Contraction Before Expansion: How AI is Poised Spark Tech Services Growth
getty
Over the last two years, something unusual has happened in the technology services industry, an industry that, for three decades, could almost be counted on for steady growth. Historically, consulting, systems integration, and outsourcing, especially those tied to the software development lifecycle (SDLC), have expanded at rates of 5 to 15% annually.
But today, growth has flatlined. Depending on the quarter, the industry has fluctuated between 1% growth and a 2% contraction. Of course, there are firms still growing, but their growth comes at the expense of others. The industry has entered what I would call a market share swap environment. Those gaining ground are doing so by taking business from competitors rather than expanding into new demand. Overall, there’s effectively no net growth in the market.
Where Did the Growth Go?
In a world where we talk so much about AI investment, it’s not translating into growth in tech services. Several forces are at work here, and together they’re reshaping the industry’s structure.
The first major factor is insourcing. Over the past two years, enterprises have reduced the share of discretionary projects they outsource to third parties. That doesn’t mean they’ve stopped doing these projects; they’re just increasingly doing them themselves using AI-assisted development tools.
A few years ago, discretionary spend was a major driver of growth for service providers. Today, a growing proportion of that spend is absorbed internally. Teams can now accomplish with a small number of developers, aided by AI tools, what once required large external delivery teams.
The second factor is the growth of Global Capability Centers (GCCs). For twenty years, the industry’s story has been one of labor arbitrage, with work shifting offshore to lower-cost locations, such as India or the Philippines. Now, however, many enterprises are choosing to build their own centers rather than rely on external providers. This too is a form of insourcing, and it’s accelerating rapidly.
When you combine these two trends – AI-enabled insourcing and the proliferation of GCCs – you get a significant reduction in third-party demand.
AI Efficiency Is Fueling a Short-Term Revenue Squeeze
AI isn’t just changing who does the work; it’s changing how efficiently the work can be done. As automation and AI-assisted delivery increase, the same projects can now be delivered faster and at lower cost.
Companies that once charged $1 million for a development project can now profitably bid $500,000 and still make their margins. This efficiency is beneficial for clients but exerts downward pressure on industry revenue overall.
Put simply, we are doing more work for less money, and that translates into flat or contracting top-line growth.
It’s highly probable that we’ll see modest industry contraction over the next six to twelve months as AI productivity gains take hold.
Why AI’s Early Impact Is Deflationary, Not Expansive
Every major technology shift creates both disruption and opportunity. Historically, when a disruptive technology enters the market, it lowers the cost of technology, and as a result, consumption expands; people buy more because they can afford more.
That’s the long-term hope for the tech services industry. But in the near term, AI’s impact is deflationary, not expansive.
We are in the early stage of AI transformation, where enterprises are still experimenting, building governance frameworks, and exploring safe deployment. That work, while necessary, doesn’t yet generate large-scale commercial demand for third-party services.
Meanwhile, AI’s first wave of productivity is being realized inside enterprises, not through outsourcing. The same tools that allow service providers to automate development are also available to clients. And many are seizing the opportunity to bring work in-house.
So we find ourselves in an unusual paradox: the most exciting technology trend of the decade is, in the short term, deflationary for the services sector.
Sustainable Growth Depends on Rethinking the Operating Model
But that’s not the end of the story. If history is any guide, the industry will rebound, not through incremental automation, but through a new phase of AI-driven transformation.
The real unlock from AI occurs not when you add AI tools to your existing operating model, but when you change the operating model itself.
Simply automating what you already do produces limited results. It’s the equivalent of paving over old roads rather than redesigning the city. The productivity gains are real but small.
The significant ROI comes when enterprises reimagine how they operate, including how decisions are made, how work is structured, and how technology, data, and people interact. When companies reach that stage, the scope of work expands dramatically.
To embed AI into an operating model, enterprises typically need to re-platform, modernize their data, and often migrate to the cloud to access and unify that data. These initiatives are extensive and create large new scopes of work, precisely the kind of large-scale transformation work where external partners play a vital role.
Define the Future Model Before Investing in the Tools
The key is sequence. Many companies are starting with the technology, investing in AI pilots, cloud infrastructure, and data modernization, without first defining what their future operating model should look like.
This is the same mistake we saw in the early stages of digital transformation. Enterprises modernized their infrastructure and migrated workloads, but because they didn’t simultaneously redesign their operating models, the ROI fell short.
The right approach is to start with a vision of the future operating model, then work backward to identify the technology and data requirements to enable it. That’s what unlocks both the transformation and the business case for investment.
When enterprises reach that stage, the work becomes far more complex and integrated. It requires consulting, system design, data engineering, change management, and often organizational redesign. And those capabilities typically come from third parties.
What Will Drive the Next Growth Cycle
This brings us to the question every industry leader is asking: When will this next phase begin?
Right now, large-scale operating model change is still rare. There are a few promising examples, but not yet enough for a pattern to emerge. As with every previous technology shift, most companies will wait to see credible AI transformation proof points before committing.
But once those early exemplars demonstrate success, adoption will accelerate quickly. And when that happens, the scope of transformation work will expand significantly, ushering in a new growth cycle for the tech services industry.
There’s also a broader economic principle at work – Jevons’ Paradox: as the cost of a technology decreases, its use increases. AI is likely to follow this same path. As AI makes technology more affordable and accessible, enterprises will consume more of it, not less.
That’s why, despite the current contraction, I remain optimistic. The industry is adjusting to a new equilibrium, or a temporary compression before the next expansion.
The Industry Reset Before the Rise
The near term will be challenging. AI-driven efficiencies, insourcing trends, and price compression will continue to put pressure on revenue and margins. But these forces are also laying the foundation for the industry’s reinvention.
As enterprises move from experimenting with AI to transforming with AI, the complexity and scope of work will grow. That’s where tech services firms have an opportunity to lead — helping clients reimagine operating models, re-platform architectures, and build the systems of execution that define the next era of business.
So yes, we are likely to see contraction before expansion. But once operating model transformation takes hold, I believe we will look back on this period not as a decline, but as the reset before the next wave of growth.
