Industry

Figures converted from INR at historical FX rates — see data/company.json.fx_rates. Ratios, margins, and multiples are unitless and unchanged. Industry-level data (NASSCOM, peer financials reported in USD) are quoted as published.

The Arena: India's IT-BPM Industry, the BPM/KPO Slice Where eClerx Lives

eClerx is not a generalist Indian IT services firm. It sits in a narrower, more specialised pocket of the India-led IT-BPM industry — the Business Process Management (BPM) and Knowledge Process Outsourcing (KPO) layer — where global enterprises hand over judgment-intensive operations (trade settlements, KYC, financial-crime checks, marketing operations, customer support, finance & accounting, digital-shelf analytics) to specialised offshore providers. That arena rides on three structural forces investors must internalise before reading anything else about the company: (a) a US$282.6 billion Indian tech-services machine that grew 5.1% in FY2025 and is now the world's primary back-office for the Fortune 2000 [1]; (b) a structurally labour-intensive cost stack of ~60% wages where ~40-50% offshore cost arbitrage is the original moat; and (c) a once-in-a-generation re-architecture under Generative and Agentic AI that is rewriting both pricing models and headcount economics in real time.

The headline numbers below frame eClerx's place in that arena.

India IT-BPM industry FY2025 (US$ B)

282.6

Industry YoY growth FY2025

5.1%

Industry workforce (millions)

5.8

eClerx FY2026 operating revenue (US$ M)

469

Industry context drawn from NASSCOM data cited in eClerx's FY2025 management discussion; the technology sector added 126,000 net new positions in FY2025 to reach approximately 5.8 million employees [2]. eClerx itself crossed US$469 million in operating revenue in FY2026, up 17.9% in dollar terms — placing it firmly in the mid-cap Indian BPM peer group [3].

What this industry actually does (a primer)

The Indian Information Technology Services classification is a wide tent. Inside it sit two very different businesses that investors routinely conflate: traditional IT services (custom software, application maintenance, infrastructure outsourcing — Infosys, TCS, Wipro country) and BPM/KPO — the management of processes, not codebases. eClerx is firmly in the second camp.

The original framing remains useful and comes straight from the company's own IPO prospectus: Business Process Outsourcing (BPO) is the outsourcing of "routine and standardized business processes that are to be carried out based on pre-determined and laid down rules"; Knowledge Process Outsourcing (KPO) is the layer above it, where the central theme is "to create value for the clients by providing business expertise rather than process expertise" — work that demands "advanced analytical and technical skills as well as decisive judgment" [4]. KPO and BPO together make up the IT enabled Services (ITeS) industry — IT-intensive processes delivered over telecom networks from a location distinct from the user [4]. The contemporary industry name BPM (Business Process Management) is the rebranding under which providers have moved up the value chain from voice-driven contact centres into analytics, automation, compliance, and increasingly AI-orchestrated operations. eClerx describes itself as a "productized services company" employing over 19,000 people globally and serving Fortune 2000 enterprises across financial markets, digital, customer operations, and technology [5].

Why BPM matters more than the IT-services average: the BPM slice is roughly one-fifth of the Indian tech industry by revenue but is more cyclical, more wage-sensitive, more concentrated in specific verticals (BFSI, telecom, healthcare, retail), and more directly exposed to the AI productivity question. Reading a BPM report by the playbook of a TCS or Infosys note will mislead you on every important dimension — pricing, margins, attrition, capex, and AI risk.

Where the work happens — and why "seats" matter

Offshore delivery still defines the economics. eClerx's offshore facilities have a capacity of around 14,700 seats spread across India, USA, Philippines, Egypt and Peru as of March 2025 [6] — up from ~10,700 India seats in March 2022 [7] and ~12,400 across India and Philippines in March 2024 [8]. That seat-count, the offshore voluntary attrition rate, and utilisation are the operational triumvirate that drives margin in this industry — and the three KPIs every analyst on every quarterly call returns to.

