimes — health-data rules under HIPAA (Health Insurance Portability and Accountability Act), financial-sector rules under the GLBA (Gramm-Leach-Bliley Act) Safeguards Rule, the EU's GDPR (General Data Protection Regulation), and a growing patchwork of state privacy laws — plus market-mandatory commercial audits (SOC 2, ISO 27001, PCI DSS) and SEC cybersecurity-incident disclosure for public companies.[35]

Two cross-cutting points. Immigration policy is effectively industry regulation: the group leans heavily on skilled-worker visas (notably the H-1B), and a sharp 2025 increase in H-1B fees hit the offshore-delivery majors, which are the largest visa users.[33] And the AI governance layer is emerging — the NIST AI Risk Management Framework is voluntary but increasingly written into contracts. Net effect across all four children: regulation shapes cost and labor supply more than it gates entry. Barriers to starting a shop are low; barriers to serving regulated clients (finance, health, defense) are meaningfully higher — which favors incumbents with established controls, clearances, and certifications.

8. Consolidation

The whole group is hyper-fragmented (group HHI 49.9) yet consolidating hard — the apparent contradiction resolves once you see that consolidation is capability-led, not evidence of a concentrating market. Two forces run at once across all four children:

• Private-equity roll-ups. IT services is a textbook PE roll-up target — fragmented, cash-generative, light on capital, with recurring revenue and a wide multiple spread between small firms (4–8x EBITDA) and scaled platforms (11–15x+). Sponsors buy a platform, bolt on smaller competitors cheaply, and aim to exit the combination at a higher multiple. PE was involved in roughly 69% of disclosed MSP deals in 2025, and take-privates have reached listed names (EQT/Perficient, Apax/Thoughtworks, Carlyle/ManTech, Bain/Guidehouse, CD&R/Presidio).[25][26][28][29][30][31]
• Strategics buying capability constantly. Accenture alone made roughly 39 acquisitions in 2024, and the offshore majors keep moving upmarket into higher-value AI and domain work.[16][31] Facilities management has its own consolidation lineage — DXC was formed from the CSC/HPE-services merger, and Kyndryl was IBM's 2021 spin-off of its managed-infrastructure unit.

Cutting across all of it is the hyperscaler and software-vendor ecosystem — Amazon Web Services, Microsoft Azure, Google Cloud, SAP, Salesforce, ServiceNow — which both feeds these firms deals (as implementation partners) and competes with them (when a client moves to the cloud, the platform itself absorbs part of the old outsourcer's role). Expect continued roll-up of the long tail even as the underlying codes stay statistically fragmented; the strategic prize is scaled, AI-fluent, compliance-heavy delivery capacity.

9. Risks

The risks are largely shared across the group, with different emphasis by child:

• AI disruption of the labor-hours model — the defining risk for the whole group, sharpest in the two big children. AI coding and delivery tools can produce the same output with fewer billable hours; if output decouples from headcount, the "bodies × hours" revenue model deflates even as demand for outcomes grows. Independent evidence is mixed — task-level speedups can be large while real-world team gains have been more modest, and AI-generated code can raise rework.[14] Markets have already derated several digital-engineering names on this fear. Whether AI is a net tailwind (more integration work) or a headwind (pricing deflation) is the central open question.
• Cyclicality of discretionary spend — IT projects are among the first things clients pause in a downturn; utilization and bookings fall fast.
• Federal-budget risk — acute for 541512/541519 govcon names. In 2025 the Department of Government Efficiency (DOGE) effort cancelled, cut, or restructured an estimated $85+ billion in federal contracts, explicitly targeting large IT integrators; near-term federal demand is genuinely uncertain.[32]
• Labor cost, attrition, and immigration policy — talent is the product; wage inflation, turnover, and higher H-1B costs pressure margins.[33]
• Fixed-price execution — cost overruns and scope creep can wipe out project profit.
• Structural cloud cannibalization — specific to facilities management (541513), whose legacy on-premises-outsourcing core is in slow secular decline as workloads move to hyperscalers.
• Client, vendor, and geographic concentration — losing an anchor account or a key cloud-partner status can move results; offshore delivery concentrates single-country (India) and currency exposure.
• Cybersecurity and M&A execution — operating clients' systems means owning the blast radius of a breach; and roll-ups can fail through integration problems, customer churn, goodwill impairment, or PE-related leverage.

10. How to invest and the outlook

Public-market routes. There is no single-industry ETF (exchange-traded fund), so treat the listed companies as differentiated proxies, not a sector basket, and match each to the children it actually touches:

• Specialist exposure to the group's core (541511/541512): the digital-engineering names — EPAM, Globant, Endava, Grid Dynamics — concentrated, mid- and small-cap, and directly in the AI cross-current.[17][18][19][20]
• Diversified exposure: Accenture, IBM, Cognizant, and the offshore majors (INFY, WIT, and India-/Europe-listed peers) — steadier, blue-chip-like, where custom build and integration are one component of a broader mix.[16]
• Managed-infrastructure (541513): the near-pure operators Kyndryl, DXC, Unisys — trading on signings momentum and the pace of the legacy-to-modern mix shift.[21][22][23]
• Federal IT (541512/541519): Leidos, Booz Allen, CACI, SAIC, and General Dynamics (GDIT) — defensive government-services names, with the DOGE overhang the key near-term swing factor.[24][32]
• Resellers/staffers (541519) and adjacent data-center REITs round out the menu — but the REITs are a capital/power bet, not a labor-services one.

