AI is redefining IT services, but earnings quality is the real test

 In brief

• AI is rewriting the economics of IT services, shifting the focus from scale and headcount growth to productivity, automation and outcome-based delivery models.
• Earnings quality now matters more than headline growth, requiring assessment of revenue sustainability, AI revenue attribution, pricing power and margin resilience.
• Due diligence should be forward-looking, evaluating AI’s long-term impact on profitability, cash flows, balance sheets and value creation.

Rethinking revenue quality in the age of AI

Revenue quality has become central to diligence. Growth in large or focus accounts can overstate performance if the long tail is flat or declining. Diligence needs to assess whether expansion is broad-based, repeatable and supported by genuine client demand rather than a narrow set of accounts. Distinguishing organic growth from acquisition-led growth is equally important, especially where inorganic revenues can mask softness in the core business.

Pricing dynamics also require scrutiny. AI productivity gains are changing client expectations and intensifying repricing discussions. Benefits once retained by service providers may increasingly be shared with, or passed on to, customers. This can alter forward revenue profiles even where volumes remain stable. Longer-duration, milestone-based and outcome-linked contracts are also making backlog less straightforward. Diligence should distinguish contracted revenues from contingent, phased or performance-dependent elements, particularly in complex AI transformation programs.

Decoding AI revenues and delivery economics

AI revenue attribution remains inconsistent across the sector. Reported AI revenues may include AI-embedded services, AI consulting, platform-led offerings, automation-enhanced managed services or productivity-led transformation programs. Each has different pricing, margin and scalability characteristics. Without a clear taxonomy, headline AI revenue can be difficult to compare and may overstate recurring or proprietary value creation, making AI financial due diligence increasingly important.

Delivery economics are shifting as pricing moves from effort-based models toward outcomes, milestones and delivered value. Traditional operating metrics such as headcount growth, utilization and rate cards are becoming less predictive. The more relevant diligence questions are whether productivity gains are measurable, whether AI assets are reusable across clients, whether pricing is defensible and whether savings translate into sustainable margin improvement rather than one-off efficiencies.

Rising judgment in earnings quality

Earnings quality is becoming more judgment-driven. A key area is the accounting treatment of AI investment, including platforms, tools and accelerators. Whether such spending is expensed or capitalized can materially affect reported profitability and comparability. Outcome-linked contracts add margin uncertainty, as historical delivery experience may not be sufficient to estimate future effort, delivery risk or gain-share economics.

Normalization is also more complex. Severance, restructuring and transformation costs need to be assessed carefully to distinguish one-off charges from recurring operating model costs. Management-identified synergies require validation to distinguish structural savings from temporary cost deferrals. Currency movements can further distort earnings, particularly where foreign exchange gains or hedge outcomes are not repeatable.

Implications for the balance sheet and cash flows

The impact of AI-led transformation extends beyond the income statement. Earnings Before Interest and Taxes (EBIT) and Earnings Per Share (EPS) can be affected by amortization of acquired intangibles and goodwill-related charges, particularly after capability-led acquisitions. Tax shields may support cash flows, but their timing, availability and sustainability require independent assessment.

The definition of net debt is also expanding. Beyond borrowings, diligence should consider committed costs, severance liabilities, stock option settlements, earn-outs and transformation-related obligations. These items can influence valuation, deal structure and post-transaction liquidity.

Cash conversion is another pressure point. As AI projects become milestone-driven and payments more back-ended, revenue recognition may run ahead of cash collection. Receivable cycles can lengthen, deferred or conditional billing can increase, and upfront investment in platforms and tools can absorb cash. Working capital analysis therefore needs to become more forward-looking and linked to delivery assumptions, including AI’s impact on cash flow and margins.

The bottom line

As the sector moves toward FY27, quarterly performance will increasingly reveal whether AI is driving structural operating improvement or merely offsetting weak demand through efficiencies and accounting judgments. In a low-growth environment, how earnings are generated matters more than how fast revenues grow. Diligence needs to evolve from hindsight to foresight — testing the resilience, visibility and durability of AI-led value creation.