The arc of the Indian IT-BPM industry, by the numbers

The single best long-form way to understand the cycle of this industry is to read NASSCOM's annual strategic reviews as paraphrased in eClerx's own management discussion year over year. Doing so reveals a clear three-year deceleration since the post-COVID FY2022 boom — followed by the AI inflection point reshaping the FY2026 outlook.

The FY2022 record explicitly called out the moment the industry "crossed $200 Bn in total revenue and 5 Mn in total workforce" — with the BPM sector specifically logging $44 billion at "the highest-ever" growth rate since 2011 of 15.5% [9]. By FY2023 the industry was at ~$245B with digital revenue at 34% of the total and workforce at 5.4M [10]. FY2024 saw growth cool to 3.8% YoY at $254B as the industry added only 60,000 net employees in a "creditable performance" against macro and geopolitical headwinds; exports crossed the $200B mark while growing just 3.3% [11]. FY2025 then saw a re-acceleration to 5.1% growth with $13.8 billion in incremental revenue and the workforce returning to net hiring [1].

The forward indicator that matters most is the NASSCOM CEO survey: 77% of tech industry leaders expect business growth, 85% expect client tech budgets to be stable or higher in FY2026, and nearly two-thirds expect AI investments to exceed 10% of total technology expenditure [12]. That last data point is the most consequential structural signal in this industry today. Money is flowing into a different kind of spend pattern than the headcount-augmentation paradigm BPM was built on.

How providers make money: the value chain, the price card, the cost stack

The historical price card

Pricing in this industry has always varied an order of magnitude across the BPO-to-KPO continuum. eClerx's own 2007 IPO prospectus reproduced a CRISIL table of average billing rates that — even though the absolute dollar levels have drifted — still anchors the relative ordering of work-types today:

No Results

Source: CRISIL Annual Review of IT enabled Services, June 2007, as reproduced in eClerx's draft red herring prospectus [13]. The dollar amounts are dated; the shape of the curve — voice/F&A at the bottom, KPO analytics/legal/research at the top — still maps directly onto how eClerx and its peers position themselves today. Moving "up the curve" from voice contact-centre to financial-crime-compliance, KYC, advanced analytics and product-data engineering has been the multi-decade strategic vector of every successful Indian BPM firm.

For historical context on where Indian ITeS came from — and to put today's $282.6B headline in perspective — the IPO prospectus also reproduced NASSCOM's ITeS revenue progression from FY2002-2006:

From ~$1.6B at the turn of the millennium to ~$7.2B by FY2006 [14], to ~$282B today. The trajectory is one of the great industrial-services growth stories of the last 25 years — and the question facing investors in FY2026 is whether AI breaks that trajectory or extends it.

The cost stack — why this is a wage business

Profitability in BPM is mostly a function of one number. In FY2025, employee benefits expense was 59.84% of total revenue for eClerx, with cost of technical sub-contractors at 2.36% and other expenses at 11.89% — leaving an EBITDA margin of 25.91% [15]. The whole industry runs on roughly the same arithmetic. Move the wage line by 200 basis points and the EBITDA line moves the same. The IPO prospectus called this out a generation ago — "wage increases in India may prevent us from sustaining our competitive advantage and may reduce our profit margin" — and it remains the single biggest controllable margin driver [16].

The demand mix: which verticals matter most

The interesting feature of modern Indian BPM is not the size of the addressable market but the concentration within it. Across the listed peer set, four verticals dominate: BFSI (banks, capital markets, insurance), CMT (communications/media/telecom), high-tech/manufacturing, and retail/F&L. A fifth "emerging" category (predominantly F&A automation served to SMBs) is growing faster than the others.

eClerx's mix — updated through Q4 FY2026 — is illustrative of where a mid-cap BPM player sits in 2026:

Source: eClerx Q4 FY26 investor presentation [17]. Two structural shifts worth a closer look: BFSI is shrinking as a share (43.3% → 40.8%) even though it is growing absolutely, because Emerging is growing at a far faster clip (5.2% → 8.4%, driven by Personiv F&A serving SMBs). Fashion & Luxury continues to be the soft spot — explicitly called out by management as a sluggish vertical in FY2026 with green shoots expected in H1 FY2027 [3].

