ESN and AI: New Rates, Margins, and Recruitment

AI-native low-cost players are shaking up IT service companies, while the effects of AI on margins remain heterogeneous. Hiring policies are shifting towards AI expertise and upstream consulting, and two out of three companies are already modifying their pricing. KPMG and Numeum conducted a study among approximately 200 companies and quantified these shifts.

AI-native Competition and Margins: High Risks, Mixed Results

Low-cost AI-native companies are cited by 80% of respondents as a major risk to the value chain, ahead of clients internalizing projects using AI tools (53%) and designing their own agents (48%). On the profitability front, the picture is mixed. Among large companies and mid-sized enterprises (ETIs), 36% report a slight margin improvement attributable to AI, and 6% a net improvement, while 27% have not yet measured this effect. Among small and very small enterprises (SMEs and TPEs), 21% report a net improvement, but 42% do not quantify it. Conversely, margins have deteriorated for 8% of large companies and ETIs and for 7% of SMEs and TPEs, with AI investments exceeding observed gains. On the commercial front, 33% of players report increased pressure on prices.

Recruiting Differently: Upskilling Profiles and Increased Selectivity

62% of IT service companies (ESNs) and ICT firms already see an impact of AI on their hiring policies. For 32%, this translates into increased selectivity and higher expectations for AI skills, with 19% making targeted reductions in certain roles and 11% compensating through internal upskilling. Over the next 18 months, 43% prioritize hiring AI specialists (data, LLM, agent technology), 42% seek profiles that bridge technical and business roles, and 39% aim for candidates with 3 to 8 years of experience. Profiles with solid experience, such as architects or technical leaders, are favored by 31%, while 18% are leaning towards juniors "augmented by AI." Essential skills include critical thinking and the ability to validate AI-generated results (39%), expertise in business framing and high-level consulting (22%), practical mastery of tools, from prompting to copilots (17%), managing hybrid human-agent systems (12%), as well as autonomy and continuous learning (10%). Regarding recent graduates, 47% of respondents require proficiency in AI upon entering the job market. When it comes to recruiting juniors, 47% believe the situation remains stable, 22% observe a significant decrease, 9% note an increase, and 22% have not yet made a decision. Within teams, 39% do not foresee changes to the staffing model, 24% anticipate a decrease in the number of juniors but with a higher average skill level, 16% expect a distribution between highly experienced profiles and AI agents or tools, while only 5% plan to increase junior recruitment through AI.

Massive Training Remains the Exception, Cultural Adaptation Progresses Gradually

In 20% of ESNs and ICT firms, over 80% of employees have already been acculturated to generative AI. However, the most common scenario involves a limited scope, between 10% and 30% of employees, observed in 28% of companies. Advanced training dedicated to AI remains unevenly distributed across organizations.

Pricing in Transition: Towards More Subscriptions and Created Value

Two out of three companies are evolving their pricing policies. The most frequently cited avenues are unit-based or fixed pricing (59%), subscriptions (42%), value-based or results-based billing (33%), monetization of assets or licenses (24%), and success-based compensation (13%). Conversely, 30% maintain overall unchanged rates. In the short term, 37% believe that the shift towards value-based billing represents the most plausible market trend; monetization of platforms and agents ranks second, followed by downward pressure on daily rates and the rise of operational activities. On the client side, 75% of surveyed companies report increased expectations for productivity through AI, and 62% express a heightened demand for upstream consulting. AI-native players are already imposing new models, such as success-based compensation. Some executives report projects billed at a few thousand euros that, according to them, generate value for the client estimated in the hundreds of thousands of euros.

Increased Productivity, Primarily Used for Assistance, and Evolving Strategies

37% of surveyed companies report a productivity gain exceeding 20% on projects thanks to AI. The use remains primarily focused on production assistance for 86% of respondents, while 11% employ it to autonomously produce certain deliverables. These developments occur while ESNs still predominantly bill on a time-and-materials or fixed-price basis, even as AI accelerates team work. 58% of ESN and ICT leaders anticipate a strong or disruptive impact of generative and agent-based AI on their business model, according to a KPMG and Numeum study conducted among approximately 200 companies, which documents both mission sales and sought-after profiles. Strategically, 52% of large companies have made a shift, compared to 38% of ETIs, 32% of SMEs, and 30% of TPEs; within these last three categories, between 57% and 60% declare they are in a phase of adaptation.