Services GenAI 360: The Tide Won't Rise Evenly

What You Need to Know:

• Generative Artificial Intelligence (GenAI)/Agentic is fundamentally changing IT and Business Services delivery and operating models
• Competitive advantage will increasingly hinge on Intellectual Property (IP)-led delivery, partnership ecosystems & deep domain & process expertise.
• Such factors will be key drivers to influence investment criteria as the AI cycle unfolds.
• Our survey of enterprise IT decision-makers and Services vendors reflects mixed views on GenAI as a Services growth driver.
• However, there's agreement on the shift to non-full-time-equivalent (FTE), IP-led models longer term.

The TD Cowen Insight

As GenAI fundamentally changes IT and business process management (BPM) Services relationships, sector investing criteria will change. Amid the bear case displacement narrative, there's indiscriminate Services pressure, but we expect separation over 2026/2027 as enterprise GenAI adoption accelerates. The competitive advantage will increasingly hinge on IP-led delivery, deep partner ecosystems and real utility leveraging domain knowledge.

Our Thesis

GenAI and Agentic technology is the most controversial topic impacting the Services ecosystem, both in business performance and the investment backdrop. Many Street participants question the future role that IT & BPM Services vendors will have in a world where GenAI/Agentic becomes prevalent, with the most dire predictions seeing an existential threat to the industry. We don't hold as distressing a view, believing the broad brush of sector pressure is overdone. We question the timeframe for structural displacement risks and think there's little credit being given for AI-driven demand and evolving models that can supplement profitability. However, we acknowledge fundamental changes are afoot.

We believe key considerations for evaluating Services companies will transition with a greater emphasis on native IP development, partnership networks, and successful monetization of GenAI/Agentic solutions in new structures that blur operating models.

This transition will best be measured through:

• contract mix,
• partner-derived revenue and tiering,
• and per capita revenue and operating income key performance indicators (KPIs).

To be clear, enterprise GenAI adoption is in its infancy. This is an evolving controversy to be revisited.

As buying behavior normalizes post-COVID spend digestion, we believe we will enter a period where the structural impacts of GenAI/Agentic tech will be better determined and increasingly visible in Services providers' performance. However, we don't see this tide rising evenly. Rather, this is apt to accelerate diverging performance, and competitive advantages will increasingly hinge on IP-led delivery that relies on deeper partnerships and greater industry domain and process expertise.

GenAI 360: Tracking 360 Key Accounts for Business Use AI Adoption

Seeking input from both sides, we surveyed senior IT leaders in enterprise buyer organizations and Services vendors. We aggregated detailed interviews with roughly 21 senior IT decisionmakers in large organizations across industry verticals. These individuals spanned C-suite roles and senior directors of engineering and cloud services. We also posed detailed questions to approximately 15 public and private IT/BPM Services providers.

We leveraged proprietary data from our IT Services Spending Survey; across these three survey populations, we detail commonalities and divergence in current behaviors and future perceptions around GenAI. The synthesis of key messages from both sides of the conversation provides perspectives on the current state of GenAI/Agentic solution demand and development, whether it's impacting IT and Services vendors' spend, how Services vendors may be participating, if delivery models, hiring and contracting are being impacted and how these dynamics may evolve over time.

We also performed a wide-ranging analysis of vendors' current GenAI/Agentic solution development across IT and BPM Services coverage to separate marketing vs. practice to determine where real offerings exist (Services & IP). Considering the criticality of collaboration needed to succeed in this arena, we analyzed the breadth and maturity of partnership depth with hyperscalers, enterprise Software as a Service (SaaS) and data vendors and the most notable foundational large language model (LLM) vendors.

Financial and Industry Model Implications The early-stage nature of GenAI/Agentic adoption leaves models more dependent on near-term cyclical factors versus gradual structural change underlying our thesis. However, stabilization that has set in for most is set to reveal those with tailwinds from early GenAI/Agentic development efforts.

Opportunities reside in GenAI prerequisites (Services work around advisory, data, process, infrastructure), in net new technology acceptance model (TAM) generation via adjacencies or areas that were previously cost prohibitive (i.e., GenAI helping to reduce costs in areas like legacy code modernization) and by diversifying revenue streams via monetizing IP driving increased blurring of Services models.

Risks reside in higher productivity commitments that enterprise IT buyers now expect (20-30% expected productivity increase often cited per our surveys) that has weighed on gross margins amid a slower growth period, as well as potential insourcing and vendor consolidation due to GenAI efficiencies.

Successful adoption internally to drive increased productivity and favorable mix shift from the development of more IP solutions is expected to be visible via steadily increasing per capita revenue and operating income. Based on our analysis of vendors' current solutions and partnerships, we don't see that tide rising evenly.

What to Watch?

The pace of scaled enterprise GenAI/Agentic adoption

After GenAI/Agentic exploration & experimentation in small scale engagements over 2023 through the first half of 2025, Services vendors broadly cited a turn toward scaling a smaller number of initiatives as enterprises pivoted to a greater focus on return on investment (ROI). An evolution to more consistent industry bookings/revenue tracking vs. ad hoc disclosure is key, and select vendors have evolved segment reporting to better reflect data and AI-related solutions activity. Most organizations (about 75%) rate themselves between 2 and 3 (scale of 5) for organizational readiness, indicating moderate readiness/developing stages with pockets of advancement. While a subset (about 20%) report high readiness driven by strong infrastructure, organized data and formalized AI programs, the majority cite data governance, talent, organizational process alignment and legal/compliance as the gating constraints.

Receptivity to "build" versus "buy" decisions

Amid industry/market speculation of replacement, direct enterprise adoption of foundational model vendor products will be critical to monitor for IT/BPM Services & Enterprise Software sector fortunes.

Evolving contract terms and conditions in Services relationships that reflect a convergence of Services and Software

Such evidence of traction in more strategic GenAI Services and IP commercialization would resonate in quarterly/annual contract mix moving away from FTE-based/time and materials (T&M) contracting and appear in financials via expanding per capita revenue and per capita profitability for Services winners.

M&A activity may accelerate foundational model provider or Big Tech acquiring niche specialists AI consulting/Services providers and digital engineering vendors to accommodate "last mile" needs while valuations sit at deeply depressed levels.