These vertical mixes are not the same across peers. Sagility is overwhelmingly healthcare-focused; ExlService is heavily insurance and BFSI; Genpact is broadest with Financial Services, Consumer/Healthcare, and High-Tech & Manufacturing as three named segments; Firstsource leans healthcare and BFSI; Hinduja Global is more contact-centre/voice heavy. Vertical concentration matters because the underlying demand cycles are different: BFSI follows global bank regulatory cycles (a known driver eClerx's CEO has explicitly tied to FCC growth) [3]; CMT follows telco capex and US regulatory politics around offshoring; healthcare follows US payer/provider cycles.

Where the cycle sits right now

Three operational KPIs reveal the live state of the cycle better than any P&L line. They are the metrics every BPM CFO opens with on the quarterly call: client concentration, utilisation, attrition, plus the ACV (Annual Contract Value) of new deal wins as the leading indicator.

eClerx Top-10 client concentration

59%

Q4 FY26 offshore utilisation

74.2%

Q4 FY26 offshore voluntary attrition

21.7%

Q4 FY26 new deal ACV (US$ M)

Concentration. Top-10 client concentration for eClerx has come down to 59% from 63%-64% — a healthy sign of portfolio diversification that reduces concentration risk as the company scales [18]. The FY2025 annual report disclosed 63% of revenue from the top 10 clients [19]. Industry-wide, very high client concentration is normal in mid-cap BPM and is a permanent risk factor — large clients can in-source, switch providers, or be acquired, any of which can move the revenue line meaningfully.

Utilisation. Q4 FY2026 utilisation was 74.2%, down from an "eight-quarter high" of 76.5% in Q3 FY2026 [20]. This 70-77% band is the industry's normal operating range — peers run similar numbers. Utilisation moves up when delivery teams are full and pricing is firm; it moves down when management is investing ahead of demand or when discretionary work pulls back.

Attrition. Q4 FY2026 offshore voluntary attrition was 21.7%, up marginally from Q2/Q3 but well below the cycle peak of 28.8% seen in Q2 FY2025 [21]. The IPO prospectus a generation ago listed attrition as the industry's number-one challenge — "despite the availability of trained manpower in the industry, there continues to be a lot of movement both within and across companies, thus driving up attrition rates" [22]. It remains true. Anything sustained above 25% is a cost and quality red flag; anything sustained below 18% suggests demand has cooled. The IPO prospectus also frankly noted that the work is "labour intensive" and success "depends in large part upon our ability to attract, hire, train and retain qualified employees" [23].

Deal wins (ACV). Q4 FY2026 new-deal ACV (excluding the CLX fashion subsidiary) was US$46.1 million [24] — the high end of the historical range. ACV is the closest thing this industry has to an order-book metric. A range of ~$20M/quarter to ~$45M/quarter for a mid-cap player like eClerx is normal; sustained moves up or down tend to lead revenue growth by two to three quarters.

Billing mix: T&M, BPaaS, onshore

Most BPM revenue is still billed on a Time & Materials basis — for eClerx, T&M was approximately US$280 million of the ~US$284 million standalone revenue in FY2025, with fixed-price work at just ~US$4 million [25]. The transition the industry is now living through is the shift toward BPaaS (Business-Process-as-a-Service — platform + outcome-linked pricing) and outcome-based contracts. eClerx discloses its BPaaS share of revenue explicitly: it has hovered around 18-21% over the last eight quarters, with onshore revenue running another 18-21% of the mix [26]. That billing-mix evolution is one of the most important things to watch over the next three years — a higher BPaaS share theoretically decouples revenue from headcount and is the operational expression of "AI-led productivity." It is also a structural margin question. If BPaaS scales without proportional cost-base reduction, gross margins compress; if AI compresses the cost-base faster than the price-base, margins expand.

