IT managed services is a $400bn global industry growing at high single digits. The headline number hides three different businesses stacked on top of each other. At the top sit the global IT outsourcers running mainframes, data centres, and applications for the Fortune 500. In the middle sit the mid-market integrators wrapping security, cloud, and helpdesk into recurring contracts. At the bottom sit tens of thousands of regional MSPs serving SMBs on a per-seat or per-device fee.
They all call themselves "managed services". They look almost nothing alike on the income statement.
This primer is for the investor trying to size a position, underwrite an LBO, or short a tier-one outsourcer. The largest players, including IBM, Accenture, Cognizant, TCS, HPE, DXC, Capgemini, NTT Data, Atos, and Wipro, operate at 10–20% adjusted EBITDA on multi-year contracts. The PE-backed mid-market roll-ups operate at 15–25% EBITDA with high recurring revenue and 90%+ gross retention. The bottom of the market operates at single-digit margins with one tech in a van.
Mixing them up is how investors misprice this sector.
The next 5,000 words cover industry structure, unit economics by tier, the companies that matter, the questions a good expert call needs to cover, and the debate that will define returns over the next 36 months: whether AI compresses delivery margins faster than it expands wallet share.
What Is IT Managed Services?
A managed service is the ongoing operation of a customer's IT environment for a recurring fee under a service-level agreement. The provider owns the outcome. The customer pays a predictable monthly or annual amount and stops worrying about whether the email server is up, whether the firewall is patched, or whether the helpdesk picks up the phone at 2am.
This is the line that separates managed services from project-based IT consulting. A systems integrator gets paid to deliver a project and leaves. An MSP gets paid every month for as long as the lights stay on. The model lets enterprises shift from capex to predictable opex, which is why finance teams love it and why the recurring revenue base is so sticky.
The services cluster into five buckets. Infrastructure management covers servers, storage, networks, data centres, and increasingly the cloud accounts customers run on AWS, Azure, and GCP. Application management covers ERPs, CRMs, custom software, and the SaaS estate. Security management covers firewalls, endpoint protection, identity, and the SOC functions that have become the highest-margin part of the industry. End-user services cover the helpdesk, device provisioning, and the people who answer when something breaks. Communication and collaboration covers Teams, Zoom, voice, and the networking that holds it together.
Most customers buy two or three of these from the same provider. The economics are driven by how many of the five a single MSP can sell into a single account. Wallet share is what separates a 5x EBITDA business from a 12x business.
Why It Matters to Investors
The sector matters for three reasons that have nothing to do with how interesting IT plumbing is.
First, it is one of the largest pools of recurring contracted services revenue outside SaaS. The global market is estimated at $401bn in 2025 and projected to grow at roughly 10% CAGR to $847bn by 2033. IT outsourcing alone, the enterprise tier sitting above the SMB MSP market, is $618bn growing more slowly at around 3.3% to $752bn by 2031. The growth differential between the enterprise tier and the mid-market is the single most important thing to understand about the sector. Mid-market MSPs are growing two to three times faster than the global outsourcers.
Second, it is the most active roll-up theatre in private equity right now. There is over $400bn in PE dry powder targeting technology services. The gap between mature platforms commanding 12-14x EBITDA and small sub-scale MSPs trading at 4x has created a structural arbitrage that has run for the better part of a decade. Evergreen Services Group alone completed 47 acquisitions in 2025, taking its cumulative total above 100. Whether that arbitrage holds, narrows, or breaks is the most important question for anyone underwriting a platform investment in this sector.
Third, AI is finally hitting delivery economics in a way that matters. Half the cost base of a typical MSP is people. If GenAI agents and AIOps platforms can automate even 30% of L1 and L2 tickets, gross margins on legacy contracts move 500 to 1,000 basis points. That savings either shows up as multiple expansion for the provider or as price compression demanded by the customer, depending on who has pricing power.
The buyers think they are about to get a discount. The PE-backed platforms think they are about to print money. Both cannot be right.
Industry Structure
The cleanest way to think about the industry is in three tiers that almost never compete directly.