What to examine everywhere: separate organic growth from acquisitions, and watch bookings, backlog, utilization, attrition, pricing versus wage inflation, delivery geography, customer/fixed-price concentration, cash conversion, and acquisition discipline. Compare EV/sales or EV/EBITDA only among companies with similar service mixes and children.

Private-market routes. This is one of the most accessible industries for direct private ownership, given the low average firm size and the ~$34M small-business threshold. Options run from buying or building a single small shop or MSP (search funds, lower-middle-market buyouts), to backing a PE roll-up platform consolidating the long tail, to private credit against recurring contract cash flows. Diligence should center on recurring/repeat revenue, gross margin by delivery location, bench levels and retention, subcontractor and customer/founder concentration, contract assignability, IP ownership, security controls, and — because these are labor-led businesses with little hard collateral — cash-flow quality and leverage discipline.

Near-term outlook (forward-looking judgment, not settled fact). The group is best understood as high-quality but cyclical exposure to corporate and government technology spending, with an unusually large, still-unresolved question mark over how AI reshapes its economics. The plausible shape: AI-driven integration and modernization demand rises, the AI data-center boom lifts the facilities-management child even as its legacy core shrinks, and federal cybersecurity/defense work stays firm — while the labor-hours pricing model faces genuine pressure and higher immigration costs bite delivery margins. Expect a widening gap between winners and losers: firms that turn AI into higher-value, outcome-priced work, own reusable delivery assets and defensible domain expertise, and hold trusted regulated-sector relationships should prosper; those selling undifferentiated hours, commodity resale, or poorly controlled fixed-price work face pricing and volume pressure. For private buyers, the roll-up thesis holds as long as the multiple spread between small and scaled firms persists — but AI raises the bar on which small firms are worth buying.

