CEOs Aren’t Thinking Big Enough with AI

Many companies are using AI to improve what already exists—automating tasks, speeding up workflows, helping teams work a little faster.

Those are valuable steps forward, but too many companies stop there before asking the bigger questions. What if AI could change how the business works in fundamental ways? Or open the door to entirely new products and streams of revenue?

The most successful CEOs are already asking these questions, pushing forward with AI and looking beyond automation to reshape core functions with GenAI and agentic systems. Their businesses are inventing new, AI-powered offerings where their customer relationships, expertise, or data give them an edge. In some cases, those moves are unlocking $200 million or more in new value.

“AI is our electricity moment,” says Jessica Apotheker, a BCG managing director and senior partner and chief marketing officer of BCG and BCG X. It’s an apt metaphor: When factories first began swapping gas lamps for light bulbs, it added a few hours of productivity. But when factory owners took the bigger step of redesigning their factories for electric machines, the shop floor suddenly became a hundred times more effective.

“CEOs have to decide if they’re chasing small wins,” says Apotheker, “or if they are going to reimagine how the business transforms for the 100x payoff.”

That choice lies at the heart of what we call the imagination gap: the distance between what most companies are attempting with AI and what the technology makes possible. Closing that gap requires not just vision, but the readiness to rethink what the business can become—and the willingness to act before the path is fully clear.

The Barriers to Imagination

Many companies are struggling to move beyond the experimentation phase of AI. That’s not because the technology isn’t ready, but because the company’s leadership isn’t.

In a time of economic uncertainty and heightened scrutiny, there’s even less room for initiatives and investments that seem loaded with risk. Blue-sky thinking sounds good in theory, yet in practice it makes many leaders uneasy. There’s no clear owner of innovation—no “chief reimagineer” in the org chart. Even when bold ideas do surface, they’re often blocked by bureaucracy, risk aversion, or the pressure to deliver the next quarter’s numbers.

“There’s a lack of willingness among leaders to look beyond the quarter or the year and say, ‘What are my next frontiers, and how am I going to invest?’” says Lee Robertson, a BCG managing director and partner.

As a result, many organizations aim low. They focus on incremental automation, which is useful but leaves so much potential untapped. Often, leaders don’t even ask what new approaches are possible, in part because many lack a strong grasp of what AI can actually do.

In their defense, AI is evolving faster than most companies can keep up with. By the time a CEO understands the capabilities of the latest model, the technology has already moved on. But that doesn’t mean it’s too early to act. Today’s tools are powerful enough to drive meaningful transformation. “With current technology, there’s nothing that cannot be reimagined,” says Nacho Hafner, a BCG managing director and partner.

Whether reshaping how the business works or inventing something entirely new, an AI transformation begins with a mindset shift. Instead of asking what’s safe to try, the CEO must be willing to imagine a future-state business that doesn’t yet exist—say, a new digital interface for customers—and then guide the organization backward from there. That means getting comfortable with risk and learning to move forward, even with unknown variables.

Two Paths to Value

Companies are taking a range of approaches to AI. Deploying off-the-shelf models can streamline tasks and speed up work. The upside is evident right away: teams gain confidence with the technology and the company’s investments show quick wins.

But when the goal is to go bigger by reshaping functions and imagining new business models, systems-level thinking becomes vital. As the technology evolves, it gives leaders the freedom to revisit what once felt impossible.

“When you plan to implement GenAI’s capabilities,” says Robertson, “you suddenly have permission to ask, ‘What if I could eliminate all my traditional bottlenecks?’”

Reshaping for Advantage

In many companies, a typical performance workflow—like optimizing a marketing campaign—still unfolds in a slow, linear way. Analysts pull data into dashboards, media managers interpret results, and cross-team coordination adds friction. It’s a process that can take a week or more, and much of the effort goes into assembly, not impact.
 
With GenAI, teams can reshape that same workflow end-to-end. Data flows automatically into live models. Draft recommendations come ready for teams to refine and approve. What once took several days or weeks can now happen in less than a day, sometimes within hours.

“To truly reshape a function, leaders must encourage teams to think boldly—beyond incremental gains,” says Arnaud Bassoulet, a BCG managing director and partner. “That means challenging teams to reimagine entire processes from the ground up, embracing disruptive thinking, and exploring new models for collaboration, operating rhythm, and organizational design.”

It’s also a good way to transform the most high-stakes areas of business, including R&D. In consumer goods, for example, teams can use AI to speed up development cycles in a leaner operating model. The revamped process can deliver more breakthrough products while freeing researchers to focus on high-impact discovery and innovation. Key milestones like insights-to-concept, document generation, and consumer feedback loops can be supported by AI, with human expertise still central to the process.

Consider a vital R&D process like a multiyear stability study. It usually takes months, including four weeks of drafting and two weeks of approval. With AI, companies can eliminate slower parts of the process and cut the entire production workflow to less than two weeks—all while maintaining the quality of R&D concepts.

“It’s important for the CEO to create the right incubation space for this sort of transformation,” says Apotheker. “From day one, push for teams to think about how to reinvent their workflows with AI. Look for business impact on a smaller scale and then scale fast when the benefits are clear.”

Inventing What Doesn’t Exist (Yet)

Where the Reshape approach offers a new view of how work gets done, Invent imagines what new value a business could bring to market.

“It’s the same muscle—just applied to a more ambitious goal,” says Hafner. “Often, once companies see what’s possible internally, the next question becomes: If I can do this for myself, can I offer it to my customers?”

