Key Takeaways
- Revenue declined 4.8% CC, margin improved: Higher furloughs in the UK and right-shifted Oracle program starts drove the top-line drop, but EBITDA margin still expanded 60 bps to 16.1% on AI-led efficiency and forex tailwinds.
- AI investment is closing the competitive gap: Management says continued AI capability build has eliminated the technology differentiator versus larger players, opening the door to compete for bigger core digital engineering and AI-led deals.
- A deliberate shift to outcome-based contracts is underway: Mastek is proactively moving clients from time-and-material to outcome-based deals offering 15–30% productivity benefits, a short-term revenue headwind meant to build trust and unlock larger engagements.
- The U.S. reset is structurally complete but not yet visible in revenue: The North America leadership overhaul is 100% done and digital engineering capability is now competitive, yet the quarterly run rate has fallen back to pre-acquisition levels; management expects sustainable improvement only in FY27.
- What to watch next quarter: Whether the right-shifted Oracle programs actually start and convert to revenue, any clarity on the unexplained Middle East decline, and the labour code adjustment that is expected but unquantified.
Revenue Decline: What Broke and What Reverses
- Furloughs took half the hit: Furloughs accounted for roughly $2.7M of the Q3 revenue impact, with 70% of that concentrated in the UK and some in U.S. public sector; management calls this normal Q3 seasonality and expects no significant furlough impact going forward.
- Oracle project timing took the other half: 60% of the Oracle business is project-driven, so revenue ramps down when programs go live and ramps up with new wins; several transformation programs went live during Q3 while new program starts right-shifted to Q4, creating an air pocket that should begin to fill next quarter.
- Middle East decline is unexplained: Management acknowledged a sharp decline in the Middle East geography but, when asked directly, began an answer about WAR revenue and was cut off before completing the explanation; the prior quarter’s claim that restructuring was "almost over" now sits against a fresh deterioration with no clarity on cause.
- U.S. revenue has reset lower: The U.S. quarterly run rate has fallen to $20M–$23M, back to pre-acquisition levels from roughly $30M post-acquisition; core fundamentals (leadership, digital engineering capability) have been rebuilt, but the revenue inflection is still at least a couple of quarters away.
Umang Nahata: "I don't have a firm time commitment here. But I think next year would be the year when you start seeing U.S. starting to improve sustainably in our business."
Margin Walk and Labour Code Uncertainty
- Operational efficiency led the margin improvement: EBITDA margin rose 60 bps QoQ to 16.1%, with 43 bps coming from AI-led operational efficiency and 17 bps from forex tailwinds; subcontractor costs also dropped 0.7% QoQ, partly due to ramp-downs in Secured Government Services.
- Labour code impact was neutralized this quarter, but Q4 is uncertain: The labour code change created a 0.7% headwind (INR 6.4 cr), which was fully offset by a one-time credit from leave accumulation reversal at onsite; clarifications from the labour department are awaited in the next 1–2 months, and management has gone conservative by taking the retrospective impact upfront, estimating an ongoing hit of $150,000–$200,000 per quarter pending those clarifications.
- Management is comfortable with a 16.5%–17% EBITDA band: The stated target range is 16.5%–17%, and the company expects to maintain steady bottom-line performance while continuing AI investments and the shift toward outcome-based contracts.
AI Strategy: Services, Not IP
- AI capability is now table stakes for larger deals: Management asserts that continued AI investment has eliminated the technology differentiator gap with larger competitors, enabling Mastek to compete for bigger core digital engineering and AI-led engineering deals that were previously out of reach.
- The IP strategy is deliberately contrarian: Mastek has no intention of building platform IP or monetizing AI tools; the focus remains entirely on service execution and a partnership model with hyperscalers, with any IP created used only as a differentiator for winning deals and demonstrating client success.
- Productivity benefits are being proactively offered to clients: AI-led productivity gains are being passed through across a large number of customers, including UK Public Sector engagements, as a deliberate strategy to create mind space for expansion; management notes that top 30-odd customers still represent a very small wallet share relative to their overall digital spend.
- ROI on AI investments is claimed to be strongly positive: Management states the long-term return on initial capability investments and commercial discounts is "very, very positive," with the overall bottom-line impact expected to become healthier as the strategy matures.
Umang Nahata: "We actually have a very contrasting view to that. We do not want to get into the IP and the product business as far as AI is concerned... I think our focus is executing services, and we will look at continued focus on service execution."
Outcome-Based Contracts: A Deliberate Top-Line Headwind
- The model is shifting from T&M to outcome-based: Mastek closed several deals this quarter moving customers from time-and-material to outcome-based contracts, offering 15% to 30% productivity benefits in exchange; this creates a short-term revenue headwind but is intended to build customer trust and create space for larger, longer-duration deals.