---

Sources

1. U.S. Census Bureau, 2022 NAICS — 54151 Computer Systems Design and Related Services and its four component industries (541511, 541512, 541513, 541519): definitions, illustrative activities, and cross-references to 513210 / 518210. https://www.census.gov/naics/?details=54151&input=54151&year=2022
2. U.S. Census Bureau, 2022 Economic Census — Concentration of Largest Firms (NAICS 54151): receipts ~$600.6B; 125,645 firms; CR4 11%, CR8 15.7%, CR20 24%, CR50 32.3%; HHI 49.9 (Histometrics ingested ground-truth federal statistics). https://data.census.gov/table/ECNSIZE2022.EC2200SIZECONCEN?n=54151
3. U.S. Census Bureau, County Business Patterns 2023 (NAICS 54151): 145,813 establishments; 2,050,626 employees; ~$250.8B annual payroll; ~$63.5B Q1 payroll (Histometrics ingested ground-truth federal statistics). https://www.census.gov/data/datasets/2023/econ/cbp/2023-cbp.html
4. Histometrics child primer, NAICS 541511 Custom Computer Programming Services — receipts ~$278.5B; 1,012,250 employees; CR4 11.8%, HHI 54.7 (2022 Economic Census; 2023 CBP).
5. Histometrics child primer, NAICS 541512 Computer Systems Design Services — receipts ~$243.3B; 766,160 employees; CR4 21.1%, HHI 189.6 (2022 Economic Census; 2023 CBP).
6. Histometrics child primer, NAICS 541513 Computer Facilities Management Services — receipts ~$23.8B; 103,562 employees; CR4 13.8%, HHI 80.3 (2022 Economic Census; 2023 CBP); includes adjacent data-center REIT and market context.
7. Histometrics child primer, NAICS 541519 Other Computer Related Services — receipts ~$55.0B; 168,654 employees; CR4 23.2%, HHI 186.6 (2022 Economic Census; 2023 CBP).
8. U.S. Small Business Administration, Table of Small Business Size Standards (NAICS 54151 codes = $34M average annual receipts; 541513 = $37M), 2023. https://www.sba.gov/document/support-table-size-standards
9. U.S. Census Bureau, County Business Patterns methodology (employer-only coverage) and Nonemployer Statistics (freelance/sole-proprietor firms tracked separately). https://www.census.gov/programs-surveys/cbp/technical-documentation/methodology.html · https://www.census.gov/econ/overview/mu0500.html
10. Mordor Intelligence, Custom Software Development Market (mid-teens-percent projected CAGR; major players), 2026. https://www.mordorintelligence.com/industry-reports/custom-software-development-market
11. MarketsandMarkets, System Integration Services Market (~$764B by 2030); IBISWorld, IT Consulting in the US, 2025. https://www.prnewswire.com/news-releases/system-integration-services-market-worth-763-81-billion-by-2030--marketsandmarkets-302595204.html
12. MarketsandMarkets, Services for Data Center Market — Global Forecast to 2030 (~$116B in 2025 to ~$321B by 2030), 2025. https://www.marketsandmarkets.com/PressReleases/services-for-data-center.asp
13. IoT Analytics / Bain & Company, Data-center infrastructure — AI-driven capex toward ~$1 trillion by 2030, 2025. https://iot-analytics.com/data-center-infrastructure-market/
14. Bain & Company, From Pilots to Payoff: Generative AI in Software Development, Technology Report 2025. https://www.bain.com/insights/from-pilots-to-payoff-generative-ai-in-software-development-technology-report-2025/
15. U.S. Bureau of Labor Statistics, Software Developers, QA Analysts, and Testers: Occupational Outlook Handbook (+15% employment 2024–2034), 2025. https://www.bls.gov/ooh/computer-and-information-technology/software-developers.htm
16. Accenture plc, Fiscal 2025 Annual Report (~$69.7B revenue; ~786,000 staff) and advanced-AI bookings disclosures. https://www.accenture.com/us-en/about/company/annual-report
17. EPAM Systems, Inc. — revenue, headcount, and market data. https://stockanalysis.com/stocks/epam/
18. Globant S.A. — revenue and market data. https://stockanalysis.com/stocks/glob/
19. Endava plc — Fiscal 2025 results (NYSE-listed ADR). https://investors.endava.com/
20. Grid Dynamics Holdings, Inc. — revenue and market data. https://stockanalysis.com/stocks/gdyn/
21. Kyndryl Holdings, Inc., Q4 and Full-Year Fiscal 2025 Results (signings ~$18.2B) and Form 10-K. https://investors.kyndryl.com/news-releases
22. DXC Technology Co., Form 10-K, fiscal year ended March 31, 2025. https://www.sec.gov/Archives/edgar/data/1688568/000168856825000029/dxc-20250331.htm
23. StockAnalysis, Unisys (UIS) Revenue (~$1.95B), 2026. https://stockanalysis.com/stocks/uis/revenue/
24. SEC filings and results for Leidos (LDOS), Booz Allen Hamilton (BAH), CACI International (CACI), SAIC (SAIC), and General Dynamics (GD/GDIT), FY2024–FY2025. https://www.sec.gov/cgi-bin/browse-edgar
25. EQT / Perficient, EQT Completes Acquisition of Perficient (~$3B, closed Oct. 2024), 2024. https://www.perficient.com/about/newsroom/news-releases/2024/eqt-completes-acquisition-of-perficient
26. Thoughtworks, Inc., Thoughtworks Completes Transaction to Go Private (~$1.75B, Apax Funds, 2024). https://investors.thoughtworks.com/news-releases
27. Ensono / KKR, KKR Completes Acquisition of Ensono. https://www.ensono.com/insights-and-news/newsroom-and-press/kkr-completes-acquisition-ensono/
28. ManTech / Carlyle, ManTech Announces Completion of Acquisition by Carlyle (~$4.2B, 2022). https://www.globenewswire.com/news-release/2022/09/14/2516381/0/en/ManTech-Announces-Completion-of-Acquisition-by-Carlyle.html
29. Guidehouse / Bain Capital, Guidehouse Completes Transaction with Bain Capital (~$5.3B, 2023). https://guidehouse.com/news/corporate-news/2023/guidehouse-completes-transaction-with-bain-capital
30. Presidio / BC Partners, CD&R to Acquire Presidio from BC Partners, 2024. https://www.bcpartners.com/news/cdr-to-acquire-presidio-from-bc-partners/
31. Dealroom, What Is a Private Equity Roll-Up Strategy? (2025); Solganick, Technology Services and IT Consulting M&A Update (~69% PE share of 2025 MSP deals; Accenture ~39 acquisitions in 2024). https://dealroom.net/blog/what-is-a-private-equity-roll-up-strategy · https://solganick.com/industry-sectors/technology-services-msp-mergers-acquisitions/
32. GovSpend / Washington Technology, DOGE contract terminations in FY2025 (~$85B+; targeted IT integrators), 2025. https://govspend.com/blog/doge-terminations-in-fy25-what-the-numbers-say-and-whats-still-to-come/
33. Business Today, Infosys, TCS, HCL, Wipro: Impact of H-1B visa fee hike, 2025. https://www.businesstoday.in/markets/stocks/story/infosys-tcs-hcl-wipro-impact-of-h-1b-visa-fee-hike-495020-2025-09-22
34. Holland & Knight, CMMC Regulations: Key Questions for Defense Contractors (2025); FedRAMP and FAR/DFARS. https://www.hklaw.com/en/insights/publications/2025/11/cmmc-regulations-key-questions-and-answers-for-defense-contractors
35. U.S. HHS, HIPAA Security Rule; U.S. FTC, GLBA Safeguards Rule; U.S. SEC, Cybersecurity Risk Management, Strategy, Governance, and Incident Disclosure (2023). https://www.hhs.gov/hipaa/for-professionals/security/index.html · https://www.ftc.gov/legal-library/browse/rules/safeguards-rule · https://www.sec.gov/rules-regulations/2023/07/s7-09-22