Across every industry, organizations are investing heavily in the potential of artificial intelligence to reshape how they operate and grow. Nearly 80% of executives expect AI to significantly contribute to revenue by 2030, yet only 24% know where that revenue might come from.
This isn’t an awareness gap. It’s an architecture gap.
The companies already capturing AI’s value aren’t waiting to discover it through pilots and proofs-of-concept. They’re engineering it through deliberate choices about how work gets designed, how human and digital workers come together, and how productivity savings are reinvested.
From our work with enterprises across every major industry, a clear divide is emerging.
Some organizations are bolting AI onto legacy workflows and gaining marginal productivity. Others are redesigning how value gets created and building growth trajectories competitors can’t replicate. By 2030, this won’t be just a short-term positioning advantage. It will determine who remains in business. The difference comes down to three architectural choices that separate AI-first enterprises from everyone else.
Redesign Work Itself, Don’t Just Augment It
Most AI adoption fails because organizations are automating fundamentally broken processes. They’re making inefficient work more efficient—and wondering why transformation doesn’t happen.
AI-first enterprises start with a different question: If we were designing this work today with no legacy constraints, what outcome do we want? And what combination of human judgment and AI capability achieves that outcome best?
Nestlé provides a powerful example of a more than a centry-old global enterprise. The company isn’t just adding AI features to existing systems. They’re building an AI-powered enterprise architecture that understands their entire product ecosystem, supply chain, and consumer relationships in ways generic models never could. The goal isn’t incremental improvement—it’s the capability to deliver superior products faster while creating more personalized experiences for employees and customers.
Riyadh Air represents the opposite end of the business spectrum—a startup with no legacy constraints. But the principle is identical. The airline is building an AI-native operation from day one, with a unified architecture connecting operations, employees, and customers as a single intelligent system.
The insight both share is that the digital backbone isn’t just infrastructure. It’s the intentional architecture that allows humans and AI to work as integrated capabilities, creating adaptability that compounds over time.
Build Proprietary Intelligence, Not Just Access to Models
By 2030, everyone will have access to powerful AI models. The winners will have customized AI that knows their business better than any third-party AI possibly could.
L’Oréal isn’t just using AI to accelerate R&D. They’re building a custom AI foundation model trained on their proprietary formulation data, scientific research, and sustainability requirements. These models will give their scientists capabilities no competitor could replicate, enabling new scientific possibilities that wouldn’t otherwise exist.
In our recent survey, more than half of executives expect their competitive edge to come from AI model sophistication specifically. Sophistication also comes from proprietary data, custom models tuned to specific challenges, and continuous learning loops. Organizations need multi-model portfolios – some proprietary, some licensed, all integrated into architectures that evolve as quickly as their markets.
The most valuable companies won’t be those with the most data. They’ll be the ones that turn data into AI-driven decisions at scale, with intelligence competitors can’t mimic by simply licensing better models.
Engineer Growth Loops, Not Just Efficiency Gains
Most AI strategies fail because they treat productivity as the destination.
Executives expect AI to boost productivity by 42% by 2030. But if you bank those gains as cost savings, you’ve fundamentally misunderstood the opportunity. AI-first enterprises treat productivity as fuel by reinvesting efficiency gains into new products, services, and markets.
The pattern works like this: AI-driven efficiency frees capital and talent. That freed capacity funds innovation in new markets. New markets generate new data. New data trains better AI. Better AI creates more efficiency. The loop accelerates.
L’Oréal scientists won’t just make formulations faster—this speed will allow them to explore sustainable ingredients that were not economically feasible before. Nestlé isn’t just optimizing supply chains—they’re using those gains to build direct consumer relationships that transform how people interact with their products. Riyadh Air isn’t just building a new airline—they’re stripping out fifty years of legacy in a single stroke that will define the next decade of aviation.
This creates exponential divergence. While laggards optimize margins, leaders accelerate into new markets, building capabilities that compound. By 2030, the gap won’t be measurable in productivity percentages. It will be measurable in entirely different business models.
The Questions That Determine Who Wins
The next era of growth won’t be predicted. It will be engineered. Leaders must answer three uncomfortable questions now:
- If we redesigned our operations with AI-first principles, what would we stop doing entirely? Not what would we do faster, rather, what would we eliminate? Most organizations discover that 30-40% of their workflows exist solely to compensate for constraints that AI removes. But elimination requires courage optimization avoids.
- What proprietary intelligence could we build that competitors can’t replicate? Not what AI can you license, but what AI could you engineer—built on the human expertise unique to your organization—that is so deeply tuned to your business that competitors would need a decade to catch up?
- Are we banking productivity gains or reinvesting them into growth loops? Cost savings are finite, but growth loops are exponential. Which one is your strategy building?
By 2030, the companies that can answer these questions won’t just be more productive. They’ll be operating in markets competitors didn’t know existed, with capabilities competitors can’t build, and business models competitors can’t afford.
The real risk isn’t moving too fast on AI. It’s engineering too slowly while competitors redesign the game entirely.
The opinions expressed in Fortune.com commentary pieces are solely the views of their authors and do not necessarily reflect the opinions and beliefs of Fortune.
The post I lead IBM Consulting, here’s how AI-first companies must redesign work for growth appeared first on Fortune.
