How Big Is the IT Services Sector, Really? Hong Kong, Singapore and the United States Measured From Official Statistics

The short answer: To size the IT services market in Hong Kong, Singapore or the United States, read the official statistics; and they do not agree on what "IT services" means. Hong Kong's Census and Statistics Department counted 8,179 companies with 54,605 persons engaged in IT and information services in 2024. Singapore's Department of Statistics counted 19,227 establishments in a wider group. The US Census Bureau counted 145,813 establishments in computer systems design and related services in 2023. The three numbers cannot be added or ranked. Where a size table is published, in Hong Kong and the United States, the finding is the same: fragmentation, with most providers very small.

Search for "IT services market Hong Kong" and the results are mostly paid analyst estimates and web pages that quote each other. The official statistical systems of Hong Kong, Singapore and the United States publish counts of companies, people, receipts and value added for the sector, free, with methodology notes. Almost no managed IT content reads them.

This report does. It was written for a buyer who has heard a provider call itself a "market leader" or who wants to know how deep the local supply of IT providers really is. It is also a reference for the rest of this series. Earlier reports could size the provider landscape only in words: our report on the Hong Kong and Singapore provider landscape described the structure of the market without a denominator, and our report on private-equity roll-ups of managed IT providers found no Hong Kong or Singapore deal data at all. The denominator exists. It sits in statistical tables rather than in marketing.

A disclosure. Brocent sells managed IT services in these markets, so it is a participant in the sector this report measures. We have tried to handle that by using only official statistics for every sizing claim, by naming the edition and reference year of each one, and by confining Brocent's own material to one short, labelled section near the end and a few cross-links. We make no market-share estimate for Brocent or for anyone else.

Key findings

• Hong Kong, 2024 (Census and Statistics Department). The group "information technology service activities and information service activities" had 8,179 companies, 54,605 persons engaged, business receipts of HK$94.3 billion and industry value added of HK$46.2 billion. The wider ICT sector had about 17,000 companies, 119,414 persons engaged and value added of HK$183.0 billion, 5.9% of GDP at basic prices.
• Singapore, 2024 (Department of Statistics). The group of SSIC 62-63, which covers computer programming, IT consultancy, information services and online marketplaces, had 19,227 establishments and operating revenue of S$168.2 billion. The infocomm industry revenue dataset that IMDA publishes openly stops at 2017.
• United States, 2023 and 2022 (Census Bureau). NAICS 5415, computer systems design and related services, had 145,813 establishments and 2,050,626 employees in 2023. Its 2022 revenue was US$648.8 billion, of which IT technical support was US$55.9 billion.
• The three boxes are different boxes. Hong Kong counts companies, Singapore counts establishments and the US counts establishments with employees. Each classification draws the sector's edge in a different place. We do not add the figures, rank them or convert them into one number.
• Fragmentation is a measured fact. In Hong Kong 89.2% of the 8,179 companies had fewer than 10 persons engaged, and 132 companies, 1.6%, had 50 or more. Those 132 took 57.5% of the group's receipts. In the US, 83.1% of NAICS 5415 establishments had fewer than 10 employees.
• What the numbers cannot say. None of these series identifies a managed service provider, measures subcontracting, ranks providers or says who is good. They are a denominator, not a league table.

Why the vendor reports cannot answer this

Three kinds of document usually answer the question "how big is the IT services market", and none of them is suited to a buyer who wants a defensible number.

The first is the paid analyst estimate. Analyst firms publish forecasts of IT services spending, but the full reports are sold, and the public figures that circulate are press-release headlines whose method a reader cannot inspect. We have not used any of them, because a figure whose method cannot be opened cannot be checked.

The second is the content-marketing page. A web search for the market size of IT services in a given city returns pages that cite one another, often without naming the series, the year or the definition. A number with no named series is a claim, not a statistic.

The third is the vendor survey. These are useful for opinions, such as what providers say about hiring or pricing, but they sample the vendors who respond. They do not count the sector.

The official statistical systems are different in the one respect that matters: they are legally mandated surveys or administrative counts, run by statistical offices, with published classifications and methodology. Their weaknesses are real, and we state them, but they can be opened and checked. The Hong Kong survey is conducted under the Census and Statistics Ordinance and is a mandatory survey. Singapore's Department of Statistics and the US Census Bureau publish their methods as well.

Three statistical systems, three boxes

A statistic is only as meaningful as the box it counts, and the three jurisdictions use three different classification systems.