Unlike Reshape, which focuses on internal transformation, inventing a new AI-based product or service points your energy toward market-facing initiatives. “Invent is a source of future revenue—but it’s also a potential source of competitive advantage,” said BCG’s Nicolas de Bellefonds, a managing director and senior partner.

A Personalized Beauty-Product Experience. L’Oréal shows how the combination of bold ambition and awareness of customers’ needs can spark entirely new kinds of value. In beauty, especially in digital and self-service channels, shoppers often face an overwhelming number of choices with little personalized advice. L’Oréal saw an opportunity to close that gap with a novel kind of experience.

The company built Beauty Genius, an AI-powered, always-on virtual advisor. Customers can use the tool to receive personalized product recommendations, from a new shade of eye shadow to a new skin care routine, directly on their device. The experience uses a combination of AI capabilities: vision AI for selfie-based analysis, GenAI for tailored recommendations, and behavioral data to personalize each interaction.

“It was a dream to bring service to a self-service world,” said Othman Bennis, L’Oréal’s chief digital and marketing officer. “We did what we always do when we create products—we listened to consumers.”

Rolled out across multiple markets, the service has reached reach over 10 million consumers and driven 30% to 70% higher engagements rates than static product content.

Fleet Management Insights in Real Time. To invent successful new products, a CEO can also look at a company’s strongest assets from a new angle—something logistics company Penske did well.

“Companies are sitting on decades of customer data—feedback, interviews, surveys, and increasingly, real-time data from sensors and connected systems—that used to be hard to capture or analyze at scale,” says Dutch MacDonald, a BCG managing director and partner. “Now, with AI, you can mine that data, create valuable insights, and quickly test against both real and synthetic users.”   

For Penske’s truck leasing business, a core challenge was visibility. Fleet managers lacked real-time insights into performance drivers like fuel efficiency, vehicle downtime, and maintenance costs. To close that gap, Penske built Catalyst AI, a platform that lets users—fleets of all sizes that need to maximize every cent on fuel and every minute on the schedule—compare real-time vehicle data against the metrics of similar or top-performing vehicles. The solution can also analyze individual truck or branch location performance.

The platform was codeveloped with customers, whose early feedback helped shape the user experience and data features. In the process, Penske transformed an internal asset into a commercial-grade AI product, strengthening its position as a data-driven leader in a field where knowledge makes a difference.

How CEOs Can Raise Their Ambition—and Build Toward It

The AI-powered workflows, the new products, the $200 million bets—they all start with the CEO. Only the chief executive has the mandate, authority, and trust to push the business to build new workflows and imagine what doesn’t exist yet. CEOs can get moving with several strategic choices.

Create the right incubation space for AI initiatives to flourish. Avoid big task forces; leaders should build focused teams and keep them lean, especially when reshaping existing workflows. Executive sponsorship of AI transformation goes a long way, too, so make it clear to the business that you support blue-sky thinking around AI.

Structure your AI ambition like a portfolio. AI transformation is not a single bet, but rather a sequence of moves to build momentum and value. Target returns of two to three times your investment from a mix of AI initiatives—clear internal plays balanced with a few bold, high-upside bets. “Some of these moonshots won’t work,” says Robertson. “That’s okay.” The goal is to learn quickly, govern clearly, and keep the ambition alive across multiple time horizons.

Use a test-and-run loop to imagine new ways of working—or of reaching customers. AI innovation doesn't follow a traditional waterfall. Companies should continually test their assumptions and be prepared to change course if necessary. Three steps make up the testing plan:

1. Discover. Identify shifts in customer behavior and map where your strongest assets can make a difference. Decide where you have the right to win. “When starting, the CEO should ask: where does AI change the economics of your value chain?” says de Bellefonds. “What data or AI assets could you monetize, not just use?”
2. Validate. Create prototypes rapidly. Be ready to test ideas with customers, because their input is invaluable. When you get traction with an idea, scale it quickly. Leave behind anything that doesn’t get customers or employees to notice. 
3. Build. Move to scale only when the risk is low enough to justify real investment. As new products evolve, keep the customer involved throughout.

Lock in business leadership support. Imagining new business models calls for both strategic sponsorship and cultural support. Find a senior business owner—a division head or product leader—who deeply understands the customer and doesn’t balk at barely proven ideas. This person can be a key ally for the CEO who’s pushing for transformation. They must be open-minded; many AI builds don’t fit cleanly within existing architecture or operating models.

“AI innovation runs against how most leaders are wired—to protect how things run today,” says Robertson. “That’s why you need someone willing to go along with you who says ‘Let’s try it’ instead of ‘This doesn’t fit.’” 

Set your sights beyond what feels feasible today. A frequent pitfall for leaders is letting today’s capabilities define tomorrow’s ambition. For the CEO, this means setting direction ahead of the curve and anchoring goals to where the business needs to go, not what’s easy in the current moment.

“The technology is moving faster than your ability to transform,” says de Bellefonds. “If you aim where the tech is now, you’ll arrive too late.”  

Build AI talent and a learning culture now, before you fall behind.  In every sector, there are already AI-first companies emerging with lighter infrastructure, faster build cycles, and fewer constraints. A savvy entrepreneur can board a flight and build the next breakthrough app before the plane touches down.

Those disruptors are now your competition, and employees must help you keep pace. “Companies can’t fast follow,” says Apotheker. “If you’re not already building the talent and culture, you won’t be able to keep up.” 

The CEOs that succeed won’t just optimize with AI. They’ll use it to imagine entirely new business models and to rethink how value is created, how revenue flows, and how customers are engaged. That means moving with urgency, investing with ambition, and building the conditions for bold ideas to scale. This isn’t just about GenAI—it’s about who’s willing to lead the next era of reinvention.