Capturing the value of AI: from "experimenting" to "rewiring"

Experimenting with AI has never been easier. Getting value from it has never seemed harder. Between the two lies a shift in mindset that McKinsey sums up in one word: rewiring. The goal is no longer to add AI to the business, but to rewire how the business works.

Companies have overwhelmingly crossed the usage threshold. According to the 2026 edition of McKinsey's State of AI, 89% of organizations regularly use AI in at least one business function. But fewer than half, 44%, say they have scaled it across the enterprise, up from 38% a year earlier. KPMG's Global AI Pulse Survey is even starker: only 11% of organizations, which it calls "AI leaders", treat AI as an enterprise-wide transformation rather than a bolt-on to what already exists.

On results, the picture remains modest. Only 37% of McKinsey's respondents see any effect of AI on their operating income, and just 6% attribute at least 5% of their EBIT to it. Meanwhile, boards are losing patience: in Dataiku's 2026 survey of 685 CIOs, 97% report rising board pressure for AI return on investment, and 76% expect their job to be at risk if their company shows no measurable gains by the end of 2027. The window for experimentation is closing.

Experimentation: a useful phase that became a trap

Experimenting has real merits. Pilots make it possible to learn fast, test tools at low cost and familiarize teams with a new technology. For two years, it was the right approach: nobody yet knew exactly what generative AI could do, or where.

The problem is that many companies never left that phase. Pilots follow one another, each proving the technology works within a narrow scope, and none changes how the company actually operates. KPMG pinpoints the dividing line: the most advanced 11% redesign their processes before deploying AI, while the rest layer it onto existing workflows. Experimentation then becomes a permanent state, reassuring and expensive.

What "rewiring" really means

The word comes from McKinsey, which popularized it in the book Rewired, published in 2023. The core idea is simple: a technology-driven transformation does not succeed because the technology is deployed, but because the organization using it is rewired. A business-led roadmap, talent, operating model, technology, data, adoption at scale: technology is only one of six dimensions that must evolve together.

The 2026 data confirm this. Among the top performers identified by McKinsey, 73% have fundamentally redesigned their workflows, compared with 25% of other companies. These same companies are far more likely to have strong leadership commitment and a process for measuring impact, twice as likely to devote more than 15% of their IT budget to AI, and three times as likely to be scaling agents across functions.

Rewiring, then, is not a more ambitious AI project. It is a transformation project in which AI is the lever.

Workstreams to make the shift

Moving from experimentation to rewiring does not mean rebuilding everything. It means changing how initiatives are chosen, designed and managed. Four workstreams are decisive.

The first is to think in domains rather than use cases. A use case improves a task; a domain, such as the sales cycle, the supply chain or pricing, brings together a set of processes that can be rethought end to end. BCG estimates that 70% of AI's potential value is concentrated in a handful of core functions: sales and marketing, manufacturing, supply chain and pricing.

The second is to redesign the work before deploying the tool. Which tasks disappear, which are handed to AI, which remain human, and how roles change: these questions must be settled before deployment, not after. This is precisely what separates McKinsey's 73% of top performers from the 25% of other companies.

The third is to set the expected value from the outset. Every rewired domain needs a quantified target, in margin, turnaround time or revenue, and a way to measure it. Without that target, it is impossible to tell whether the company has left experimentation behind or simply made the pilot bigger.

The fourth is to commit senior leadership and the resources that go with it. Rewiring affects organizational structures, roles and budgets. It cannot be driven by an innovation team or by IT alone: it requires trade-offs that only the executive team can make.

The pitfalls of the transition

Leaving experimentation behind carries its own risks. Three mistakes come up again and again.

The first is scaling too early. Rolling a pilot out company-wide before its value has been demonstrated only multiplies costs. Scaling should be a reward for what has proven itself, not a reflex.

The second is confusing rewiring with reorganization. Moving boxes on an org chart, creating an AI department or appointing an AI lead changes nothing if the way people work stays the same. It is processes and decisions that need rewiring, not just structures.

The third is neglecting governance. In Grant Thornton's 2026 survey, 46% of executives cite governance failures as a leading cause of AI underperformance, and 78% lack confidence that they could pass an independent AI governance audit within 90 days. The deeper AI is built into core processes, the more strategic its oversight becomes.

From pilot to system

The move from experimenting to rewiring marks a change in kind. The question is no longer whether AI works, but how the company should work with it. It is a less technical, more demanding question, and one that belongs squarely to the executive team.

The stakes are significant. According to BCG, only 5% of companies are "future-built", meaning structured to capture AI value at scale, while 60% report only minimal revenue and cost gains. The gap between these two groups has nothing to do with the technology they have access to, and everything to do with how deeply they have changed.

For a business leader, the right question is no longer "how many pilots have we launched?" but "which process have we completely rethought this year, and what has it delivered?" That is the true measure of the shift from experimentation to transformation.

Hymeria's position

Before rewiring, you need to know where. Hymeria Consulting helps business leaders identify the domains where AI can create the most value, assess the possible scenarios and build the business case needed to decide. Our studies, analyses and business plans combine senior-level scoping, more than 40 specialized AI agents producing the analysis, systematic validation by an expert with over 15 years of experience, and an actionable deliverable in 5 to 10 days, within a fixed budget agreed before work begins.

Related service: Innovation & Technology, technology landscape, make-or-buy analysis and product roadmap in 5 to 10 days.