The trillion-dollar gap that is closing

The global technology services industry exists largely because business software is complex, generic and hard to implement. AI is changing all three at once.

Think about how it actually worked. You bought Salesforce, then you hired Accenture to make it work. You bought SAP, then you hired Capgemini to spend 18 months configuring it to almost fit your business. You bought Oracle, then you hired Deloitte to build the integrations. You bought ServiceNow, then you hired TCS to manage it.

The entire services ecosystem fed directly on the gap between what the platform promised and what your organisation actually needed. Accenture, Deloitte, KPMG, Capgemini, Wipro, TCS and hundreds of smaller firms all built their business models around closing that gap. It was enormous because the software was generic by design. Closing it required armies of consultants, integrators and project managers, often for years and often at a cost that exceeded the software itself many times over.

This was not a failure of the services firms. It was a rational response to a real problem. The software genuinely was that complex and that generic. Someone had to bridge the gap, and they built trillion-dollar businesses doing it.

Now the gap itself is closing. If the software is purpose-built for your specific problem from the start, you do not need a consulting firm to translate it, a systems integrator to bolt it together or an outsourcing partner to keep it running. The model that sustained the services industry for forty years starts losing its reason to exist, not because the services firms got worse but because the problem they solved is going away.

This is already visible in the numbers. The major consulting and outsourcing firms have been reporting slower growth, reduced deal sizes and pressure on margins for the past 12 months. Most are attributing it to macroeconomic headwinds. Some of that is genuine. The rest is structural, and the AI capability jumps in late 2025 and early 2026 accelerated the structural part dramatically.

The smarter services firms can see where this is heading. Instead of selling labour to help customers use someone else's software, they are starting to use AI to build the finished product themselves. An agency that used to charge $500,000 to implement and customise a CRM can now build a purpose-built alternative for a fraction of that and sell it as a product with recurring revenue, margin and scale. The agencies of the future will not look like consulting firms. They will look like software companies.

If you work in or buy from the services industry, the question worth sitting with is uncomfortable but important. How much of what you are paying for is solving a problem that was created by the software itself rather than by your actual business?

How does your organisation actually score?

Get a free AI Disruption Snapshot in five minutes — no credit card.