• Hong Kong uses the Hong Kong Standard Industrial Classification, Version 2.0 (HSIC). The survey covers "information and communications", codes 58 to 63, and within it the group "information technology service activities and information service activities", codes 62 to 63. The Census and Statistics Department's separate ICT sector measure is wider again, because it also includes companies that distribute and manufacture ICT goods.
• Singapore uses the Singapore Standard Industrial Classification. The Department of Statistics' table that we read had been updated from SSIC 2020 to SSIC 2025. Its group SSIC 62-63 is titled "Computer Programming, Information Technology Consultancy, Information Services & Online Marketplaces". Online marketplaces sit in the same group as IT consultancy, which tells you how wide the box is.
• The United States uses the North American Industry Classification System (NAICS). Its code 5415 is "Computer Systems Design and Related Services". In the Census Bureau's Service Annual Survey the revenue of that code is broken down into sources that include custom application design and development, systems design and integration, IT technical support, IT infrastructure and network management, and hosting and IT infrastructure provisioning. We could not open the Census Bureau's current NAICS definition page, because its server refused our request, so wherever this report says what 5415 contains it relies on those revenue lines from Census's own survey, not on a definition we read.

Two consequences follow. First, an "IT services" figure in one jurisdiction is not the same thing as an "IT services" figure in another, even before currency or year. The Singapore group includes online marketplaces. The Hong Kong group includes information services. The US code includes hosting. Second, none of the three boxes is "managed IT services". A managed service provider that supports a company's laptops, networks and help desk is counted, in each system, among a much larger set of firms that write software, run data centres or advise on systems. The statistics measure the neighbourhood, not the street.

For that reason this report never adds the three headline numbers, never ranks the three markets by size and never computes an "Asia plus US" total. Each market is read on its own terms, with its classification, unit and reference year stated in every line.

Hong Kong by the Census and Statistics Department

Hong Kong has two official views, and they answer different questions.

The ICT sector view. The Census and Statistics Department publishes an annual booklet, *Hong Kong as an Information Society*. We read the 2026 edition, published on 30 June 2026. It says: "In 2024, there were some 17 000 companies in the ICT sector with around 119 400 persons engaged, representing 3.2% of the total number of employed persons in Hong Kong." Its table gives the exact figures: 17,043 companies and 119,414 persons engaged. Value added was HK$183.0 billion, up 9.0% on 2023, and equal to 5.9% of GDP at basic prices. The department marks that last ratio as subject to revision.

The same booklet splits the sector. In 2024 the companies that distribute ICT goods made up 48.1% of the sector's companies, and the companies that provide ICT services 51.8%. In employment the proportions were reversed: providers of ICT services held 63.3% of persons engaged, and distributors 36.3%. So, by our arithmetic, the sector figure of about 17,000 companies includes roughly 8,200 whose business is distributing ICT goods. That is why the second view is closer to what a buyer means.

The industry view. The department's *Annual Survey of Economic Activities* reports a group "information technology service activities and information service activities" inside the information and communications sector. We read the 2024 report, issued in November 2025, and the underlying tables through the department's open data interface. For 2024 they show:

• 8,179 companies, up from 7,643 in 2019;
• 54,605 persons engaged, up from 52,368 in 2019;
• business receipts of HK$94.3 billion; and
• industry value added of HK$46.2 billion.

The whole information and communications sector, which adds publishing, broadcasting and telecommunications, had 11,510 companies, 101,953 persons engaged, receipts of HK$259.1 billion and value added of HK$104.0 billion.

Our own arithmetic on those tables gives an average of 6.7 persons engaged per company in the IT and information services group, and the number of companies grew about 7% between 2019 and 2024 while persons engaged grew about 4%. That is steady, not explosive. It describes a sector that has added firms slightly faster than it has added people.

Two cautions apply. The most recent figures are for 2024, so as of November 2026 the data are almost two years after the end of the reference year; statistical offices publish with a lag, and that lag is a feature of the method rather than a failure. The department also publishes business receipts indices for service industries, which would give a more current signal. We did not use them, because an index of change is not a level and would not add to the denominator this report sets out to build.

Singapore by the Department of Statistics and IMDA

Singapore's picture has a data-quality story of its own.

The Department of Statistics' Table Builder carries the Services Survey Series. The table we read, *Key Indicators By Industry Group In All Services Industries*, was last updated on 4 March 2026 and we generated it on 4 November 2026. It states that its data exclude own-account workers, and that the classification was updated to SSIC 2025. For SSIC 62-63 it shows:

• establishments: 16,156 in 2022, 17,841 in 2023 and 19,227 in 2024, an increase of 19% over two years; and
• operating revenue: S$136.6 billion in 2022, S$152.7 billion in 2023 and S$168.2 billion in 2024, an increase of 23% over the same two years.

