Understanding the AI Stages of Adoption: A framework for business leaders

In June of 2024, I introduced the concept of the AI Stages of Adoption (AISA), a framework for understanding where organisations are in their AI journey. Since then, I’ve had countless conversations with business leaders about how this framework helps them navigate their transformation. Today, I want to share a deeper perspective on AISA and how you can use it to accelerate your organisation’s AI adoption.

The Origins of AISA

The framework was born from a realisation that our traditional ways of thinking about technology adoption simply don’t work for AI. Back in 2015, as we helped enterprises move to the cloud, we used the Cloud Stages of Adoption (SOFA) framework. It showed a linear journey through the adoption stages:

1. Project - Organisations start with individual cloud projects, typically focused on new applications or specific workloads. Teams gain hands-on experience with cloud services through proof-of-concepts and discrete initiatives. This stage is characterised by learning, experimentation, and building initial cloud competencies.
2. Foundation - Organisations establish their cloud foundation by implementing governance frameworks, security controls, and networking architecture. They develop standardised processes, create landing zones, and build Cloud Centres of Excellence (CCoE). This stage focuses on creating scalable, secure infrastructure for broader cloud adoption.
3. Migration - Organisations move existing applications and workloads to the cloud at scale. This involves portfolio analysis, application assessment, and systematic migration using strategies like re-hosting, re-platforming, or refactoring. Teams develop migration factories and standardised processes for efficient transitions.
4. Reinvention - Organisations leverage cloud-native capabilities to transform their business. They modernise applications, adopt serverless architectures, and implement advanced services like AI/ML. This stage focuses on innovation, business agility, and creating new revenue streams through cloud-enabled capabilities.

As customers moved through the stages (a linear journey), they also started to build out their Cloud Centres of Excellence (CoEs). The thinking was that by identifying which stage the customer was in, we could deploy the correct cloud adoption strategy and build a plan which removed friction for our customers.

SOFA also put the ‘digital native’ customers alongside enterprise customers, showing that those customers who had used cloud from the outset had a natural advantage in terms of their ability to execute technology innovation at speed.

Unlike cloud adoption, which typically follows a linear path driven by IT, AI initiatives emerge simultaneously across different parts of the business. There isn’t a single ‘migration’ or ‘modernisation’ journey. Instead, we see multiple adoption paths happening concurrently, each with its own pace and objectives.

This distinction is crucial. While cloud adoption was primarily an IT-driven journey, AI adoption is happening across the entire business simultaneously, often driven by business units seeking immediate value. Different AI initiatives can deliver vastly different returns on investment - some might be quick wins with immediate impact, while others might require significant investment before showing value.

There’s also the perspective that everyone is an AI native, some first movers exist, but in the main, most organisations have found themselves launched into AI adoption by simply being here when AI ‘happened’.

Understanding the Stages

Let’s break down each stage of AI adoption and what it means for your organisation:

Experimenting Stage: The Dawn of AI Adoption

This is where most organisations begin their journey. It’s characterised by exploration and testing, often without formal oversight or governance. This is where we frequently see ‘shadow AI’ emerging, with business units independently experimenting with tools like ChatGPT or Claude to solve immediate business challenges.

The hallmarks of this stage are ad-hoc initiatives driven by individual departments, no formal AI strategy or governance, and limited budget allocation. You’ll see proof-of-concept projects and heavy reliance on third-party solutions and consumer AI tools.

The challenge here isn’t starting experiments - it’s managing them effectively. Without proper governance, these experiments can create both opportunities and risks.

Adopting Stage: From Experiments to Enterprise

As pilot projects demonstrate clear value and executive sponsorship grows, organisations begin to move from isolated experiments to more structured adoption. This transition is often triggered by successful proofs of concept, increasing demand from multiple business units, and critically, the realisation that uncontrolled AI adoption poses business risks that need to be managed.

In this stage, organisations begin to formalise their AI initiatives. You’ll see initial governance frameworks emerging, dedicated budgets appearing, and multiple departments showing interest. The formation of AI Centres of Excellence often begins here, along with the development of initial AI policies.

Optimising Stage: Refining and Scaling

Once AI initiatives are formally established and showing consistent results, organisations naturally progress to optimisation. This shift happens when the focus moves from ‘getting it working’ to ‘making it better’ - typically marked by increasing maturity and sophistication of governance structures and growing internal expertise.

