EXLS Q2 2025 Earnings Call Transcript

EXLService reported Q2 revenue of $514.5 million, up 14.7% year over year, with adjusted EPS rising 20.3% to $0.49 and data/AI-led revenue increasing 17% to 54% of total sales. Management raised full-year 2025 guidance to $2.05 billion-$2.07 billion revenue and $1.86-$1.90 adjusted EPS, while also authorizing a $125 million accelerated share repurchase. Margins were slightly pressured by continued AI investment, with adjusted operating margin at 19.6% and management guiding for a modest second-half moderation.

Analysis

EXLS is quietly becoming a leveraged AI implementation story rather than a generic IT-services grower. The key second-order effect is that as more revenue migrates to usage/outcome pricing, the company should see a widening gap between unit economics and reported revenue quality: lower upfront billing volatility, but potentially higher lifetime value and stickier renewal economics if deployments embed deeply in client workflows. That makes the business less economically elastic than the market likely assumes, because a rising share of mission-critical workflows becomes harder to rip out once the model is trained on proprietary process data.

The moat is less about “AI capability” in the abstract and more about controlled access to domain datasets plus workflow integration. That should create a winner-take-more dynamic in verticals where compliance and traceability matter, especially insurance, healthcare, and banking, while smaller AI-first vendors get relegated to point solutions or pilots. A more subtle implication is that EXLS may increasingly compete against consulting firms on transformation spend, but with a more attractive ROI narrative; that can pressure legacy BPO peers to either accept lower-margin AI transition work or lose wallet share.

The near-term risk is not demand but margin timing. Management is explicitly choosing to reinvest into product, which means the P&L could look optically capped for the next 2-3 quarters even if bookings remain strong. The market may underappreciate that headcount leverage is already kicking in, so if deployment quality stays high and deal conversion holds, FY26 could be the first year where operating leverage re-accelerates without requiring another macro tailwind.

The contrarian angle is that the stock may deserve a higher multiple than plain-vanilla services, but not yet a software multiple until the commercial model proves durable at scale. The main question is whether outcome-based pricing expands margins or simply passes more value back to clients; that will determine whether this is a sustained rerating story or a temporary AI enthusiasm trade.

Key Decisions for Investors

• Long EXLS into any post-earnings digestion over the next 2-6 weeks; thesis is multiple expansion from a services multiple toward a vertical-AI compounder if conversion of pipeline into usage-based revenue continues.
• Add to EXLS on a pullback to the prior breakout area; use a 3-6 month horizon and target a rerating driven by FY26 margin inflection, with downside limited if revenue growth stays >10% and pipeline remains double-digit.
• Pair trade: long EXLS / short a legacy BPO or IT-services peer with weaker AI differentiation over the next 1-2 quarters; the spread should widen as buyers pay up for proprietary workflow data and verticalized models.
• For risk-managed exposure, buy EXLS call spreads 4-6 months out to capture a potential rerating while capping premium burn if management’s reinvestment keeps near-term margins compressed.
• Avoid shorting the stock solely on second-half EPS deceleration; instead, wait for evidence of deal slippage or margin miss, because the setup is more likely to punish late bears than early buyers.