The group is the widest of the three boxes in this report, because it includes online marketplaces. A companion table with industry detail shows, for 2024, 8,407 establishments in computer programming activities, 6,944 in IT consultancy and computer facilities management, 2,150 in other IT and computer service activities, and 1,450 in data analytics, processing and hosting. The detail table does not publish operating revenue for those classes, so we cannot tell a buyer how much of the S$168.2 billion is IT consultancy, how much is programming and how much is platform revenue. Dividing revenue by establishments gives a figure that looks like an average firm, but an average hides very different firms, and the table does not publish the distribution.

Singapore's Infocomm Media Development Authority (IMDA) used to publish a dataset of infocomm industry revenue by market segment, including IT services, as open data. That dataset on data.gov.sg covers 2001 to 2017 only, and its record shows a last update in June 2024. We say this plainly because the dataset is still the first result for several searches, and its figures describe a market from before the pandemic.

What IMDA does publish now is wider than the sector. Its Annual Report 2025/2026 says the digital economy accounted for 19.3% of GDP in 2025, and that the tech workforce grew to 222,200 professionals. It also says that "growth is mostly driven by non-Information & Communications sectors, which account for 58% of tech employment." That statement is the reason not to read the tech workforce as the IT services sector. Most people in tech jobs in Singapore, by IMDA's own count, work outside the information and communications industry. The digital economy is a measure of how much of the economy runs on digital technology. It is not a measure of how many companies sell IT services, and the two must not be conflated.

We did not find a table of Singapore IT services establishments by employment size, comparable to the size tables in Hong Kong and the US. The Department of Statistics does publish an enterprise landscape table by SMEs and non-SMEs, but it covers the whole economy, not this industry: it shows that about 99% of Singapore's enterprises are SMEs, and that tells a buyer nothing specific about IT. We will not infer a Singapore size distribution from a different table.

The United States by the Census Bureau and BLS

The US offers the most detail and the longest lag.

Establishments and people. The Census Bureau's County Business Patterns for 2023, a file we downloaded from its open data site, reports for NAICS 5415 a total of 145,813 establishments, 2,050,626 employees in the pay period including 12 March, and an annual payroll of US$250.8 billion. An establishment is a location, not a company, and employees exclude the self-employed, so these counts are not comparable one-to-one with Hong Kong's companies.

Revenue and what it is made of. The Service Annual Survey, which we read through the St. Louis Federal Reserve's FRED data service, put the revenue of employer firms in NAICS 5415 at US$648.8 billion in 2022, up from US$582.6 billion in 2021. It reports sources of revenue, and two lines come closest to what a managed service provider sells: IT technical support, US$55.9 billion, and IT infrastructure and network management, US$46.7 billion. A third, hosting and IT infrastructure provisioning, was US$37.8 billion. Custom application design and development, at US$159.9 billion, is the largest product line apart from "all other operating revenue". These revenue lines come from a different survey and year from the 2023 establishment count, so we do not combine them. The series available through FRED did not carry values after 2022 at the time we retrieved it.

A more current employment signal. The Bureau of Labor Statistics' Current Employment Statistics series for computer systems design and related services showed 2.355 million employees for September 2026. That figure is preliminary and comes from a monthly establishment survey with a different method and coverage from County Business Patterns. We cite it only to show that a current number exists, and we do not reconcile it with the 2023 figure.

The US lag is therefore visible in the table of sources: 2023 for counts, 2022 for revenue and a different survey for the current month.

The number that matters to a buyer: companies by size

Of everything the three systems publish, the distribution of companies by size matters most to a buyer. It is also the figure that most marketing copy never mentions.

Hong Kong, IT and information services, 2024. The Census and Statistics Department publishes this group by number of persons engaged:

• Under 10 persons engaged: 7,294 companies, 89.2% of companies, holding 38.2% of persons and 17.7% of receipts.
• 10 to 49: 753 companies, 9.2% of companies, 27.0% of persons and 24.9% of receipts.
• 50 to 99: 69 companies, 0.8% of companies, 9.1% of persons and 17.7% of receipts.
• 100 to 199: 42 companies, 0.5% of companies, 10.4% of persons and 14.6% of receipts.
• 200 or more: 21 companies, 0.3% of companies, 15.3% of persons and 25.2% of receipts.

Counts are from the department's published table; the shares are our arithmetic on the same table.

Four statements follow from these figures.