Geography and currency: an export industry

The Indian BPM industry is, structurally, an export business serving the United States and Western Europe. For eClerx in FY2024-25, 92% of revenue originated in the US and Western Europe [27] — and 86% was billed in US Dollars, 9% in Euros, and 5% in Sterling and other currencies [28]. That geographic and currency concentration is broadly true of every Indian peer except those whose mix is more domestic-India. Two consequences:

The industry's growth tracks US corporate technology spending and Western European banking budgets far more than Indian GDP. Read the JPMorgan, Bank of America, Verizon, Comcast and Adobe budgets, not the Reserve Bank of India outlook.
FX matters in both directions. A weaker INR lifts reported INR revenue and margins (because costs are largely INR-denominated and revenue is USD-denominated); a stronger INR compresses them. Most Indian BPM firms hedge a large share of their forward USD receivables — eClerx hedges ~80% of USD receivables, with FY2027 hedges sitting at an average of INR 89.89/USD [29]. The hedge book caps both upside and downside from FX moves.

Competitive structure: a fragmented, mid-concentration industry — with one important caveat

The Indian BPM/KPO industry is genuinely fragmented. Even at the top, there is no single player with 20%+ share of BPM exports. The biggest pure-plays in eClerx's competitive set look like this:

No Results

Sources: per-peer financial snapshots from latest published filings. INR figures translated to USD at FY-end period rates (FY2025 ~₹85/US$, FY2026 ~₹88/US$).

The bubble chart below illustrates the return-on-equity vs. net-margin dispersion across the peer set — and shows why "Indian BPM" is misleadingly homogenous as a label:

Three things jump off this chart:

eClerx, EXLService and Alldigi cluster in the high-margin, high-return quadrant. All three are "knowledge-heavy" BPM rather than voice-heavy.
Genpact is the volume leader by an order of magnitude. Its ~11% net margin reflects scale at the cost of premium-segment economics.
Hinduja Global Solutions is the cautionary tale of the voice-heavy, undifferentiated end of the market — its 0.1% net margin in FY2026 is a quiet but important reminder that "BPM" includes both well-run franchises and structurally challenged ones.

A peer-set caution worth flagging plainly

The peer set indexed in this run was auto-selected by a screen for India BPM exposure and market-cap proximity. It is not a perfect match for eClerx's business model. Sagility is essentially a healthcare-only payer/provider BPM (a vertical eClerx barely touches). Hinduja Global is voice-heavy contact-centre work. Alldigi (formerly Allsec) is a small-cap with a customer-and-employee-experience focus. The most directly comparable peers — by service mix, by vertical concentration in BFSI and digital, and by the move into analytics and AI-orchestrated operations — are Genpact, EXLService, and Firstsource. Use the broader six-company peer set for benchmarking labour-cost discipline and revenue-growth; use the narrower three-company set for benchmarking BFSI/analytics economics and pricing power.

eClerx's own competition disclosures repeatedly underscore that its arena is "highly competitive" — including being named "a major contender" or "aspirant" across multiple Everest Group PEAK Matrix assessments in Customer Experience Management, RPA Products, F&A Outsourcing, Marketing Services and BFS Operations [30]. The company also explicitly described the Customer Operations vertical as competing in a "highly competitive landscape in the BPO industry" [31].

The AI inflection: deflation, productivity, or both?

Of every theme an industry primer for an Indian BPM company could cover in 2026, this is the one that defines the next five years. The arena that was built on offshoring labour at $10-15/hour is being re-architected by generative and agentic AI in real time. Three things are happening simultaneously:

