On November 5th, McKinsey dropped their 2025 AI report, and it validates something I’ve been seeing in the trenches every single week.
The gap between AI hype and business reality isn't just wide; it’s massive.
According to the report, nearly 90% of companies are now using AI. That sounds like victory, right?
Wrong. 67% remain stuck in "pilot purgatory."
When I read data like this, I don't just see percentages. I see the faces of frustrated founders and executives I talk to. They bought the enterprise licenses, they ran the weekend hackathons, and they have Slack channels full of cool prompts. But when they look at their P&L at the end of the month, the impact is effectively zero.
Here is what the data tells us, why it’s happening, and the uncomfortable reality of what it takes to fix it.
The "Impact Gap" is Bleeding You Dry
The McKinsey survey of nearly 2,000 executives paints a stark picture: AI adoption is mainstream, but AI transformation hasn't started.
Three numbers tell the whole story:
Only 33% of companies have moved beyond initial pilots.
While 64% say AI drives innovation, just 39% report any real EBIT (earnings before interest and taxes) impact.
51% of organizations have already seen AI backfire, mostly due to inaccuracy.
This is what I call the "Impact Gap." It’s the expensive chasm where organizations bleed time, money, and morale chasing cosmetic productivity tweaks instead of real value.
As we often discuss here at First AI Movers, if your focus is still on "proving AI works," you are already behind. The technology works. The failure point is almost always human workflow integration.
What the Top 6% Do Differently
The report identified high performers—the top 6% of companies reporting at least a 5% EBIT contribution from AI.
They aren't just smarter; they behave differently. If you want to move out of pilot purgatory, you need to adopt their playbook. They do three things everyone else ignores:
They don't pave cow paths. Average companies try to slap AI onto broken, legacy processes to make them faster. High performers rebuild the workflow entirely assuming AI exists. They don't use AI to write emails faster; they redesign their systems so that 50% of those emails never need to be written in the first place.
They set innovation goals, not just efficiency targets. Leading companies allocate over 80% of their AI investment to transforming core operations and creating new offerings. If your main AI goal is "saving 2 hours a week per employee," you are aiming too low to see ROI.
Leadership is obsessed with execution. This is the hardest truth: companies with engaged leadership are 3x more likely to scale AI successfully. AI strategy cannot be delegated to the IT department. If the C-suite isn't personally involved in how workflows are changing, the initiative will die in the middle management layer.
Why I Built My Business Around "The Boring Stuff"
Reading this report reminded me exactly why I run First AI Movers and offer my advisory services.
It’s frustrating to see smart leaders get seduced by the shiny objects of AGI and sci-fi promises, only to crash against the rocks of poor governance and undefined workflows.
I didn't start selling my services because I wanted to capitalize on a trend. I started selling them because I saw too many organizations failing at the "last mile" of AI implementation.
I realized that companies don't need more tools; they need a framework for the unsexy, difficult work of redesigning how work gets done. They need governance that runs alongside deployment, not as an afterthought when things break.
That gap between "buying a tool" and "redesigning a workflow" is where my expertise lives. It’s why I insist on deep-dive audits before we deploy anything. It’s why I focus so heavily on retraining your team, not just fine-tuning models.
If you are tired of being in the 67% stuck in pilot purgatory and are ready to do the hard work required to join the 6% actually seeing EBIT impact, let’s talk.
The Takeaway
Stop measuring your AI success by how many seats of ChatGPT you’ve bought. Start tracking how fast your core workflows are adapting to this new reality.
The companies moving into value creation today aren't chasing magic. They're solving real, boring business problems with narrow AI—and they have the discipline to follow through on execution.
AI adoption is now table stakes. Workflow transformation is where the next decade's winners will be built.
Let's do this—together.
by Dr. Hernani Costa Founder, First AI Movers