
Ask the internet how big the IT outsourcing market is and you'll have a dollar figure in three seconds. Follow it back and you'll usually land on a paywalled report with no sample, no stated method and nobody willing to put their name on it. Most IT outsourcing statistics circulate exactly that way. This page carries no such number. Refusing to print one is the most useful thing on it.
Key takeaways
Three different quantities get called "the outsourcing market," and they are not interchangeable. IT services spend is what buyers spend on IT services in total, outsourced or not — Gartner's Services segment forecast of $1,570 billion for 2026 [N2]. Contract value is annual contract value newly booked in large deals during a single quarter, counting only commercial outsourcing contracts of $5 million or more, and it is not revenue recognized [N12] [N13]. Outsourcing revenue is what providers actually earn for outsourced work, and no methodology-backed global figure for it exists in public form. These three quantities cannot be added together, compared with each other, or converted into one another.
Everything below is a figure someone actually measured, labeled with what it counts, who counted it, and the period it covers. Where a number is a survey self-report rather than a measurement, it says so. Same when it's a forecast rather than an actual.
One thing is missing on purpose. This page publishes no outsourcing market-size figure, because no methodology-backed one exists in public, fetchable form: nothing pairs a named analyst, a stated method and a total-market dollar value for outsourcing specifically. If you came looking for IT outsourcing statistics 2026 in the shape of a single headline number, that absence is the finding. What exists instead are two adjacent measures. This page uses those, clearly labeled.
Two more scope notes. This is a global page. For country-level developer-population data, our Philippine developer statistics study is the companion piece, and the Philippine figures here appear only as destination context. Want the concept explained rather than the numbers? Our staff augmentation guide covers that ground. Software outsourcing statistics specifically (application, engineering and embedded work) sit inside the deal-flow and AI sections below rather than in a table of their own.
Two organizations publish figures adjacent to outsourcing with methodology attached. Neither one is the outsourcing market, and the difference matters more than the digits.
Gartner measures IT services spend. Its July 2026 forecast puts the Services segment at $1,570 billion in 2026, up 5.3% from $1,492 billion in 2025 [N2]. That is spend on IT services in total, including a great deal of work nobody ever outsources. Gartner's method rests on "rigorous analysis of the sales by over a thousand vendors across the entire range of IT products and services," supported by primary research [N4]. For context, total worldwide IT spending is forecast at $6.37 trillion in 2026, up 14.2% [N1], with services the slowest-growing major segment other than communications: data center systems are up 62.5% to $822 billion, software up 15.5% to $1,468 billion, IaaS up 29.3% to $287 billion [N3]. The money is moving to infrastructure and software, not to services.
ISG measures contract value. The ISG Index counts commercial outsourcing contracts with annual contract value of $5 million or more [N12], and has done so for 95 consecutive quarters [N40]. Its Q2 2026 combined-market figure was a record $42.4 billion, up 43% year over year [N13].
Three quantities, three different things. IT services spend is what buyers spend, outsourced or not. Contract value is what was newly booked in large deals during one quarter, not revenue recognized. And outsourcing revenue, the number everyone actually wants, is neither. Adding them, comparing them or converting between them produces nonsense. ISG's own 2026 revenue growth forecast for managed services is 2.1% [N24] — an order of magnitude below its bookings headline, from the same firm, in the same release.
It measures newly booked contract value across the combined market, and the growth inside it is cloud consumption. The 43% headline is the most misread number in this market, so start by taking it apart.
That growth is not outsourcing. Managed services grew 2.7% to $10.9 billion in Q2 2026, while as-a-service soared 65% to a record $31.5 billion [N14]. Inside that, IaaS surged 78% to $25.8 billion and SaaS rose 25% to $5.7 billion [N15]. ISG chief AI officer Steve Hall adds a caveat worth quoting: "A meaningful portion of current infrastructure spending is coming from a relatively small number of frontier AI model providers, including companies like OpenAI and Anthropic, which are purchasing enormous amounts of compute capacity" [N16]. The record is cloud compute, bought by a handful of labs.
