Artificial Intelligence (AI) Maturity in Small and Medium-Sized Enterprises: A Framework of Internalized and Ecosystem-Embedded Capabilities

This study proposes a novel, context-sensitive AI maturity framework specifically designed for small and medium-sized enterprises (SMEs) that reconceptualizes maturity as a multidimensional, non-linear, and ecosystem-embedded capability comprising eight dimensions, five levels, and four development pathways to better reflect the unique resource constraints and organizational realities of SMEs.

The Gist

Imagine you are trying to teach a small family bakery how to use a high-tech, AI-powered oven.

Most existing "maturity models" (the rulebooks for how to get good at AI) are written for giant industrial bakeries. They assume you have a massive budget, a team of 20 engineers, a formal board of directors, and a warehouse full of data. They say, "To be 'Level 5' (the best), you must build your own custom oven from scratch, hire a PhD chef, and write a 100-page safety manual."

If a small bakery tries to follow those rules, they will fail. They don't have the money, the staff, or the time. They might give up, thinking they aren't "mature" enough to use AI.

This paper argues that the rulebook is wrong for small businesses. It proposes a new way to look at AI maturity that fits the reality of Small and Medium-sized Enterprises (SMEs).

Here is the breakdown of their new framework using simple analogies:

1. The Core Idea: It's Not About the "Oven," It's About the "Kitchen"

The authors say AI isn't just a piece of technology you buy; it's a skill you build.

• Old View: "How expensive is your oven?" (Technology-focused)
• New View: "How well does your whole kitchen work together?" (Capability-focused)

For a small business, you don't need to build your own oven. You might just need to know how to use a smart, pre-made oven from a vendor, trust the instructions, and bake better bread. The paper calls this "Ecosystem-Embedded." It means your AI maturity depends not just on what you own, but on who you work with (vendors, cloud platforms, consultants).

2. The 8 Ingredients of Success (The Dimensions)

Instead of just looking at "Technology," the paper says you need to balance 8 different ingredients to make a good AI "dish." For a small business, these look different than for a giant corporation:

1. The Boss's Mindset: In big companies, a committee decides. In small ones, the owner decides. If the owner is curious and willing to try, that's a huge advantage.
2. The Team's Skills: You don't need a team of AI scientists. You need a generalist who can learn on the job and ask the right questions.
3. The Data (The Flour): You don't need a warehouse of data. You just need clean data for the specific problem you are trying to solve.
4. The Tools: Using the right tool for the job, not the most expensive one.
5. Fitting into the Routine: Does the AI actually help you bake bread faster, or does it just sit there? It needs to be part of the daily workflow.
6. Right-Sized Solutions: Don't buy a Ferrari if you need a bicycle. Use simple AI that solves your specific problem.
7. Learning: Did it work? If not, tweak it. Small businesses are great at this because they can change their minds quickly.
8. Safety: You don't need a 50-page legal contract to be safe. You just need to be honest about risks and treat customers fairly.

3. The Journey is Not a Ladder (Non-Linear Progression)

Traditional models say you must climb a ladder: Step 1, then Step 2, then Step 3.
The paper says AI maturity is more like a garden.

• You might have a beautiful flower garden (great at marketing AI) but a messy vegetable patch (bad at data).
• You might skip a step, go back, or jump ahead.
• You don't have to be "perfect" at everything to be successful. You just need to be good at the things that matter for your business.

4. The Four Types of AI Businesses (The Archetypes)

The paper identifies four different "personalities" of small businesses using AI. None of them are "wrong"; they are just different strategies:

• The Curious Explorer: "I'm just playing around." They try AI tools because the boss is interested, but there's no big plan. It's messy, but it's a start.
• The Broad Implementer: "I bought a subscription to everything." They use many AI tools from vendors (like a pre-made meal kit). They get things done, but they don't really understand how the kitchen works inside.
• The Focused Specialist: "I'm going to master one thing." They ignore everything else and use AI to become the absolute best at one specific task (e.g., predicting exactly how much flour to order). This is often the sweet spot for small businesses.
• The Advanced Leader: "We are the masters." They have a perfect mix of internal skills and outside help. They use AI to run the whole show. (Very rare for small businesses, but possible).

5. Why This Matters

This framework is a relief for small business owners.

• It stops the shame: You don't have to feel "immature" because you don't have a giant data center.
• It offers a map: It tells you, "Hey, you can be successful by being a 'Focused Specialist' who relies on good vendors, rather than trying to be a 'Giant Corporation' on a budget."
• It helps policymakers: Governments and consultants can stop trying to force small businesses to act like big ones. Instead, they can build better support systems (ecosystems) that help small businesses access the tools they need.

In a nutshell:
The paper says, "Stop trying to build a skyscraper when you live in a cottage. You don't need a skyscraper to be successful. You just need a cottage that is warm, safe, and perfectly suited for the people living inside it."

