The Build Decision Changed the Relationship

A third of enterprise organizations skipped their last software purchase because they could build it themselves with AI coding tools. That is not a procurement trend. That is a relationship change.

I have been in more build-vs-buy conversations than I can count. For most of my career, the answer was almost always buy (which I know is a little strange coming from the guy whose teams build custom software): faster time to value, lower total cost, fewer risks, stay in your lane because software is not your core business...you get the point. That framing served clients well. It still does, in the right situations. But McKinsey's State of AI 2026 found that 32% of organizations have already skipped at least one software purchase because they could build the equivalent internally using agentic coding tools. Among the AI high performers (the 6% attributing more than 5% of EBIT to AI), that number climbs to nearly 50%. I have been watching this shift play out in client conversations in real time. The "should we buy this platform?" question is getting replaced with "wait, can we just build this?" And sometimes the answer is actually yes.

The part I want to be careful about, though, is throwing out the buy-side argument entirely. The EBIT impact of AI held flat at 37% of companies despite all this building activity. One in five organizations is limiting AI use specifically because of operating costs. The capability to build did not automatically produce the judgment to build the right things, in the right order, with the right governance underneath (I have seen some genuinely impressive things built that nobody needed, at significant cost, because nobody asked that question first).

The consulting opportunity has always been judgment, not labor. The firms that will survive this shift are not the ones who argue against it. They are the ones who reposition around what clients genuinely cannot do for themselves: figuring out what is worth building, measuring whether it worked, and knowing when to stop before the sunk cost calculus kicks in.

If a third of your clients can now build the thing you sell, what exactly are you selling? And where do you make up that third?