Within outsourcing proper, service lines move in opposite directions. In Q2 2026, ITO declined 3% to $7.7 billion while BPO rose 34% to $2.3 billion, and ER&D fell 6% to $931 million even as its deal volume rose 34% [N21]. Across the first half: ITO down 5.6% to $15.5 billion, BPO up 47% to $4.8 billion, ER&D down 2.8% to $1.8 billion [N23]. "Outsourcing is growing" is not a sentence that survives contact with this data.
Deal counts held roughly flat: 725 managed services contracts in Q2, up 3% [N17], with seven mega-deals of $100 million or more, down from eight, their combined ACV off 27% [N18].
Then the finding that should reframe the whole page. New-scope ACV hit a record $8.2 billion, up 14%, with the number of deals signed representing nearly 75% of all managed services bookings in the quarter [N19]. Read that carefully: the 75% attaches to deal count, not to booking value. Hall's reading: "Enterprises are actively reshaping their sourcing portfolios. Much of this activity appears to reflect work moving between providers and changing operating models rather than entirely new outsourcing demand" [N20]. Three quarters of the bookings are churn and restructuring, not appetite.
Deloitte's Global Outsourcing Survey, drawing on more than 500 executives globally in its 2024 edition [V1], is the standard citation for motive. Read it as stated intent and self-assessment, not measured behavior.
On direction of travel, 80% of executives say they plan to maintain or increase investment in third-party outsourcing [V2]. Cost has lost its monopoly as a motive: skilled talent and agility now rank alongside cost reduction as key drivers [V3]. Scope has widened too, with 50% using outsourced services for front-office capabilities such as sales, marketing and R&D [V7]. The machinery for running all this is thinner than the ambition. 70% report their vendor management function is not fully mature [V8].
Most statistics pages stop there, which is how they get the story wrong. The same survey found that 70% of executives have selectively insourced scope previously held by a third party over the last five years [V5]. Seven in ten. That's the closest thing to a credible disappointment signal in any of this data, and it is not a failure rate; it is a record of scope coming back.
The captive center is the other half of the picture. 78% of surveyed organizations use Global In-house Centers today [V6]. ISG's GCC study, covering 200 firms that recently set up a GCC or plan to, found 90% intend to increase the scope of work handled by their GCC within the next 12 months [N31]. The work they name most often: finance and accounting operations (31%), technical support and IT operations (27%), customer service (23%) and software development (22%) [N32]. ISG also charts GCC headcount growing faster than IT services headcount from 1Q23 to 1Q26, though the GCC series is India-only against a global IT services series [N7].
Captives are a substitute for outsourcing, not evidence of it. Anyone selling you GCC growth as proof that outsourcing is booming has the arrow pointed the wrong way.
This is where survey opinion and measured decline sit side by side, and they need to be kept apart.
The opinion: 83% of Deloitte respondents say they are using AI as part of their outsourced services, while only 25% report reductions in vendor service costs or improvements in service quality [V4]. Both halves are self-reported. Only 20% say they are developing strategies to manage digital workers [V9], which is a reasonable proxy for how early the governance work is.
The measurement: ISG recorded application development and maintenance under pressure [N26] and ER&D ACV down in Q2 [N21], with software engineering down 27% and embedded engineering down 23% across the first half [N38]. Hall's explanation of those declines is that "traditional labor-intensive work is increasingly displaced by LLMs," with providers facing "pricing deflation and more provider-funded, AI-powered transformation embedded within contracts" [N25]. He's more specific elsewhere: LLMs are displacing human labor "especially in areas like software engineering and embedded engineering" [N26]. The declines are measured. The causation is his reading of them, and worth weighing as such.
ISG's own market framing is blunt: "AI is changing the objectives and the economics of outsourcing," and "AI-led deflation and services alternatives keep pressure on labor-based services" [N27]. Its AI Index scores managed services at −4.7% on economic proof since the December 2022 inflection point, against IaaS at +182% and SaaS at +66% [N28]. For managed services it records "Rev per employee is improving, but profitability is under pressure," with investment proof marked "STILL IN TEST" [N29].
The boardroom conversation has changed shape too. AI discussions on earnings calls rose more than sixfold since the ChatGPT inflection, and the tone has moved toward commercial discipline: cost, risk and governance went from 7% of the discussion in 2025 to 11% in 2026, while execution and commercialization fell from 58% to 53% [N30].
