Aokah et les leaders du secteur identifient cinq tendances qui redéfinissent les centres de compétences mondiaux

Le secteur des centres de compétences mondiaux (GCC) évolue vers des modèles portés par l’IA : 83% des GCC investissent déjà dans l’IA générative et 58% dans l’IA agentique, avec 29% qui prévoient d’intensifier leurs investissements. En revanche, l’exécution reste un point faible : plus de 72% des nouvelles configurations GCC subissent des retards matériels ou des dépassements de coûts dans les 24 premiers mois. La gouvernance est également insuffisante, puisque seules 21% des entreprises disposent d’un modèle abouti pour l’IA agentique/autonome.

Analysis

This reads less like a broad AI acceleration story and more like a redistribution of profit pools: from headcount-heavy outsourcing toward orchestration, governance, and enterprise workflow control. That is structurally negative for legacy offshore IT services that monetize utilization, and relatively positive for firms that sell process redesign, controls, and software layers that sit above the labor pool. The key second-order effect is that captive centers becoming “strategic” usually reduces vendor leverage, because the client retains the highest-margin work and outsources only the implementation bursts.

Near term, I would not expect a clean earnings impact in the next quarter; procurement cycles and org redesign slow the conversion of narrative into revenue. Over 6-18 months, however, AI-driven redesign should pressure pricing for traditional services while expanding demand for governance, observability, identity, and workflow tooling. That favors names like ACN, NOW, and MSFT more than labor-arbitrage proxies such as INFY, WIT, and to a lesser extent CTSH/EPAM.

The contrarian point is that the market may be overestimating how fast enterprises can rewire operating models. Most of the spend may go to consultants and compliance rather than durable software consumption, which means the upside for pure-play AI vendors could be smaller than the narrative suggests. The thesis is falsified if offshore IT order growth re-accelerates without margin compression, or if consulting backlog and workflow-software ACV fail to inflect despite continued GCC expansion.

Key Decisions for Investors

• Pair trade: long ACN / short INFY for 6-12 months. Thesis: GCCs internalize strategic work, but implementation and governance budgets migrate to consultants; risk/reward skews to ACN if pricing power holds and INFY shows utilization or realization pressure.
• Initiate a selective long on NOW on pullbacks into earnings over the next 1-3 months. The operating leverage comes from process redesign and governance attach rates, not generic AI hype; exit if billings or remaining performance obligations fail to accelerate.
• Avoid chasing rallies in WIT and CTSH; use strength to trim or hedge exposure over the next quarter. They are more exposed to seat-based pricing compression if enterprises shift work into captive centers and automate with AI.
• If you need a broader sector expression, prefer long MSFT over a basket of offshore services for the 6-18 month horizon. MSFT benefits from the control plane, identity, and cloud spend that GCC modernization requires, while the downside is that enterprise adoption stays too fragmented to move the needle quickly.
• Set a watch item on Indian IT guidance and consultative backlog: if large-deal TCV or margins do not weaken within 1-2 quarters, the short case in offshore services is not ready.