The mechanical case for headcount augmentation is weakening on the lowest-end work. Coding is being democratised (eClerx's CEO: "people who understand the domain can write the prompts, build the code"). Customer-service voice volumes are coming down. eClerx itself has noted "we have seen the call volumes, chat volumes coming down" but argued it has captured share at the same time [32]. This is consistent with a market that is shrinking in unit volume but consolidating into fewer, better-equipped providers.
The opportunity for outcome-based and BPaaS pricing is expanding. eClerx flagged its first large-scale Agentic AI win in Q4 FY26, with deployments planned for Q1 FY2027 [33], built its analytics and automation business to a US$90 million annual run-rate [33], and now has an outcome-linked pricing model in products like Compliance Manager (where it has "reduced the cost of refresh, let's say, by 50% because we have brought in Agentic AI") [34].
The natural 15-20% annual "roll-off" of short-term projects is unchanged for now. Management has been emphatic that AI deflation is not meaningfully accelerating client roll-offs in their book — "we don't think that number is going to meaningfully change because AI is another productivity tool, and it takes time to implement" [35]. Whether that view holds for the full industry is the single most important debate in the BPM space today.

The CEO's framing — that "service providers who can bring it all together [agentic, tech, domain, IP, process, headcount] and deliver to an outcome" will win, and those that "look at these in discrete manner" will not — is a useful frame for separating winners from losers in the BPM AI transition [36].

Industry growth has not yet been impaired. The NASSCOM CEO survey continues to point to client tech budgets being stable or higher in FY2026 (85% of respondents), with two-thirds of executives expecting AI investment to exceed 10% of overall tech spend [12]. If AI compresses pricing per unit but expands the addressable spend per client by even more, BPM revenue keeps growing — possibly at higher margins. If AI compresses pricing faster than the addressable spend expands, BPM revenue compresses. The honest answer in mid-2026 is that we do not yet know.

Regulation, politics, and structural risks

The industry has four risk vectors that have shaped its history and continue to shape its prospects:

No Results

Three of these merit specific attention because they are live in FY2026:

The NPRM (Notice of Proposed Rulemaking) on offshore call restrictions in the US. eClerx's CEO flagged on the Q4 FY26 call that the company is "closely monitoring the proposed NPRM on offshore call restrictions and maintaining active dialogue with clients on contingency planning" — with NASSCOM "lobbying with the U.S. Congress" against rules that could require telcos and cable companies to keep a certain percentage of work onshore [37]. This is a non-trivial political risk for the CMT vertical specifically. The IPO prospectus identified "political opposition to offshore outsourcing in the United States and other countries" as a foundational risk a generation ago [38]; it has not gone away.

BFSI regulatory cycle dependency. The CEO has been explicit that financial-crime-compliance growth at eClerx "is also driven from a regulatory environment. So, like in terms of if regulators increase or decrease what they want of the banks and financial institutions, that can have an effect on the overall spend." [39] The change of Fed chair is explicitly cited as a watch-item. Industry-wide, BFSI BPM demand is meaningfully geared to the regulatory cycle — KYC refresh requirements, AML thresholds, capital-markets rules.

Geographic concentration of revenue in the US and Western Europe (92%) combined with currency concentration in USD (86%) [27] means the industry is structurally exposed to American macro conditions and the INR/USD cross. There is no domestic-Indian demand cushion of any meaningful size.

The KPIs an investor should watch

Eight numbers are enough to monitor whether the Indian BPM industry — and a company within it — is in a healthy cycle:

No Results

Glossary — eight terms an investor needs to know

No Results

Bottom line for the report

The investor reading the rest of this report should hold five things in mind:

eClerx is a knowledge-heavy BPM/KPO mid-cap, not a generalist IT services firm. Read its numbers in the BPM peer set, not the TCS/Infosys peer set.
The Indian IT-BPM industry is a US$282.6B export business growing low-to-mid single digits, with the BPM sub-sector running at higher absolute rates and the AI inflection rewriting both unit economics and pricing models simultaneously.
Wages are 60% of the cost stack — every margin conversation begins and ends with this.
The cyclical KPIs to watch are concentration, utilisation, attrition, and ACV bookings. They lead the P&L by one to three quarters.
The AI question is genuinely binary at the industry level. Management's framing — that integrated AI-domain-process providers win and discrete pure-plays lose — is a useful operating thesis, but the verdict will come from the next two to three years of pricing and bookings data, not from analyst Q&As today.

References

Generated Jun 20, 2026 00:08 UTC