India, the United States and Poland are the top three GBS locations and have consistently stayed there [V11]. That ranking measures stated preference, not volume of work delivered.
It comes from Deloitte's Global Business Services Survey, the most current location ranking available. This is its 14th installment, fielded Q3 to Q4 2024, drawing on more than 2,000 respondents over eight years across 30-plus countries [V10]. It's a different instrument from the outsourcing survey above, with a different sample and a different date.
Mexico has risen among the top three preferred locations on technology and talent availability, scalability and cost, and Portugal has entered the top 10 [V12]. Roughly 55% of organizations with a global GBS leader role report over 20% average savings [V13], 50% plan to increase their footprint, and about 58% have started or plan to start on GenAI [V14].
By region, ISG's managed services ACV for the first half of 2026 ran Americas −6%, EMEA +13%, Asia Pacific +10% [N35]. The Americas decline is the first time first-half ACV has trended negative since 1H19 [N35]. Demand by vertical in Q2 was led by consumer packaged goods at +59%, transportation +28%, energy +21% and BFSI +16.5%, while manufacturing fell 11%, its third straight quarterly decline [N36].
On the Philippines, treat the numbers as destination context rather than a developer supply signal. The IT-BPM sector closed 2025 with export revenues above $40 billion, up 5% [V15], and IBPAP has set a 2026 baseline projection of $42 billion [N8]. The sector accounts for roughly 8% of Philippine GDP [N11] and hosts around 160 GCCs, second only to India in that segment [N9]. The work is overwhelmingly contact-center, and the headcount figure of about 1.97 million is a projection for end-2026 rather than a current count [V16]. Developer-population data lives in our Philippine developer statistics study, not here, and the practical mechanics of offshore development in the Philippines are covered separately.
Where the work lands determines what you have to sign. Under EU law, personal data can flow to an adequate third country "without any further safeguard being necessary," with transfers treated as equivalent to intra-EU ones [V18].
The current adequacy list covers Andorra, Argentina, Brazil, Canada (commercial organizations), the Faroe Islands, Guernsey, Israel, the Isle of Man, Japan, Jersey, New Zealand, the Republic of Korea, Switzerland, the United Kingdom, the United States (for EU-US Data Privacy Framework participants), Uruguay and the European Patent Organisation [V17].
Two of the largest delivery destinations on this page, India and the Philippines, are not on it.
That is a contracting fact, not a disqualification. Absence of adequacy does not block transfers: Standard Contractual Clauses ensuring appropriate safeguards remain a lawful ground for transferring data from the EU to third countries, with the modernized set issued in June 2021 [V19]. In practice, a vendor in a non-adequate country should arrive with executed SCCs, a transfer impact assessment and named sub-processors, and shouldn't be surprised you asked. If procurement discovers this at contract stage rather than shortlist stage, you've already lost weeks.
Attrition is the number that decides whether the team you met at kickoff is the team you have in month nine. For the wider hiring picture these rates sit inside, see our developer hiring statistics.
Aon's 2025 Salary Increase and Turnover Study covers more than 700 businesses across Indonesia, Malaysia, the Philippines, Singapore, Thailand and Vietnam, fielded July to September 2025 [V21]. Its 2025 attrition actuals: Philippines 20.0%, Singapore 19.3%, Malaysia 18.2%, Thailand 17.2%, Indonesia 15.0%, Vietnam 15.0% [V20].
Read that table precisely, because it is routinely misread. These are 2025 actuals, not 2026 projections — Aon's 2026 projections in the same study concern pay, not turnover. They are all-industry figures, not tech-specific, and technology roles in these markets are not necessarily at the all-industry rate in either direction. Anyone quoting "20% projected Philippine attrition in 2026" is reading the wrong column.
This is the section to keep. Treat each item below as something to refuse when it appears in a vendor deck.
Global outsourcing market size. No methodology-backed figure exists in public form. Any deck leading with "$X billion by 2030" should be asked for the sample, the method and the analyst's name. The honest substitutes are Gartner's IT services spend [N2] and ISG's contract value above the $5 million line [N12], each labeled for what it counts.
