Inspiration to how you continously achieve value from your AI investments​

Companies will spend $2.7trillion on AI in 2026 [Gartner, 2026]. Only 6% attribute 5% or more of theirEBIT to it [McKinsey, 2026].

The conversation has shifted from what AI can do to when it will pay back. Headlines now ask whether AI is a bubble, and investors warn that the bets must pay off – and soon [Bloomberg, 2026]. The question many companies are starting to ask is whether their AI investments return enough to justify continued funding – or whether the focus should shift.

Companies will spend $2.7trillion on AI in 2026 [Gartner, 2026]. Only 6% attribute 5% or more of theirEBIT to it [McKinsey, 2026].

The conversation has shifted from what AI can do to when it will pay back. Headlines now ask whether AI is a bubble, and investors warn that the bets must pay off – and soon [Bloomberg, 2026]. The question many companies are starting to ask is whether their AI investments return enough to justify continued funding – or whether the focus should shift.

Despite the questions on return, the road to value is becoming clearer – and it leads beyond the most obvious initiatives. Successful companies are set up to launch, test, and scale AI initiatives repeatedly. High performers are far more likely to have redesigned workflows around AI (about 75% vs. 25% of others) and to show visible senior leadership commitment [McKinsey, 2026].

As the technology improves, so do the opportunities that it unlocks. But as AI costs increasingly scale with usage [J.P.Morgan, 2026], every initiative carries a running cost, not just a build cost.It becomes critical to prioritize efforts on the initiatives that have the biggest impact.

This paper explains how to move from exploring AI initiatives to realize continuous AI value, through three pillars: Design, Build and Transform.

AI transformation is an innovation discipline, not a technology discipline. It shifts the focus from what can be built to what will bring the most value, with initiatives prioritized based on results.

Despite the questions on return, the road to value is becoming clearer – and it leads beyond the most obvious initiatives. Successful companies are set up to launch, test, and scale AI initiatives repeatedly. High performers are far more likely to have redesigned workflows around AI (about 75% vs. 25% of others) and to show visible senior leadership commitment [McKinsey, 2026].

As the technology improves, so do the opportunities that it unlocks. But as AI costs increasingly scale with usage [J.P.Morgan, 2026], every initiative carries a running cost, not just a build cost.It becomes critical to prioritize efforts on the initiatives that have the biggest impact.

This paper explains how to move from exploring AI initiatives to realize continuous AI value, through three pillars: Design, Build and Transform.

AI transformation is an innovation discipline, not a technology discipline. It shifts the focus from what can be built to what will bring the most value, with initiatives prioritized based on results.

We see three main challenges standing between AI experimentation and AI value: no clear direction, no structure for turning ideas into results, and an organization not ready to work differently. To bridge this gap, we use our three-pillar framework to encourage innovation, rapid validation, and continuous enablement of the organization.

1. Design: an operating model that sets the ambition, structures how the portfolio is managed, and plans how to enable the organization.
2. Build: a way of working that prioritizes use cases on expected value through ongoing stage gates, so only the best progres
3. ‍Transform: organizational enablement that evolves how leaders and employees work with AI.

‍

However far along the AI journey you are, these fundamentals will help you focus your efforts where they matter most – and, if done right, will make the results repeatable.

We see three main challenges standing between AI experimentation and AI value: no clear direction, no structure for turning ideas into results, and an organization not ready to work differently. To bridge this gap, we use our three-pillar framework to encourage innovation, rapid validation, and continuous enablement of the organization.

1. Design: an operating model that sets the ambition, structures how the portfolio is managed, and plans how to enable the organization.
2. Build: a way of working that prioritizes use cases on expected value through ongoing stage gates, so only the best progres
3. ‍Transform: organizational enablement that evolves how leaders and employees work with AI.

‍

However far along the AI journey you are, these fundamentals will help you focus your efforts where they matter most – and, if done right, will make the results repeatable.