The top tier is the global IT outsourcer. These businesses took on the mainframe estates and global infrastructure of the Fortune 500 in the 1990s and 2000s. They sign 5 to 10 year contracts measured in hundreds of millions of dollars, often with a transition phase where they absorb thousands of customer staff onto their own payroll. Kyndryl, the largest pure-play in this tier after spinning out of IBM in 2021, generated revenues around $15bn for fiscal year 2025 with adjusted EBITDA margin of approximately 18%, up about 130bps year over year. The peer set is IBM, Accenture, Cognizant, TCS, HPE, DXC, Capgemini, NTT Data, Atos, and Wipro. Together they control around 40% of global IT outsourcing spend.
This tier is mature, slow-growing, and structurally challenged. Customers have spent fifteen years trying to renegotiate the original deals down. Kyndryl's revenue is still declining in constant currency because management is deliberately walking away from inherited zero-margin and low-margin third-party content in customer contracts. The pivot is from labour-intensive infrastructure work to advisory, hyperscaler partnerships, and higher-margin consulting. Kyndryl Consult is the part of the business that is actually growing.
The middle tier is the PE-backed national or super-regional MSP. These are companies in the $20m to $500m revenue range serving mid-market enterprises, typically 500 to 5,000 employees per customer. Platform MSPs at this scale operate ConnectWise, ServiceNow, or Autotask as their delivery stack, hold SOC 2 and ISO 27001 certifications, and have professional management teams capable of integrating bolt-on acquisitions. The active PE-backed platforms include Alpine's Evergreen, Oval Partners' New Charter Technologies, PSP Partners' Ntiva, Court Square and Berkshire's Thrive, OMERS' Integris, and Nautic's VC3. This is where the M&A activity is.
The bottom tier is the local MSP serving small businesses. Roughly 25% of these businesses operate at breakeven or lose money. Many were founded by a single engineer who bought a few clients off a friend and never built operating infrastructure. These businesses sell for 2x to 5x EBITDA when they sell at all. Their role in the investment narrative is as acquisition fodder for the middle tier.
The reason this matters is simple. A 9.9% blended market growth rate is misleading. The top tier is growing low single digits. The mid-market is growing 10–15%. The security-led mid-market is growing 20–25%. An investment thesis built on the headline number is an investment thesis built on a number that does not exist in any single sub-sector.
Unit Economics
The economics get clearer the smaller you go.
At the top tier, gross margins on infrastructure outsourcing run 15–25%. The contracts are huge but loaded with pass-through hardware and software. Adjusted EBITDA margins of 18% at Kyndryl scale, with adjusted free cash flow around $550m on $15bn of revenue, give you a sense of the steady-state economics. Free cash flow conversion is mediocre because the contracts require ongoing capex commitments to customer environments.
At the mid-market, the economics are substantially better. A well-run MSP at this tier should generate 50–60% gross profit margin on services as the average, with top performers above 70%, and overall gross profit margin of 65–75% across the blend. Best-in-class adjusted EBITDA hits 19%+, while the median lands at 8–12%. The gap is explained almost entirely by service-line discipline and pricing. Industry benchmarks point to 60–70% gross margin as the target, 10–30% EBITDA margin as healthy operating performance, $100+ revenue per endpoint per month, and 10–20% growth as the goal.
Revenue mix inside a mid-market MSP is the single most important variable in valuation. Pure managed-service gross margins run around 46%. Products run around 26%. Projects run around 19–24% depending on the quarter. Mix shift toward managed services is what drives multiple expansion. Mix toward products and projects holds it back.
Security is where the unit economics get genuinely interesting. Hybrid MSP/MSSP firms reported average monthly recurring revenue of $8,900 per security client in 2024, more than double the revenue generated from traditional IT clients. MSPs offering managed detection and response see roughly 35% higher EBITDA margins than peers, and healthcare-focused MSPs hit 28% EBITDA on HIPAA compliance work. Security is also where customer churn drops to almost nothing because no CIO rips out their SOC provider when the threat landscape is what it is.
Customer concentration and contract length matter as much as margin. Best-in-class MSPs have over 85% of revenue recurring, gross retention above 95%, and average contract length above three years. That combination is what justifies the 11–14x EBITDA multiple at the top end of the M&A market.
The Key Players
The investible universe splits into three groups.
The first is the listed global outsourcers. The reference set in order of scale is Accenture, TCS, Cognizant ($19.4bn revenue, 347,500 employees), Infosys, Capgemini, HCLTech, Wipro, DXC Technology ($14.4bn revenue, 130,000 employees), Kyndryl ($6.3bn revenue, 90,000 employees), and Unisys ($5.4bn revenue, 16,400 employees). IBM is the elephant in the room because most of its services revenue now flows through Red Hat and the consulting business rather than what would have historically been called outsourcing. These are macro and IT-spend correlated stocks. The thesis is on contract pipeline, AI-led productivity, and offshore mix.