- Longer contract tenures are being negotiated: Total order backlog has grown faster than the 12-month backlog because Mastek is negotiating 3–5 year contracts instead of 1-year renewals, offering clients discounts or productivity efficiency in return for the longer commitment.
- UK government efficiency pressure adds to the dynamic: Continued pressure from the UK government to deliver 15% or higher efficiency across all government projects may have a short-term impact on revenue, compounding the voluntary shift toward outcome-based pricing.
Umang Nahata: "Our current approach is to move from time and material contracts to more outcome-based contracts. In fact, in this quarter also we have closed a few deals where we have moved our customers from time and material to an outcome-based contract and also offering anywhere between 15% to 30% productivity benefits to our clients."
Vertical and Geographic Signals
| Area | Trend | Commentary |
|---|---|---|
| Healthcare & Life Sciences | ▲ Investment priority | Global demand described as "really strong" across UK, US, and AMEA; learnings from NHS and global payer/provider organizations consolidated into a singular strength; growth in this vertical called fundamental to long-term sustained growth. |
| Financial Services | ▲ Early momentum | Won one new large account in a circa $20M deal; recognized as a contender in financial services by analyst firms for the first time; being developed as the third sector after Healthcare and Public Sector. |
| Oracle practice | ▼ Project-timing drag | 60% project-driven revenue means lumpiness; programs that went live in Q3 created a revenue hole, while new program starts right-shifted to Q4; Oracle sees Mastek as a top-20 global partner leading AI initiatives and awarded 3 recognitions at AI World. |
| UK & Europe | ▬ Mixed | Good demand in Healthcare and important renewals in Secured Government Services, but continued government efficiency pressure and high furlough concentration create near-term friction. |
| North America | ▼ Reset complete, revenue lagging | Leadership overhaul is 100% done and digital engineering capability is now competitive, but quarterly run rate has fallen to $20M–$23M; AI-led engineering is creating a much bigger digital engineering pipeline, especially as champion challenger to new customers. |
| Middle East | ▼ Sharp decline, unexplained | Management acknowledged the decline but was cut off before completing the explanation; prior restructuring was described as "almost over," making the fresh deterioration a point of unanswered concern. |
Guidance & Commitments
| Commitment | Specifics | Timeline |
|---|---|---|
| Revenue recovery | Q4 to show improvement in both top line and bottom line from Q3 (no specific numbers given) | Q4 FY26 |
| Oracle program starts | Right-shifted transformation programs scheduled to begin | Q4 FY26 |
| Furlough impact | No significant furlough impact going forward | Q4 FY26 onward |
| Labour code adjustment | Adjustment expected; ongoing impact estimated at $150K–$200K per quarter pending clarifications | Q4 FY26 (clarifications in 1–2 months) |
| U.S. sustainable improvement | "Next year would be the year when you start seeing U.S. starting to improve sustainably" | FY27 |
| FY27 vs FY26 | FY27 expected to be a stronger year than FY26, as core fundamental changes are largely complete | FY27 |
| EBITDA margin | Comfortable maintaining 16.5%–17% | Ongoing |
| Bottom-line trajectory | Expect to maintain steady bottom-line performance and continue improving over the coming period | Ongoing |
Umang Nahata: "I feel comfortable that we are more or less there in terms of whatever core fundamental changes that we had to deliver. As we go forward into FY '27, we are expecting FY '27 to be a stronger year as compared to FY '26."
Key Metrics
| Metric | Q3 FY26 |
|---|---|
| Operating revenue | ₹905.7 cr, up 4.2% YoY, down 3.7% QoQ in INR terms; down 4.8% QoQ in constant currency |
| EBITDA margin | 16.1%, improved 60 bps QoQ; 43 bps from AI-led operational efficiency, 17 bps from forex tailwinds |
| Net profit margin | 11.7%, improved 149 bps QoQ |
| 12-month order backlog growth | 18.4% YoY |
| Total order backlog growth | More than 30% YoY |
| Furlough impact | $2.7 million, representing 50% of the QoQ revenue decline; 70% of furlough impact from U.K. |
| Labour code impact | ₹6.4 cr, approximately 0.7% of revenue; negated by onetime leave provision reversal of 50–60 bps |
| New customers added | 17 in Q3 |
| Large deal win | Circa $20 million new account in financial services sector |
| Subcontractor cost | Down 0.7% QoQ as a percentage of revenue |
| Oracle business mix | 60% project-driven, ramps down after go-live and ramps up with new wins |
| Ongoing labour code impact estimate | $150,000–$200,000 per quarter going forward |
| U.S. quarterly run rate | $20M–$23M, down from approximately $30M post-acquisition |