• Nearly nine in ten companies have fewer than ten people. Their average is 2.9 persons engaged. These are the proprietors, small software houses and one-site support businesses that make up the long tail of the landscape.
• A small number of companies hold much of the capacity. The 132 companies with 50 or more persons engaged, 1.6% of the total, employ 34.8% of the persons and take 57.5% of the receipts. The 21 with 200 or more alone take 25.2% of receipts.
• Depth is scarce. A buyer who needs a duty rota, a help desk that runs through the night, or an engineer who covers for another engineer, needs a provider with enough people to build one. In this group, a provider of 50 or more persons engaged is a company in the top 2%.
• The mid-sized band is thin but real. The 753 companies with 10 to 49 persons engaged hold 27.0% of persons and 24.9% of receipts. No statistic says which of them are managed service providers.

Note what "persons engaged" counts. It includes everyone working in the company, from software developers to sales and administration, so the number of engineers able to staff a service desk is lower than the figure suggests. The table therefore overstates the supply of rota-ready support staff. It cannot understate it.

The United States, NAICS 5415, 2023. County Business Patterns reports establishments by number of employees. Of 145,813 establishments, 83.1% had fewer than 10 employees and 89.9% had fewer than 20. There were 6,871 establishments with 50 or more employees, 4.7% of the total, and 170 with 1,000 or more. The unit here is a location, so one company with many offices appears more than once.

Singapore has no equivalent published table that we could find, as noted above.

The shape in the two jurisdictions where we can see it is similar: a very large number of very small units and a short list of large ones. That observation is not a claim that the two markets are equivalent, since the units, classifications and years differ. It does bear on what this series has said before. The consolidation described in our report on private-equity roll-ups is a story about acquirers assembling scale from exactly this long tail. The size table is the statistical fact behind that story, in the one jurisdiction where a mandatory survey measures it.

For a buyer, three implications follow.

1. A coverage claim has to be tested against scale. A provider that says it covers a metropolis around the clock with a team of eight has described a rota that cannot exist without overtime, a partner or a sub-contractor. That is not necessarily a defect, but it is a question to ask.

2. Continuity is a size question. In a company of three, the engineer who knows your environment is a large fraction of the whole business. If that person leaves, your knowledge goes too.

3. A "market leader" claim needs a denominator. In Hong Kong's IT and information services group, the top 21 companies by size class are 0.3% of companies. "Leading" in a market of 8,179 companies can mean very different things, and the claim does not say which.

Hong Kong specifically

The focus market of this series deserves its own reading, because Hong Kong buyers ask the question most directly: how many real choices do I have?

What an official statistic lets a Hong Kong buyer infer:

• The number of local companies is large. Eight thousand companies in IT and information services, or about 17,000 in the wider ICT sector, is not a thin market.
• The number of companies with scale is small. There are 132 companies with 50 or more persons engaged, and they include software developers, system integrators, resellers' service arms and the local operations of multinational firms. The set of them that sells managed support to small and mid-sized businesses is smaller again, and no official statistic says by how much.
• Supply is stable, not surging. Companies grew about 7% and persons engaged about 4% between 2019 and 2024. The sector's value added rose 9.0% in 2024, to HK$183.0 billion for the ICT sector. Revenue, not headcount, drove the growth.
• The sector is not a recruiting engine. Total persons engaged in the wider ICT sector fell from 129,535 in 2014 to 119,414 in 2024, while its value added, in current prices, rose from HK$133.7 billion to HK$183.0 billion. The sector employs fewer people and produces more in nominal terms. We draw no conclusion about why; the statistics do not say.

What it cannot tell a buyer:

• Which companies are managed service providers. The classification has no such category.
• Which companies deliver their own work. The surveys count companies and persons engaged, not who does the work or for whom. Whether a company subcontracts is invisible here.
• Which companies are good. No statistic measures quality, response time or customer satisfaction.
• Whether a price is fair. Receipts per company are not prices. Our pricing benchmarks for Hong Kong, mainland China and Singapore cover that different question.

For wages and the cost of employing an engineer, which are the cost side of this market, see our report on the true cost of an IT hire and Brocent's wage-to-rate research, which this report deliberately does not repeat.

A sample of one

A report on a sector should say where the author's own company sits in it, and should say it carefully.

Brocent's Hong Kong entity, Brocent Cloud Service (Hong Kong), is described on our Locations page as the company's Greater China regional hub since 2016, with "15+ engineers". We do not publish the entity's total headcount or receipts, and this report does not infer them. If the entity's persons engaged fall between 10 and 49, which the "15+ engineers" statement allows but does not prove, then it would sit in the 10-to-49 band above, alongside 752 other Hong Kong companies of that size. That band holds 9.2% of the group's companies and 27.0% of its persons.