An operating model for AI development defines a recipe for success and guides every initiative from first idea to scale.

An AI operating model consists of three core components:

• • Aspiration: Setting a vision tied to business outcomes rather than tool adoption, formalizing the mandate, and defining the governance to steer it.

• • Portfolio: Identifying opportunity spaces and defining how to manage and execute the initiatives in the portfoli

• • Enablers: Assessing and defining the capabilities, skills, tools, and processes needed to deliver on the aspiration.

An operating model for AI development defines a recipe for success and guides every initiative from first idea to scale.

An AI operating model consists of three core components:

• • Aspiration: Setting a vision tied to business outcomes rather than tool adoption, formalizing the mandate, and defining the governance to steer it.

• • Portfolio: Identifying opportunity spaces and defining how to manage and execute the initiatives in the portfoli

• • Enablers: Assessing and defining the capabilities, skills, tools, and processes needed to deliver on the aspiration.

Build is a way of working that, like a venture investor, makes many cheap early bets and concentrates investment on those that prove their value. It continuously evaluates initiatives – failing fast and scaling what works – across four stages:

1. Discovery gathers business needs into a long list of opportunities.
2. Validation tests the long list against desirability, viability and feasibility to narrow it down.
3. Build delivers a minimum viable product (MVP) that tests the value of the prioritized initiative.
4. Scale rolls out the initiative across the organization.

Stage gates are planned checkpoints for evaluating initiatives. Based on performance against predefined KPIs, each gate has the mandate to proceed as planned, pivot by incorporating new input, or park the initiative until decided otherwise.

Build is a way of working that, like a venture investor, makes many cheap early bets and concentrates investment on those that prove their value. It continuously evaluates initiatives – failing fast and scaling what works – across four stages:

1. Discovery gathers business needs into a long list of opportunities.
2. Validation tests the long list against desirability, viability and feasibility to narrow it down.
3. Build delivers a minimum viable product (MVP) that tests the value of the prioritized initiative.
4. Scale rolls out the initiative across the organization.

Stage gates are planned checkpoints for evaluating initiatives. Based on performance against predefined KPIs, each gate has the mandate to proceed as planned, pivot by incorporating new input, or park the initiative until decided otherwise.

Redesigning how work gets done, rather than layering AI on top of existing workflows, is what separates organizations that capture value from those that do not. This means continuously equipping leaders and employees to drive their part of the AI journey. For leaders, that means prioritizing hard –prioritizing initiatives with a value driving business case over initiatives with sponsorship – and learning to lead a hybrid workforce of people and agents. For employees, it means literacy beyond prompt training: using AI in real work, trusting it enough to rely on it yet checking it, and redesigning their own processes with it.

The aim is coherence. The initiatives you prioritize should match your organization’s maturity, so that initiatives and workforce evolve in parallel. Pilots that fail often do so not because the models are weak, but because they land in an organization that cannot yet absorb them.

Redesigning how work gets done, rather than layering AI on top of existing workflows, is what separates organizations that capture value from those that do not. This means continuously equipping leaders and employees to drive their part of the AI journey. For leaders, that means prioritizing hard –prioritizing initiatives with a value driving business case over initiatives with sponsorship – and learning to lead a hybrid workforce of people and agents. For employees, it means literacy beyond prompt training: using AI in real work, trusting it enough to rely on it yet checking it, and redesigning their own processes with it.

The aim is coherence. The initiatives you prioritize should match your organization’s maturity, so that initiatives and workforce evolve in parallel. Pilots that fail often do so not because the models are weak, but because they land in an organization that cannot yet absorb them.

No matter how far you've come on your AI journey, a structured approach to AI development helps you get the most from your investment. Here's how we can help you apply an innovation lens to your AI portfolio, wherever you are today.

No matter how far you've come on your AI journey, a structured approach to AI development helps you get the most from your investment. Here's how we can help you apply an innovation lens to your AI portfolio, wherever you are today.