The second is the listed cybersecurity and infrastructure plays adjacent to managed services. AT&T Cybersecurity, Secureworks, the managed-security divisions of CrowdStrike and Palo Alto Networks via their partner ecosystems, and the listed European players like Atos and Capgemini sit here. Security-specialist MSSPs captured 32.1% of revenue in 2025 while hyperscaler-aligned MSSPs grew at 13.98% CAGR. The investment debate is whether the specialists hold their margin advantage as hyperscalers bundle security into the cloud consoles.
The third is the PE-owned mid-market consolidators. The active platforms include Evergreen Services Group (Alpine Investors), New Charter Technologies (Oval Partners), Ntiva (PSP Partners), Thrive (Court Square Capital and Berkshire Partners), Integris (OMERS Private Equity), and VC3 (Nautic Partners). A newer category of AI-native rollups, backed by venture capital rather than buyout PE, includes Shield Technology Partners (Thrive Capital and ZBS Partners) and Titan (General Catalyst). Shield is funded with over $100m from ZBS Partners and Thrive Holdings, has acquired seven MSPs through end of 2025, and is targeting double that by end of Q1 2026. OpenAI took an equity stake in Thrive Holdings in December 2025.
The PE platforms are not publicly listed but they are the most important part of the sector for hedge funds. They are the marginal buyer setting the price for the public outsourcers (who compete for the same customers) and for the bolt-on supply (which is finite).
The AI Disruption Debate
This is the section that will date fastest. It is also the section that matters most.
The bull case for the managed services industry is that GenAI agents and AIOps platforms automate 30–50% of L1 and L2 helpdesk work, level 1 SOC triage, and routine infrastructure monitoring. Labour cost drops 20–30% on the affected workflows. Automating 50% of L1/L2 support tasks can reduce labour costs by up to 30%, dropping directly to EBITDA, and AI-driven RMM platforms can deliver 22% higher gross margins than peers without them. The MSP keeps the spread because the contract is priced per seat or per device, not per ticket. Margins expand. Multiples expand. The PE thesis works.
The bear case is that the customer figures this out. Gartner projects at least 40% of enterprise SaaS spend will shift to usage-, agent-, or outcome-based models by 2030, with seat-based revenue share declining from 21% to 15%. SaaS companies with heavy per-seat exposure saw multiple compression in late 2024 and 2025 as analysts flagged "AI seat risk", and when net revenue retention drops, multiples compress quickly. If managed services pricing moves from per-seat to per-resolution or per-outcome, the labour savings get passed straight back to the customer. The MSP becomes a low-margin transaction processor.
The honest answer is that both will happen, on different timelines and in different parts of the market. The mid-market is sticky on seat-based pricing because the customer does not have the procurement sophistication to demand outcome-based contracts. The enterprise tier will move first because that is where the procurement teams live and where the contracts come up for renegotiation every three years. Kyndryl is already explicitly walking away from low-margin signings in favour of project-focused transformative work, which tells you how management at the largest pure-play sees the trajectory.
The PE-backed mid-market roll-ups have a window. How wide it is depends on how fast outcome-based pricing migrates down from the enterprise to the mid-market. The honest range is two to five years. Anyone telling you it is longer is selling something.
The Roll-Up Question
The PE roll-up thesis in managed services has been one of the most successful sector strategies of the last cycle. It is also the strategy most at risk of breaking.
The thesis works like this. Acquire a platform MSP at 8–10x EBITDA. Bolt on smaller MSPs at 4–6x EBITDA. Standardise the tech stack onto a single PSA and RMM platform. Centralise the NOC and SOC. Layer in higher-margin security and cloud services. Cross-sell into the acquired customer base. Exit the combined platform at 12–14x EBITDA five years later. The math is canonical: platform MSPs command 10–14x EBITDA while smaller MSPs trade at 5–7x, and the multiple arbitrage funds the strategy.
It worked. FOCUS Investment Banking, which advises on the majority of MSP M&A, reports a growing imbalance between qualified buyers and mature sellers, creating opportunities for well-positioned platforms. But the supply of integration-ready mid-market MSPs is finite, and the cost of capital has changed.