This stage is characterised by established AI governance, clear ROI measurements, regular training programmes, and integration with existing systems. You’ll see documented AI strategies, regular AI steering committee meetings, and standardised development practices.

Transforming Stage: Reshaping the Business

The journey from optimisation to transformation occurs when AI stops being just a tool and starts reshaping how the business operates. This transformation is characterised by AI becoming integral to core business processes and new AI-enabled products or services emerging.

In this stage, you’ll see an AI-first mindset emerging, business processes being redesigned, and new AI-enabled products and services being developed. The focus shifts from operational efficiency to strategic advantage.

Scaling Stage: Enterprise and Ecosystem Impact

In this final stage, the organisation has fully embraced AI and is focused on scaling its adoption not just across the internal business, but throughout its entire ecosystem. This broader vision of scale recognises that in today’s interconnected business environment, the greatest value often comes from extending AI capabilities beyond organisational boundaries.

Within the enterprise, scaling involves expanding infrastructure, increasing user adoption, and leveraging AI to drive growth and competitive advantage. You’ll see enterprise-wide AI adoption, mature governance frameworks, and an innovation culture. AI drives significant revenue, operations are automated, and the organisation has developed advanced AI capabilities.

But true scaling extends far beyond internal operations. Forward-thinking organisations are creating AI-powered networks with their suppliers, partners, and customers. These might take the form of shared data ecosystems that improve supply chain visibility, collaborative AI models that enhance industry-wide practices, or open-source initiatives that accelerate innovation across sectors. Some organisations are establishing data partnerships with academic institutions or industry consortia, pooling resources to tackle larger challenges than any single entity could address alone.

The ecosystem approach to scaling brings new opportunities and challenges. Organisations must balance the benefits of shared innovation with data privacy and competitive concerns. They need governance frameworks that work across organisational boundaries and technical infrastructures that support secure collaboration. However, those who successfully navigate these challenges often find they can achieve impacts far greater than through internal scaling alone.

At this stage, organisations often become active contributors to the broader AI community, sharing learnings, contributing to open-source projects, and participating in standard-setting initiatives. This engagement helps shape the evolution of AI technology and practices while ensuring access to cutting-edge developments and talent.

The Reality of Multiple Journeys

Here’s the crucial insight: organisations rarely progress through these stages in a linear fashion. Instead, different parts of your business will likely be at different stages simultaneously. Your marketing team might be in the transforming stage with AI-driven content creation, while your operations team is experimenting with predictive maintenance.

This multi-speed reality is both a challenge and an opportunity. It allows different parts of your organisation to progress at their own pace, learning from each other’s experiences. However, it also requires sophisticated governance and coordination to ensure these parallel journeys align with your overall strategic objectives.

The Dynamic Technology Landscape

The multi-speed nature of AI adoption is further complicated by the rapidly evolving technology landscape. Unlike traditional enterprise technology that follows predictable upgrade cycles, AI capabilities are advancing at an unprecedented pace. New foundation models, tools, and platforms emerge almost daily, each bringing new possibilities and challenges.

This rapid evolution means that technology decisions made during the experimenting stage might need revisiting by the time an organisation reaches adoption or optimisation. A foundation model that was state-of-the-art six months ago might now be superseded by more capable alternatives. Development platforms and tools are constantly expanding their capabilities, sometimes rendering custom-built solutions unnecessary.

Successful organisations approach this dynamism as an opportunity rather than a challenge. They establish regular review cycles to reassess their technology stack, vendor relationships, and skill requirements. This isn’t just about keeping up with the latest trends – it’s about understanding how new capabilities might unlock additional value or accelerate existing initiatives. The goal is to maintain enough flexibility to take advantage of new developments while ensuring stable, sustainable operations.

The impact extends beyond technology choices to talent and skills development. As AI capabilities evolve, so do the skills needed to leverage them effectively. Organisations need to continuously update their training programmes and recruitment strategies to ensure their teams can work effectively with the latest tools and techniques. This might mean shifting focus from deep expertise in specific models to building adaptable teams who can quickly learn and apply new capabilities.

This dynamic environment reinforces the value of the multi-speed approach. While some parts of the organisation can experiment with cutting-edge capabilities, others can focus on stabilising and optimising their use of proven technologies. The key is maintaining awareness of the evolving landscape while making deliberate, strategic choices about when and where to adopt new capabilities.

