The conversation around generative AI continues to be dominated by confidence and promise. Senior executives speak openly about the potential for better customer experience, better products, and fundamentally new business models enabled by AI. I share that conviction; the technology absolutely can deliver new operating models and superior economics.
But as we look toward 2026, particularly in the tech services market, there is a growing gap between belief in AI’s potential and the willingness to commit capital and change how organizations actually operate. That gap will define the next two years, making this a recalibrating year for tech services and AI-led transformation.
The Wait-and-See Problem in AI Transformation
When firms are faced with the risk of fundamentally changing how they operate, most prefer to wait until others go first. They want proof that the returns are real and a clearer view of how transformation is executed in practice. That instinct is understandable. Transformation cycles are long; they typically play out over 18 months to five years.
However, we are still early in that cycle for AI. AI-driven operating models have not yet emerged at scale. There are strong use cases, and there is real evidence of value, but not enough examples that create conviction for broad, enterprise-wide investment. As a result, we are in a prolonged wait-and-see period. Early adopters still need to demonstrate outcomes compelling enough to justify the disruption and capital required. Until that happens, investment and demand will remain gated by uncertainty around the “how,” not the “whether.”
A Flat Tech Services Market Is Not an Accident
Looking back over the past two years, the tech services market has behaved in an unusual way. Historically, the industry has grown at roughly 4.5 percent on average, with contractions limited to major shocks, such as the 2008 recession. Even during Covid and the immediate post-Covid period, growth resumed quickly. Yet the last eight quarters have been essentially flat. This is not because demand for technology has disappeared. Instead, growth has been absorbed elsewhere.
One major factor has been increased insourcing. Global capability centers and captives, particularly in India and other low-cost locations, have expanded rapidly. Companies are building capabilities internally that they once sourced externally.
At the same time, AI-enabled productivity gains are reshaping application development and the software development lifecycle. Code generation is the most advanced and most impactful AI use case in the enterprise today. Teams are already achieving 25 to 35 percent productivity gains, meaning the same amount of work can be delivered with fewer people. The combination of insourcing and productivity-driven efficiency has effectively neutralized industry growth.
What Will Change in 2026?
As we look ahead to 2026, it is quite possible that the tech services market will move from flat to contracted. Code generation continues to improve at a rapid pace. Internal technology teams and service providers alike are redesigning their operations to capture these benefits. By the end of 2026, advanced teams could see productivity improvements of 60 to 70 percent in parts of the SDLC.
This creates significant revenue compression for service providers tied closely to traditional application development models. It also challenges long-held assumptions about the balance between in-house and external delivery. While new AI-driven demand will emerge, it is unlikely to be unleashed quickly enough to offset the compression caused by productivity gains.
That said, this contraction will not be evenly distributed.
Where Growth Will Still Exist
Business process services are in a much stronger position. AI tools are not yet mature enough to drive the same level of revenue compression in BPO, and organizational understanding of how to apply AI in these areas is still developing. As a result, BPO is likely to continue growing through 2026.
The midmarket also stands out as a growth segment. Firms between roughly $500 million and $10 billion in revenue are earlier in their AI journeys, and often lack the internal scale to absorb work through insourcing alone. For these companies, external partners remain critical.
Industry differences will matter as well. Technology companies, financial services firms, and large banks are far more advanced in applying AI and are already generating substantial demand for specialized services. Functionally, marketing is emerging as an area where AI maturity is higher than in most other business functions. By 2026, I expect to see a noticeable pickup in demand for transforming how marketing operates.
The Operating Model Is the Real Constraint
Across all of this, one rule remains constant: AI transformation requires an operating-model change. Productivity gains in the SDLC are just the beginning. Over time, we will see pressure on where work is done, and how it is organized. Some workloads that were historically offshored may move closer to the business. Higher productivity, tighter collaboration, and aligned time zones can offset traditional labor cost advantages, but this shift will be gradual.
The net effect will be a flat-to-contracting market combined with changing delivery economics. That combination will put intense pressure on service providers, particularly those anchored in legacy delivery models.
What 2026 Means for Buyers and Providers
For service providers, 2026 will be a difficult year. Price competition will be aggressive, margins will be under pressure, and differentiation based purely on scale or cost will continue to erode.
For customers, the environment will be more mixed. Prices are likely to fall, but navigating new delivery models will create complexity and disruption. Many organizations will struggle to adapt their operating models fast enough to fully capture AI’s benefits.
Consulting, however, will see stronger demand. As companies confront the reality that technology alone does not deliver transformation, they will need help redesigning how they operate.
A Delayed, Not Denied, Transformation
None of this suggests that AI transformation will fail. On the contrary, conviction that AI can deliver value is deep and widespread. What is missing is clarity on execution and the willingness to invest, ahead of proven models becoming available.
There will be a modest recovery after 2026, but sustained growth will not return until firms are ready to commit capital and embrace true operating-model change. Until then, the market will continue to recalibrate.