The macroeconomic environment of 2024 and 2025 has disciplined the market. Cost of capital has stabilised at higher levels, and acquirers can no longer rely on cheap debt to fuel growth. Platform models that prioritise organic growth and operational maturity are outperforming those that accumulated headcount through serial M&A.
Industry insiders see the 2026 market exposing roll-ups that chased scale without building operational sophistication. Stronger platforms will acquire underperforming providers, the bottom of the market will fold or sell, and PE-backed platforms that miss their forecasts will see headcount reduction, service quality decline, and client attrition.
The investment-grade question is which platforms have actually built operating infrastructure and which are essentially holding companies for fifteen disconnected MSPs sharing a logo. The two look identical from a Form ADV. They look very different on a Q3 EBITDA bridge.
What to Ask in a Channel Check
The right Woozle expert call on this sector covers ten to fifteen questions. The order matters because the answers compound.
Start with the customer side. What is the customer's renewal cycle and have they ever benchmarked the contract? Is the pricing per seat, per device, per ticket, or outcome-based? What did the customer outsource that they would now bring back in-house if they could? Who is the second-source provider and what would it take to switch? How much of the contract is pass-through hardware and software versus pure service?
Move to the provider's delivery model. What is the offshore mix, and which roles are offshore? Filipino professional labour reduces cost 60–80% versus US-based hiring, and offshore Indian talent runs similar arbitrage, so the offshore mix is one of the largest drivers of contract margin. How many tickets does an average L1 engineer handle per day, and what percentage are now auto-resolved? What PSA and RMM platform is in use, and how integrated is the security stack?
Then ask about the competitive set. When this contract last went to RFP, who showed up, and who won the technical evaluation versus the commercial one? Where do the global outsourcers lose to the PE-backed mid-market platforms, and where do they win?
Finish with the AI question. It matters most for forward returns. What percentage of the customer's expected savings over the next renewal cycle is being driven by AI productivity assumptions? Has the customer asked for an outcome-based price or a per-resolution price? If yes, how did the provider respond? If no, how long until they do?
Five expert calls across the right cross-section will give you a defensible view in two weeks. That cross-section is customer-side procurement leads, MSP delivery heads, ex-Kyndryl or ex-DXC account leads, MSSP SOC managers, and PE-backed MSP integration directors. Doing this from desk research alone gives you a deck that looks the same as everyone else's.
The Expert Map
Investors looking to do primary research in this sector should be talking to seven distinct profiles.
The first is the customer-side CIO or VP of IT operations who owns the managed services contract. They know what the SLA misses look like, what the provider's true response time is, and what the renegotiation lever is. This is the most important call in the chain.
The second is the procurement lead who priced the last renewal. They know what the market clearing price is and what the alternative bids looked like.
The third is the delivery lead inside the MSP, typically a VP of Service Delivery or a Director of NOC/SOC operations. They know what the true automation rate is and what the unit cost per ticket actually looks like.
The fourth is the account director on the customer side at the MSP. They know what is up for renewal, what is at risk, and where the wallet share is growing or shrinking.
The fifth is the integration lead at a PE-backed platform. They know which acquired companies are actually being onboarded versus which are still operating as standalone businesses on different tech stacks two years post-close.
The sixth is the ex-employee of a large global outsourcer who has moved to a mid-market competitor. They know exactly where the global outsourcers are losing deals and why.
The seventh is the channel partner at Microsoft, AWS, or a major security ISV who sees the managed services partner landscape from the vendor side. They know which MSPs are growing their licensing book and which are not.
A Woozle workstream on this sector will typically run three to seven of these profiles in parallel over a two to four week window.
How Woozle Helps
Woozle Research runs primary research workstreams on IT managed services and the adjacent IT services categories every quarter. Our network covers the active customer base of every public outsourcer, the delivery teams inside the PE-backed roll-ups, the security specialists, and the procurement organisations that price the renewals.
We typically run one of three workstreams on this sector. Single-name channel checks across the customer base of a listed outsourcer to size revenue, renewal risk, and competitive pressure for the next two to three quarters. Diligence support for PE platform acquisitions or recapitalisations, including customer reference calls and synthetic competitive set surveys. Thematic studies on the AI delivery margin question, with structured interviews across customer, provider, and channel-partner sides of the contract.
If you have a position to size or a deal to underwrite in this sector, get in touch. We will scope the work in 48 hours and deliver the first round of interview transcripts inside two weeks.
