Un post sul AI Is Not Just Accelerating Startups. It Is Rewriting the Rules for Anyone Who Builds Businesses.

What happens when the venture creation playbook is rewritten by artificial intelligence — and why most leaders are still looking in the wrong direction.

A few weeks ago, McKinsey published a report titled "How to build businesses faster and better with AI" (March 2026, McKinsey Business Building). Worth reading. Not because it says anything particularly revolutionary — many of us in the field had been sensing these dynamics for some time — but because it puts concrete numbers on structural shifts that those operating in the tech and VC ecosystem observe daily on the ground.

One figure stands out above the rest. In 2025, 61% of corporate ventures generated more than ten million dollars in revenue, up from 45% in 2023. And the time required to reach those revenue levels fell from 38 months in 2023 to 31 months in 2025. That is not a marginal improvement. That is a structural change in the economics of business building.

But — and this is precisely the point I want to develop — these numbers do not tell only a story about AI. They tell a story about the quality of human decisions that AI amplifies. This distinction matters enormously, and it tends to get lost in the noise of technological enthusiasm.

The real problem is not speed. It is what you do with it.

McKinsey identifies three dimensions of value that AI brings to venture building: creativity and innovation cycles, go-to-market velocity, and productivity per person. All three are correct, and the report argues convincingly that these gains compound quickly: smaller teams can remain small for longer, reduce coordination overhead, and improve capital efficiency without slowing progress.

The problem is that most organisations use these tools to do the same things faster. And doing the same wrong thing faster is accelerated waste, not transformation.

I have seen private equity funds that adopted AI for deal sourcing without redesigning their selection criteria. I have seen companies that automated recruitment workflows without asking themselves what kind of leader they actually need in the next phase of growth. Speed without strategic clarity is noise, not signal.

McKinsey makes this point, though in more technical language: the goal is not simply to execute current processes more rapidly. AI enables a more fundamental shift, one that moves the critical moment of learning earlier in the venture life cycle, front-loading customer validation, product iteration, and market-signal detection before large capital commitments are made. The ventures that capture the most value from AI are not those that merely automate existing processes. They are those that use AI to ask better questions earlier, fail faster on weak ideas, and concentrate resources on opportunities with genuine product-market fit.

Translated into plain language: the value lies not in faster execution but in failing earlier on weak ideas, and concentrating resources on what actually has traction. This requires intellectual honesty and strategic discipline — qualities that no AI model can install into an organisation.

The talent question: where almost everyone in the market gets it wrong

One of the more interesting passages in the McKinsey analysis concerns team composition. The core shift, the report argues, is not simply combining domain experts with AI talent, but deliberately scaling expertise by turning it into a hybrid human-agent capability. Through agentic systems, tacit knowledge embedded in documents, processes, and experienced individuals can be extracted, structured, and reused at scale.

This connects directly to what I encounter every day in my executive search and advisory work. The question I receive most frequently from CEOs at this moment is: "Do I need to hire an AI lead, a Chief AI Officer, someone who can bring AI inside the organisation?"

The correct answer is never simple. And the question itself often reveals a fundamental confusion: the search is for a person who will bring AI in, when the real challenge is building an organisation capable of using it.

The distinction is substantial. A brilliant CAIO inside a company with a change-averse culture and poorly structured data does not produce value. It produces mutual frustration. Conversely, a leadership team already oriented toward data-driven decision-making, even without a dedicated AI figure, tends to integrate tools organically and extract real competitive advantage from them.

What McKinsey calls the "agentification of expertise" — transforming the knowledge of top performers into scalable systems — is precisely the challenge I see emerging in the most advanced conversations I have with boards. It is not a technology challenge. It is an organisational and change management challenge. The technology, at this point, is the easier part.

The number that most CEOs ignore

McKinsey confirms that 67% of companies that prioritise business building outgrow the market, and that each dollar of new-venture revenue creates roughly twice the enterprise value of a dollar generated in the core business.

This number should appear in every board meeting deck. Not as a curiosity but as a strategic argument.

And yet most large companies — not only Italian ones — continue to treat innovation as a discretionary cost, cuttable in times of margin pressure. It is a mistake that compounds with interest. Those who step off the AI-first venture building wave today will attempt to re-enter three years from now, finding a market already consolidated around players who used that time to experiment, fail quickly, and scale the winners.

