Don’t believe everything you hear about AI: the wins are real
You’ve seen the headlines, likely splashed across industry reports and LinkedIn posts: “95 per cent of AI Projects Fail.”
It’s a statistic designed to grab your attention, and it works. It fuels skepticism and gives leaders a perfectly valid reason to hit the brakes on AI investment.
But it’s also one of the most misleading narratives in business today. The problem isn’t the statistic itself, but what it’s measuring. It’s not a failure of the technology. It’s a failure of implementation.
It’s what happens when you give a Gen Z driver the keys to a manual-transmission Ferrari and then act surprised when they stall it, strip the gears, and declare the car broken.
The car isn’t broken. The driver just never learned how to drive stick.
This is the critical distinction that the clickbait headlines ignore. The supposed “failure” of AI is a symptom of a profound mismatch between a revolutionary technology and the legacy processes it’s being forced into. It’s time to reframe the conversation from “Does AI work?” to “How do we learn to work with AI?”
The dealers who master this cultural and operational shift will not just see incremental improvements; they will unlock a new, sustainable competitive advantage.
Let’s explore the real story behind the headlines and what the MIT research actually found.
To understand what’s really happening, we need to look past the headlines and into the details of the research itself. The current confusion mirrors a well-documented phenomenon known as the “productivity paradox.”
History is filled with examples of transformative technologies; from the electric motor to the personal computer, that took years, sometimes decades, to show up in broad economic productivity numbers.
Why? Because early adopters simply used the new technology as a direct substitute for the old one. Factory owners would replace a single, large steam engine with a single, large electric motor, failing to realize that the true power of electricity was in deploying smaller motors throughout the factory, completely redesigning the workflow around them. They changed the tool, but not the work.
We are seeing the exact same pattern with AI. A landmark 2023 study co-authored by MIT’s Erik Brynjolfsson, titled “Generative AI at Work,” provides a crystal-clear example.
The study, conducted in a real-world Fortune 500 company’s customer service centre, found that deploying a generative AI assistant had a stunning impact on the ground. Worker productivity, measured by issues resolved per hour, increased by an average of 14 per cent.
The report is filled with incredibly positive findings. The biggest beneficiaries were new and lower-skilled employees, who completed tasks faster and improved the quality of their customer interactions.
The AI acted as a great equalizer, embedding the best practices of the most seasoned veterans and making them available to everyone. This has enormous implications for an industry like ours, which constantly battles high staff turnover. Imagine a new BDC agent or service advisor becoming competent in weeks, not months.
So, if the technology is so effective at the task level, why the pessimistic headlines?
Because these micro-level gains haven’t yet translated into macro-level economic growth across the entire economy. But as the history of the electric motor shows, that’s to be expected.
The value of a general-purpose technology isn’t unlocked until the organizations using it undergo the necessary, and often difficult, process of reinvention. The evidence is clear: the AI works. The challenge is that our organizations are not yet ready for it.
Your Playbook: How to “move” to the land of AI
This brings us back to the sales manager’s brilliant analogy: to succeed, dealers must stop acting like software buyers and start acting like expats moving to a new country. This isn’t about a simple transaction; it’s about deep, strategic immersion. This journey rests on three foundational pillars.
Pillar 1: Learn the Language (Build Data & AI Literacy)
You cannot hope to thrive in a new country without a basic grasp of the local language. For AI, that language is a blend of core concepts and the data that fuels them. This is where AI literacy becomes the foundational pillar of your entire strategy. It’s not about turning your managers into data scientists; it’s about equipping them with the critical thinking to ask the right questions.
A crucial part of this literacy is understanding that not all AI is created equal. The old mantra was “garbage in, garbage out,” meaning an AI was only as good as the clean, structured data you fed it. While that principle remains vital, some sophisticated AI platforms are now capable of cleaning, interpreting, and structuring messy data.
But this is precisely why literacy is so important. You must be able to distinguish between a simple AI tool that requires pristine data from your DMS, and a more advanced agentic platform that can make sense of varied data streams.
True literacy is the ability to discern the difference, to know which tool is right for which job and to understand the health of your data in relation to the specific AI you deploy. It’s the difference between blindly trusting a tool and strategically directing it.
Pillar 2: Understand the Culture (Redesign Your Processes)
In a new country, you must adapt to the local customs and ways of life. Similarly, you cannot drop a powerful AI tool into a broken or outdated process and expect a miracle. You will only be making the old, inefficient process run faster. True transformation requires a willingness to redesign your workflows around the capabilities of the AI.
This is the “cultural immersion” phase. Ask the hard questions. How does an AI-powered sales assistant fundamentally change the role of the BDC and the nature of its follow-up cadences?
How does a predictive maintenance tool in the service lane shift the service advisor’s role from reactive to proactive?
According to a 2024 review in CIRP Journal of Manufacturing Science and Technology, successful AI adoption hinges on this deep alignment of technology with organizational effectiveness. It’s about co-creating new, more effective ways of working with the technology as a partner.
Pillar 3: Live Like a Local (Foster a Culture of Experimentation)
The final stage of acculturation is feeling comfortable enough to explore, try new things, and occasionally make mistakes. In the organizational world, this is known as “psychological safety.”
It is perhaps the single most important ingredient for successful AI adoption. Your team must feel empowered to experiment with AI tools without fear of failure.
This means starting with small, controlled pilot projects in specific areas of the dealership. It means measuring everything to understand what works and what doesn’t. And most importantly, it means celebrating the learnings that come from these experiments, not just the wins. As academic research on organizational learning consistently shows, firms that cultivate this capacity for experimentation are the ones that successfully absorb and apply new technologies. This is how you build the organizational muscle required for a full-scale transformation.
The Payoff: From Lagging Indicator to Leading Edge
This journey of reinvention follows a predictable pattern known as the “J-Curve.” There is an initial dip in productivity as the dealership invests time, energy, and resources into training, process redesign, and experimentation. This is the period of disruption. But for those who persevere, this dip is followed by a sharp and sustained rise in productivity, efficiency, and profitability that far surpasses the initial investment.
The dealers who master this cultural and operational shift will not just see incremental improvements; they will unlock a new, sustainable competitive advantage.
They will benefit from faster onboarding, higher employee retention, more efficient operations across every department, and a data-driven culture that is agile enough to adapt to the next wave of disruption. The prize for getting this right is not just surviving the AI revolution, but leading it.
The keys are in your hand
Let’s return to that “95 per cent failure” statistic. The dealers who see that headline and feel a sense of fear are missing the point. The leaders who will win the next decade will look at that same number and see the single greatest competitive opportunity available today.
Because that 95 per cent isn’t a technology failure rate… it’s an organizational change failure rate. It’s a measure of how many businesses are trying to bolt a supercomputer onto a horse-and-buggy process.
By embracing the role of the student, by learning the language, culture, and processes of this new era, you position your dealership to be in the five per cent that gets it right. The keys to the supercomputer are on the table for everyone. It’s time to learn how to drive.
