The Forgotten Playbook That Could Save AI From Itself
If you’ve been in operations long enough, this moment feels familiar. AI promises transformation, but on the ground it looks like the late-90s ERP boom all over again: massive hype, half-baked pilots, and zero P&L traction.
Studies show enterprise AI failure rates are already north of 80%. For example, one recent study from MIT reports roughly 95% of generative-AI pilots never deliver measurable business value. And it’s not because the technology is flawed. It’s because we’ve handed the next major P&L transformation to the wrong department.
We're repeating one of the most expensive mistakes of the last 30 years: treating a fundamental business transformation as an IT project. The problem isn’t the technology. It’s ownership.
The Historical Mirror: Governance Failures We Should’ve Learned From
Between 1996 and 2008, four Fortune 500 companies collectively burned through hundreds of millions of dollars trying to modernize their operations through technology. Each became a case study in what happens when ownership and accountability live in the wrong place.
Each one of these failures shared the same root cause: the project wasn’t owned, governed, and led by the business function responsible for the outcome.
The pattern hasn’t changed. Today’s AI failures mirror ERP’s problems: misaligned ownership, poor data governance, and failure to scale. In fact, recent industry data shows that less than 10% of companies have embedded, mature AI governance frameworks. The names changed, but the risks didn't.
The Field Directive: A Blueprint for Success
The playbook for fixing this already exists. Cisco Systems wrote it in 1994 and every successful large-scale transformation since has followed the same model.
Facing an operational crisis that threatened the company, Cisco executed its ERP implementation on time and under budget. This wasn’t luck. It was a meticulously engineered governance model.
Here is the model, which should be your directive for AI:
Business-Led, Not IT-Led: The project was framed as a strategic business imperative, not a mere IT upgrade.
Active Executive Sponsorship: The steering committee was composed of senior leaders from all functional domains, specifically the best and brightest from the business units.
Disciplined Scope: The leadership team enforced minimal customizations, slashing risk and complexity.
This blueprint is timeless. Lenovo’s recent S/4HANA transformation followed the same path, framing it as a strategic move to reshape the company’s digital foundation and shift its operating model to “as a service”. The technology enabled the strategy. It didn’t define it.
That same governance model will separate the AI winners from the next generation of transformation failures.
The Mandate for Operations Leaders
Your AI initiative isn't a technology challenge to be delegated. It's an organizational design and business process challenge engineered for P&L impact, and you, the operations leader, must own it.
Innovation isn’t about who writes the code. It’s about who owns the outcome.
Before you approve another dollar of AI spend, you need answers to these questions:
Who owns the P&L for this outcome?
Who from the business is on the steering committee with the authority to make binding decisions?
Who is the Executive Sponsor who owns the business case and will see it through to value realization?
If AI still reports to IT in your org chart, you’re already behind. The good news is that correcting the ownership structure is the fastest way to get ahead.