Outsourcing failure rates. No methodology-backed failure, dissatisfaction or early-termination rate was found. Every circulating "X% of outsourcing projects fail" number should be treated as folklore until someone produces the study. The defensible things you can say about disappointment are narrower and better: 70% have selectively insourced scope in five years [V5], only 25% see actual gains against 83% AI adoption [V4], and nearly 75% of bookings by deal count are new-scope rather than new demand [N19] [N20].
Anything under $5 million. The ISG Index sees nothing below $5 million ACV [N12]. Mid-market work and staff augmentation are structurally invisible in the best dataset in this market, which is awkward, because that is closer to what most people reading this page actually buy.
Current destination rankings. The best-known one, Kearney's Global Services Location Index, is still on its 2023 edition as of August 2026 [N41].
Cross-source regional comparison. Gartner reports global segments, ISG uses Americas/EMEA/Asia Pacific, Deloitte ranks individual countries, Aon covers six named Southeast Asian markets. Charts that blend them are decoration.
Strip out the cloud growth and the IT outsourcing trends in 2026 are consistent across sources. Traditional IT outsourcing is flat to declining, with ITO down 5.6% across the first half of 2026 [N23]. Most of the deal activity, by count, is work changing hands rather than new work appearing [N19] [N20]. Seven in ten buyers have pulled scope back in-house at some point in five years [V5]. Nearly eight in ten run captive centers [V6], and nine in ten of the recent GCC cohort want those centers doing more [N31]. Meanwhile the work most exposed to LLMs — application maintenance, software engineering, embedded engineering — is the work declining fastest [N26] [N38].
That's not a market in decline. It's a market where the labor-arbitrage model is being unbundled and the buyers who do well are the ones treating outsourcing as a capability decision rather than a rate-card decision. That decision starts with picking honestly between the software development staffing models available to you. It's also the argument for building small, senior, accountable teams instead of large cheap ones — the model behind our dev on demand service.
No methodology-backed global market-size figure for IT outsourcing exists in public, fetchable form. The two credible adjacent measures are Gartner's IT services spend, forecast at $1,570 billion for 2026, up 5.3% from $1,492 billion in 2025 [N2], and ISG's contract value, which counts only commercial outsourcing contracts of $5 million or more in annual contract value [N12]. Neither of those is the outsourcing market, and they cannot be added together or compared with each other.
Traditional IT outsourcing is flat to declining. ITO contract value fell 5.6% to $15.5 billion across the first half of 2026, while BPO rose 47% to $4.8 billion and ER&D fell 2.8% to $1.8 billion [N23]. ISG's record $42.4 billion combined-market quarter [N13] came from as-a-service, which grew 65% to $31.5 billion, while managed services grew 2.7% to $10.9 billion [N14]. ISG's own 2026 revenue growth forecast for managed services is 2.1% [N24].
No methodology-backed failure, dissatisfaction or early-termination rate for IT outsourcing was found in public form, so every circulating "X% of outsourcing projects fail" figure should be treated as folklore until someone produces the study. The defensible substitutes are narrower: 70% of executives in Deloitte's 2024 Global Outsourcing Survey had selectively insourced scope previously held by a third party over the last five years [V5], and new-scope work accounted for nearly 75% of Q2 2026 managed services bookings by deal count rather than new outsourcing demand [N19] [N20].
70% of executives in Deloitte's 2024 Global Outsourcing Survey said they had selectively insourced scope previously held by a third party over the last five years [V5]. That is a record of scope coming back, not a measured failure rate, and it is self-reported. In parallel, 78% of surveyed organizations run Global In-house Centers [V6], and 90% of the recent GCC cohort in ISG's study intend to increase the scope of work their GCC handles within the next 12 months [N31].
83% of Deloitte respondents say they are using AI as part of their outsourced services, while only 25% report reductions in vendor service costs or improvements in service quality; both halves are self-reported [V4]. ISG measured software engineering contract value down 27% and embedded engineering down 23% across the first half of 2026 [N38]. ISG chief AI officer Steve Hall attributes those declines to "traditional labor-intensive work is increasingly displaced by LLMs" [N25]. ISG's AI Index scores managed services at −4.7% on economic proof since the December 2022 inflection point, against IaaS at +182% and SaaS at +66% [N28].