The McKinsey data also confirms something that contradicts a common misconception: lower cost per experiment is not an argument for cutting venture budgets. It is an argument for running more experiments. With AI lowering the cost of experimentation, organisations can place more small bets, exit weak ideas earlier, and concentrate capital and talent on the few that break out. This is fundamentally different logic from what most corporate innovation departments operate with today.

What "AI backbone" actually means in practice

The McKinsey report introduces the concept of an "AI backbone" — data infrastructure and AI governance as the operational foundation of every new venture. At the centre of this foundation sits data: both the venture's own real-time signals from customer interactions and product usage, and the parent organisation's institutional depth through historical benchmarks, proprietary research, and market knowledge.

Translated into board language: the competition over the next five years will not be won by the company with the most sophisticated AI model. It will be won by the company with the best-structured data and the most robust governance to use it. This is good news and bad news simultaneously. Good, because AI tools remain accessible to almost anyone. Bad, because historical data, data quality, the culture of actually using data in decisions — these require years of work, not a SaaS subscription.

This is precisely where competitive advantage is built or lost. And executive search, in this sense, returns to centre stage. Those who evaluate leaders today must be capable of reading not only financial results but the quality of the informational and decisional ecosystem that person has built around themselves. A CFO who has run a data-dark finance function for ten years is a different risk profile in 2026 than they were in 2019 — regardless of what their revenue numbers say.

The organisational lens that most technology analyses miss

After more than three decades observing how organisations respond to technological shifts — from the transition to ERP systems, to the internet, to mobile, to cloud, and now to AI — I have developed a certain allergy to uncritical enthusiasm.

Not because AI is not transformative. It is. But because every technological wave carries with it a simplified narrative that tends to obscure the human and organisational complexities that ultimately determine who wins and who loses.

My background in sociology of communication and organisational behaviour gives me a particular lens for reading these dynamics. The organisations that succeed in technological transitions are rarely those with the largest technology budgets. They are those where leadership understands that technology adoption is fundamentally a social process — governed by culture, power dynamics, informal networks, and the human resistance to changing established routines.

This is as true for AI as it was for ERP implementations in the 1990s, for internet transformation in the early 2000s, and for cloud migration a decade later. The technology itself was never the decisive variable. The decisive variable was always the quality of the human decisions made around it.

Three questions that matter more than any AI tool

The McKinsey message — AI as the new operating system of venture creation — is substantially correct. But the risk is that it gets read as a technology promise, when it is in reality a leadership challenge.

The companies that will capture disproportionate value will not be those with the highest AI budgets. They will be those with the most clear-headed CEOs, capable of answering three questions honestly.

First: which decisions do I want to make better, and earlier in the process? This is a question of strategic clarity, not technology.

Second: who do I need in my team to do so? This is a talent and culture question. The wrong team with the best AI tools underperforms the right team with adequate tools, consistently.

Third: what data do I actually have, and what data am I still pretending to have? This is a question of intellectual honesty that many organisations find genuinely uncomfortable, because the answer is usually more sobering than expected.

The rest — tools, platforms, agentic workflows, LLM architectures — is execution. Important, absolutely. But secondary to strategic clarity, and entirely dependent on it.

The operating system of venture building has changed, as McKinsey correctly states. But the operating system of leadership has not. It still runs on judgment, intellectual honesty, and the courage to ask inconvenient questions before committing capital and talent to a direction. AI amplifies whatever is already there. Which means the quality of what is already there has never mattered more.

Source reference: McKinsey Business Building, "How to build businesses faster and better with AI", March 2026.

Silvio Fontaneto is a Strategic Advisor and Executive Search specialist in Digital, Tech and AI. Author of "Stop Fearing AI" and the "The Vector" trilogy. For over 35 years he has supported organisations and leaders through technological transformation.

Explore the full Knowledge Hub: www.silviofontaneto.com 📬 Subscribe to the "AI Impact on Business" newsletter for weekly analysis: LinkedIn Newsletter Read more: www.silviofontaneto.com/articles (filter: AI Strategy)