We include this for two reasons. First, it illustrates how to place a provider in the distribution: ask for its headcount in the market, and read the answer against a distribution like the one above. Second, it shows that, if Brocent's Hong Kong entity does fall in that band, it is one of several hundred firms of similar size in a market made up mostly of very small ones, not an outlier. That is a reading of the distribution, not a statement about Brocent's actual headcount, receipts or market share, and none is claimed.

What this report does not claim

• No market-share estimate. Not for Brocent, not for anyone.
• No forecast. The series end in 2022, 2023 and 2024 depending on the jurisdiction, and we have not extrapolated them.
• No cross-market total or ranking. The three figures count different things in different units and years.
• No claim about managed services specifically. No official classification isolates them.
• No claim about Mainland China. It is deliberately deferred, and its statistical system needs its own treatment.
• No use of paywalled estimates. We did not use analyst forecasts whose method could not be opened.

Sources and how we read them

All were retrieved on 4 November 2026.

• Census and Statistics Department, Hong Kong, *Hong Kong as an Information Society*, 2026 edition, published 30 June 2026: Table 2.1 and Table 2.2 (ICT sector, reference year 2024).
• Census and Statistics Department, Hong Kong, *Key Statistics on Business Performance and Operating Characteristics of the Information and Communications Sector in 2024*, with the press release of 28 November 2025. The size-class figures are from table 640-75102, read through the department's open data interface.
• Singapore Department of Statistics, Table Builder M601471, *Key Indicators By Industry Group In All Services Industries, Annual*, and M601481, by detailed industry; data updated 4 March 2026.
• Infocomm Media Development Authority, Annual Report 2025/2026 (digital economy and tech workforce); and the data.gov.sg dataset *Revenue of Infocomm Industry by Market Segment* (coverage 2001 to 2017).
• US Census Bureau, County Business Patterns, 2023, NAICS 5415, all legal forms.
• US Census Bureau Service Annual Survey, read through the Federal Reserve Bank of St. Louis: total revenue, NAICS 5415, employer firms, IT technical support, IT infrastructure and network management and hosting and IT infrastructure provisioning.
• Bureau of Labor Statistics, Current Employment Statistics, computer systems design and related services, via FRED.

Brocent's own pages used: the Locations page, which supplies the one Brocent fact in the sample-of-one section, and the wage-to-rate research, cited but not used for sizing.

Frequently asked questions

How big is the IT services market in Hong Kong?

By the Census and Statistics Department's 2024 survey, the group of information technology service activities and information service activities had 8,179 companies, 54,605 persons engaged, business receipts of HK$94.3 billion and industry value added of HK$46.2 billion. The wider ICT sector, which also counts companies that distribute ICT goods, had about 17,000 companies and value added of HK$183.0 billion, 5.9% of GDP at basic prices.

Why can't I compare the Hong Kong, Singapore and US figures?

Each uses a different classification, unit and year. Hong Kong counts companies, Singapore counts establishments in a group that includes online marketplaces, and the United States counts establishments with employees in a code that includes hosting. The figures are best read separately.

How many IT service companies are there in Singapore?

For 2024 the Department of Statistics counted 19,227 establishments in SSIC 62-63, which covers computer programming, IT consultancy, information services and online marketplaces. Within it, 6,944 were in IT consultancy and computer facilities management. That is the closest published class to what a managed provider does, but it is not the same thing.

Is Singapore's digital economy the same as its IT sector?

No. IMDA reports that the digital economy was 19.3% of GDP in 2025 and that 58% of tech employment is outside the information and communications sector. The digital economy measures digital activity across the whole economy, not the number of IT service companies.

Are most IT providers small?

In the two places where we found a size table, yes. In Hong Kong's IT and information services group, 89.2% of companies had fewer than 10 persons engaged. In the US, 83.1% of NAICS 5415 establishments had fewer than 10 employees.

Do these statistics tell me which providers are managed service providers?

No. None of the three classifications has a managed-services category, and none measures subcontracting, service quality or price.

Why are the figures a year or two old?

Statistical offices survey, check and publish with a lag. The Hong Kong industry figures here are for 2024, the US establishment counts for 2023 and the US revenue figures for 2022. A current claim in a provider's marketing should be read against that lag.

If you are comparing providers in Hong Kong and want the sizing questions applied to a specific shortlist, you can start from our Hong Kong page or talk to a consultant.