Investment and Value Realisation

To visualise the AI Stages of Adoption, we plot them on an investment-value realisation graph. I deliberately chose ‘investment’ over ’time’ on the x-axis, as it better reflects the multiple dimensions of what organisations need to commit to AI initiatives. Investment in this context means:

• Financial investment - Direct costs like compute resources, software licences, and vendor services
• People investment - Training existing staff, hiring AI specialists, building data science teams
• Data investment - Cleaning and preparing data, creating labelled datasets, establishing data governance
• Process investment - Redesigning workflows, establishing new governance frameworks, change management
• Time investment - Both calendar time and person-hours across the organisation

What makes AI adoption fascinating is that some initiatives could be quick wins requiring minimal investment but delivering outsized value - think of using foundation models for content generation or customer service automation. Others might need substantial investment before realising value - like building custom models for complex manufacturing processes.

Measuring Success Across Horizons

Success in AI adoption isn’t measured in a single dimension or timeframe. Instead, organisations need to think about value creation across multiple horizons, each revealing different aspects of their AI journey’s impact. These horizons also align closely with an organisation’s progression through the adoption stages.

The immediate horizon, spanning the first six months, focuses on operational impact and maps closely to the Experimenting and early Adopting stages. At this point, success manifests in the volume and effectiveness of pilot projects - tracking metrics like the number of active AI experiments, user engagement with AI tools, and time saved through AI-assisted processes. Early wins are crucial for building momentum and justifying continued investment. Organisations typically see improvements in operational efficiency, with teams handling routine tasks more quickly. A customer service team might measure reduced response times, while a marketing team tracks content production velocity, and operations teams monitor early warning detection rates.

As organisations move into the strategic horizon, spanning from six to eighteen months, the metrics shift to reflect progress through the Adopting and Optimising stages. Here, measurement focuses on adoption breadth and depth: the percentage of departments actively using AI, the number of production AI models deployed, and the maturity of governance processes. This is when AI starts to influence key business metrics in meaningful ways. Beyond tracking customer satisfaction scores and revenue impact, organisations should monitor cross-functional collaboration metrics like the number of departments sharing AI resources or data. The formation and effectiveness of AI Centres of Excellence become key indicators, measured through the number of successfully scaled pilots and standardised practices adopted across teams.

The transformational horizon, extending beyond eighteen months, aligns with the Transforming and Scaling stages, where AI fundamentally reshapes the business. Success metrics here focus on enterprise-wide transformation: the percentage of core business processes enhanced by AI, the number of new AI-enabled products or services launched, and the revenue generated directly from AI initiatives. Innovation metrics become crucial - tracking not just the number of AI use cases in production, but their complexity and business impact. Organisations should measure how deeply AI is embedded in decision-making processes, perhaps through the percentage of strategic decisions supported by AI insights. Employee capability metrics also matter at this stage, such as measuring the proportion of staff who are AI-proficient and the effectiveness of AI upskilling programmes.

Throughout all horizons, organisations must balance quantitative metrics with qualitative indicators of transformation. While it’s important to track hard metrics like cost savings and revenue impact, equally valuable are signs of cultural change: increased AI experimentation across teams, growing confidence in AI-driven decision-making, and the emergence of AI champions throughout the business. The most successful organisations create feedback loops between these metrics, using insights from immediate measures to refine their approach to longer-term transformation.

Building Your AI Strategy

So how do you use this framework to accelerate your AI adoption? Here are the key steps:

1. Map Your Current State - Start by understanding where different parts of your organisation are in their AI journey. Look for patterns across units, identify governance gaps, and spot potential quick wins.

2. Plan Your Engagement Strategy - Different stages need different approaches. A business unit in the experimenting stage needs help establishing guardrails and proving value. One that’s in the adopting stage needs guidance on governance and scaling.

3. Enable Multi-Speed Progress - Support parallel initiatives while maintaining appropriate governance and control. Some departments might need to move quickly while others proceed more cautiously.

Managing Organisational Change

The journey through the AI stages of adoption is, at its heart, a story of organisational change. Success requires carefully orchestrating transformation across multiple dimensions of the business, each requiring its own approach but all working in harmony toward the same goal.

