In the face of the hype and the investments surrounding AI, companies are discovering an uncomfortable truth: even with AI tools, chatbots and the appointment of Chief AI Officers, the transformative impact that was dreamed of and promised does not materialize.
According to McKinsey, while 78% of organizations report using AI, more than 80% of them have observed no tangible impact on financial results. In the same vein, a BCG study reaches a similar conclusion: only 4% of companies create meaningful value through AI, while 74% struggle to show any impact.
According to PwC’s 2025 AI in business predictions published in 2024, 49% of technology leaders already asserted that AI was “fully integrated” into the business strategy of their company—a striking figure but rather misleading: if companies are integrating AI, they are doing so primarily to accomplish the same tasks as before… but with AI.
To make matters more complex, in the race to build the best Large Language Models, organizations focus on the quantity and quality of data, at the expense of critical governance gaps and data-related issues. But good news: it’s not too late… provided they adopt a bimodal AI strategy.
Bimodal AI Strategy: aligning optimization and innovation simultaneously
Yet, such a change of this magnitude doesn’t happen overnight. For instance, financial institutions have experienced rapid innovations and major operational changes—and under pressure to move fast. Therefore, the sector had to improve its existing operations while simultaneously introducing certain technological innovations.
Here lies the endorsement for the concept of “Bimodal IT,” as introduced by Gartner in 2014. The central principle of this bimodal strategy is to divide attention between two distinct modes:
• Mode 1 : stability, predictability and the maintenance of core systems.
• Mode 2 : agility, speed and innovation, often through new digital initiatives and customer-focused applications.
As the financial sector has shown, Modes 1 and 2 are not exclusive. The essence of bimodal thinking is not to view these modes as two successive steps, but to deploy resources, strategy and actions in parallel, to achieve bimodal outcomes: optimization of performance and productivity and value creation. And with AI, this framework becomes more relevant and useful than ever.
The race to AI: changing the wings in mid-flight
In the AI race, companies will have to manage their current business while using AI to unlock resources enabling organizational transformation, which in turn will drive the transformation of the company. For many businesses, the next two years will resemble “changing the wings in mid-flight.”
When defining AI objectives, the first step is to identify the main use cases that go beyond mono-modal approaches. These strategies can be defined as:
• Reduce costs : to implement automated solutions that boost efficiency, remove friction and optimize workflows.
• Increase revenue : to use AI to sustain production lines, optimize the supply chain and streamline processes.
• Improve satisfaction : whether it is the experience of employees, customers or partners, to facilitate interactions and increase satisfaction.
• Drive business transformation : enabling capacity for new business units, service offerings, products or geographic expansions.
These objectives are not mutually exclusive. Organizations can pursue one or more outcomes through a bimodal approach that permits parallel strategies.
As Dan Priest, Chief AI Officer at PwC US, notes: “The most successful companies will move from seeking AI use cases to using AI to realize the business strategy.” Like aerospace engineers changing the wings in mid-flight, organizations must maintain their operational altitude while simultaneously building the capabilities that will define their competitive trajectory.
Mode 1: securing critical operations
To evaluate a strategy, it is essential to identify the essential gears of a business and use AI to strengthen Mode 1 efficiency. This mode concentrates on establishing and optimizing the operations that are vital to the business, those that deliver the core value and form the foundation for innovation: what cannot fail?
What generates the main revenue? It could be a production line, a telecommunications network, sporting events, a citizen services system or any element that generates revenue and meets objectives.
In Mode 1, AI should serve to improve and optimize processes: what bottlenecks could it eliminate? What customer or employee frictions could it resolve? The application of AI to Mode 1 operations will reduce costs, increase revenues and enhance the operational experience.
Mode 2: innovate without limits
While Mode 1 delivers operational excellence, Mode 2 turns the gains in efficiency into opportunities for exponential growth. This is where technology enables speed, agility and greater innovation. The people, capital and capacity freed by the optimization in Mode 1 feed innovation in Mode 2.
This parallel approach enables launching new projects, geographic expansion, product development and the creation of new service lines. A strategic vision should steer Mode 2 initiatives, which will evolve with AI technologies and their capabilities.
Mode 1 yields linear gains through improved performance and cost reductions; Mode 2 unlocks exponential results.
These two strategies cannot thrive in isolation. Companies must strategically redeploy the time freed by AI rather than simply viewing it as a cost-cutting opportunity. Because we are at the heart of an AI revolution that offers unprecedented opportunities to combine optimization and innovation simultaneously.
An AI strategy is not about AI itself. It should help businesses achieve exponential performance and stand out in an increasingly competitive and volatile market. Organizations that master this dual approach will define the next generation of enterprises.
*Dave Wright is Chief Innovation Officer of ServiceNow
