Before You "Do AI," Decide What It's For

The rush to “do AI” has become the new corporate reflex. Every board deck, strategy summit, and investor briefing now features an AI roadmap, yet few leaders can articulate why they’re doing it beyond “we have to.” 

A recent report from MIT’s Project Nanda reveals that 95% of generative AI pilots fail to deliver measurable business impact. Billions of dollars have been spent chasing proofs of concept that never scale. The problem isn’t technological. It’s strategic. 

The lesson from MIT’s analysis is counterintuitive: it’s friction, not speed, that separates the 5% who succeed from the 95% who don’t. The organizations that win aren’t eliminating friction; they’re designing for it. Resistance, feedback loops, and workflow redesign are not obstacles to avoid. They are signals of real transformation taking hold.   

Why Most AI Efforts Stall  

When AI efforts stall, the causes are familiar: unclear vision, undefined outcomes, and an early pivot to tools. These show up in three predictable missteps:  

1. Starting with the tool, not the outcome. Pilots kick off before the problem it will solve is named. The result is motion without aim: activity that looks impressive but doesn’t move a needle that matters. 

2. Bolting on instead of building in. Pilots run parallel to real work, adding tools and dashboards while decision rights, handoffs, and routines remain unchanged.   

3. Avoiding resistance. Leaders try to make adoption seamless, skirting the hard parts that surface reality. The tension isn’t the problem; it’s the signal to redesign.  

How to Turn Ambition into Operating Reality  

For most leaders, the hardest part of AI isn’t the technology — it’s the translation. How do you take an idea as vast as “AI transformation” and make it meaningful, measurable, and credible inside your business?  

The leaders who get this right start by resisting the urge to define AI as a program. They define it as a capability, one that exists to solve something specific. They ask sharper questions:  

• What problem are we trying to solve that AI might finally let us address at scale? 

• What decisions do we want to make faster or better?  

• What kind of organization do we need to become for that to happen?  

Clarity on those questions turns aspiration into direction. From there, impact comes not from launching faster, but from learning deliberately. The organizations that succeed don’t treat friction as failure — they use it as feedback. Each challenge that surfaces, whether a process gap or a mindset barrier, becomes a clue about what needs to evolve next.  

Leaders play a distinct role in that process. They frame the “why,” connect it to business priorities, and hold the organization steady when the novelty wears off and the real work begins. They model curiosity instead of certainty, treat pilots as learning cycles rather than showcases, and create space for employees to explore how AI can enhance their work rather than displace it. 

This is how ambition turns into operating reality: not through perfect roadmaps or bigger models, but through clarity of intent, disciplined learning, and leadership that keeps the organization grounded in purpose while it experiments toward progress.   

Looking Ahead  

The real differentiator won’t be access to the same models everyone else has. It will be the purpose that guides them. The next era of transformation will reward leaders who pause to ask the right questions, define what matters most, and stay close enough to the work to see what’s actually changing.  

Before you “do AI,” decide what it’s for. Because purpose, put into practice, is what turns disruption into direction. 

AI Tools We’re Testing 

AI will not replace leadership; it will demand more of it. It will ask for judgment, for patience, and for the courage to make the abstract tangible, to turn vision into structure, and structure into progress 

We’re introducing a new segment to our newsletters where we share highlights from the AI tools we’re testing — what’s showing real potential, and where we’re still learning. These quick reviews are independent reflections from our own experimentation, not endorsements or promotions. 

• A tool that hit the mark: Boardy.ai: Boardy.ai is an AI “Super Connector” that chats with you on WhatsApp, learns your goals, and finds useful contacts, potential clients or partners through LinkedIn and its own network. It’s fast and feels like having a digital networking assistant that understands what you’re looking for and supports you along the way. 

• A tool we’re still assessing: Beautiful.ai: Beautiful.ai offers polished templates and AI-assisted layouts that make creating clean, consistent decks fast and intuitive. For creating simple slide structures, it saves time. However, for more advanced deck development, we found the structure required extra manual adjustment that added more time on the back end. We’re continuing to test where it fits best — likely as a complementary tool for fast drafting rather than end-to-end presentation development.  

Take the Next Step  

If you’re preparing your organization to embed AI, redesign how work gets done, or lead through broader transformation, now is the time to step back and define what success truly looks like. 

Join us for our November 19 webinar: Strategic Restructuring in an Era of Frequent Layoffs: Protecting Key Talent, Trust, and Performance.

We’ll unpack how to align structure, leadership, and purpose to protect performance through disruption.  

Hope to see you there!

Andrea