Awareness and education form the foundation of this transformation, but their nature evolves as organisations progress through the stages. Early on, the focus is on building basic AI literacy, helping teams understand what’s possible and what’s practical. As organisations mature, education becomes more specialised and strategic. Teams need to understand not just how to use AI tools, but how to reimagine their work in an AI-enhanced world. Eventually, this evolves into enterprise-wide knowledge management, where learning becomes a continuous, self-reinforcing process.

Stakeholder engagement emerges as a critical success factor, but it’s more nuanced than simply getting buy-in. Executive sponsors play a crucial role, not just in providing resources but in setting the tone for how AI will be embraced. Champions within different departments become the front-line advocates for change, showing their peers what’s possible through practical examples. Cross-functional groups and communities of practice emerge organically, creating networks of knowledge and support that accelerate adoption across the business.

The human aspect of AI adoption cannot be overlooked. Resistance to change is natural and often rooted in legitimate concerns about job security, autonomy, and the changing nature of work. Successful organisations address these concerns head-on, not by dismissing them but by showing how AI augments human capabilities rather than replacing them. The focus shifts from automation to augmentation, from replacement to enhancement. Teams begin to see AI as a powerful tool that helps them work more effectively rather than a threat to their roles.

Core Competency Development

The journey through the AI stages of adoption requires developing new competencies across the organisation. These skills don’t develop in isolation but rather form an interconnected web of capabilities that enable successful AI transformation.

Business leaders need to develop a sophisticated understanding of AI’s strategic implications. This goes beyond simply knowing what AI can do – it requires the ability to identify valuable opportunities, assess risks, and make informed investment decisions. For managers, the challenge is more practical: how to lead teams through AI transformation, measure success, and manage change effectively. They become the bridge between strategic vision and practical implementation.

Technical capabilities evolve along their own track but must stay connected to business needs. Some team members need to develop deep expertise in AI and machine learning, becoming the organisation’s subject matter experts. Others need enough technical understanding to effectively build and implement AI solutions. The key is building a complementary set of business and technical capabilities that allows the organisation to both innovate and execute.

Operational skills form the third critical component. Teams working with AI need to develop new ways of thinking about their work. This isn’t just about learning to use new tools – it’s about understanding how to adapt processes, interpret results, and make decisions in an AI-enhanced environment. Support teams need to develop new capabilities in governance, quality assurance, and operational monitoring. Together, these skills enable the organisation to operate effectively in an AI-transformed environment.

The art of successful AI transformation lies in developing these capabilities in parallel, ensuring they complement and reinforce each other. Organisations need technical experts who understand business implications, business leaders who grasp technical possibilities, and operational teams who can bridge both worlds. This holistic approach to skill development creates the foundation for sustainable AI transformation.

The Path Forward

As you navigate your AI adoption journey, remember these key principles:

1. Start with Business Value - Every AI initiative should tie back to clear business objectives. Whether it’s improving customer experience, reducing costs, or creating new revenue streams, the business value should drive your strategy.

2. Build for Scale - Even when starting small, design your governance and infrastructure with scale in mind. This prevents having to rebuild foundations later.

3. Embrace Multiple Speeds - Accept that different parts of your organisation will move at different speeds. Focus on coordination and knowledge sharing rather than forcing everyone to move at the same pace.

4. Invest in People - Technology is only part of the equation. Invest in building AI literacy across your organisation and developing the skills needed at each stage.

The AI revolution isn’t coming - it’s already here. The question isn’t whether to adopt AI, but how to do it effectively. Understanding where you are in your journey, and where you want to go, is the first step to successful transformation.

By using the AI Stages of Adoption framework, you can create a structured approach to AI adoption that balances innovation with governance, quick wins with long-term transformation, and technical capability with business value.

The future belongs to organisations that can navigate this complex landscape effectively. What stage of your AI journey are you in?

About the Author

Mario Thomas is a transformational business leader with nearly three decades of experience driving operational excellence and revenue growth across global enterprises. As Head of Global Training and Press Spokesperson at Amazon Web Services (AWS), he leads worldwide enablement delivery and operations for one of technology's largest sales forces during a pivotal era of AI innovation. A Chartered Director and Fellow of the Institute of Directors, and an alumnus of the London School of Economics, Mario partners with Boards and C-suite leaders to deliver measurable business outcomes through strategic transformation. His frameworks and methodologies have generated over two-billion dollars in enterprise value through the effective adoption of AI, data, and cloud technologies.