Technical Summary

Based on the provided paper, here is a detailed technical summary of the study "Artificial Intelligence (AI) Maturity in Small and Medium-Sized Enterprises: A Framework of Internalized and Ecosystem-Embedded Capabilities."

1. Problem Statement

Despite the proliferation of AI maturity models, existing frameworks suffer from significant limitations when applied to Small and Medium-sized Enterprises (SMEs):

• Enterprise-Centric Bias: Current models assume formalized governance, specialized AI teams, and abundant resources, which do not reflect the reality of SMEs (lean structures, informal governance, owner-manager dominance).
• Linear Progression Assumption: Most models depict AI adoption as a sequential, linear journey through fixed stages. This fails to capture the non-linear, experimental, and often uneven trajectories of SMEs.
• Internalization Focus: Existing frameworks treat AI capabilities as purely internal assets, ignoring the heavy reliance SMEs have on external ecosystems (vendors, cloud platforms, consultants) for AI access and deployment.
• Contextual Misalignment: Applying enterprise-scale benchmarks to SMEs often leads to misleading assessments, penalizing firms that effectively use "right-sized" AI solutions despite resource constraints.

2. Methodology

The study employs a conceptual research design grounded in a systematic synthesis of academic literature and practitioner-oriented models.

• Theoretical Foundations: The framework integrates Organizational Capability Theory, specifically the Resource-Based View (RBV) and Dynamic Capabilities Theory, with Maturity Model research.
  • RBV: Views AI not just as technology but as a bundled capability requiring complementary resources (data, skills, processes).
  • Dynamic Capabilities: Emphasizes the ability to sense, seize, and reconfigure resources in changing environments.
• Analytical Approach: The authors reconceptualize AI maturity by shifting the unit of analysis from "technological sophistication" to "configurations of organizational and ecosystem capabilities."
• Framework Construction: The study constructs a new model comprising eight capability dimensions, five maturity levels, and four archetypal development pathways, specifically tailored to the structural and cognitive realities of SMEs.

3. Key Contributions

The primary theoretical contribution is the reconceptualization of AI maturity as a multidimensional, non-linear, and ecosystem-embedded capability rather than a linear accumulation of internal technology.

A. Eight Core AI Maturity Dimensions
The framework defines maturity through eight interrelated domains, adapted for SME constraints:

1. Strategic Orientation & Leadership: Focus on owner-manager cognition and emergent strategies rather than formal roadmaps.
2. Human Capital & Competencies: Emphasis on generalist skills, learning-by-doing, and external expertise over dedicated AI teams.
3. Data Resources & Tech Foundations: Pragmatic mobilization of existing data assets rather than enterprise-scale architectures.
4. AI Application Scope & Business Embedding: Prioritization of high-impact, relevant use cases over portfolio breadth.
5. Process Integration & Operational Alignment: Systematic incorporation of AI into routine workflows.
6. Technical Sophistication & Solution Appropriateness: Deployment of "right-sized" solutions that match organizational capacity.
7. Performance Evaluation & Learning: Lightweight, outcome-oriented feedback loops.
8. Risk Governance & Responsible AI: Principle-based, pragmatic risk management rather than codified oversight bodies.

B. Five Maturity Levels
The framework outlines a developmental progression (though non-linear):

1. Discovery
2. Experimentation
3. Implementation
4. Deployment
5. Optimization

C. Four Archetypal Development Pathways
To explain heterogeneity and "equifinality" (different paths leading to similar value), the study identifies four SME archetypes:

1. Emerging Explorers: Fragmented, ad-hoc experimentation driven by owner interest; limited internal capability.
2. Broad Implementers: Widespread use of standardized/vendor solutions across activities; weak internal capability development and coordination.
3. Focused Specialists: Deep integration in specific strategic domains; combines selective internal learning with targeted ecosystem support.
4. Advanced Leaders: Internally coherent and externally coordinated configurations; sustained integration into strategy and operations.

4. Results and Framework Characteristics

The proposed framework yields several key structural insights:

• Ecosystem Embeddedness: AI maturity in SMEs is explicitly defined by the interaction between internal capabilities and external ecosystem resources (vendors, platforms). Maturity is not just about what a firm owns, but how it orchestrates external assets.
• Non-Linearity: Progression is not uniform. SMEs may advance in specific dimensions (e.g., data) while stagnating in others (e.g., governance), or they may transition laterally between archetypes without moving up a "maturity ladder."
• Contextual Fit: High maturity is redefined as the ability to deploy "fit-for-purpose" AI that aligns with resource constraints and owner-manager cognition, rather than achieving maximum technological sophistication.
• Governance Adaptation: Governance maturity in SMEs is characterized by informal, principle-based controls and owner-manager oversight, contrasting with the formalized, committee-based governance of large enterprises.

5. Significance and Implications

Theoretical Significance:

• Bridging the Gap: The study resolves the misalignment between dominant AI maturity theory (enterprise-focused) and SME realities.
• Configuration Theory: It introduces a configuration-based view of maturity, explaining how different combinations of internal and external resources can yield comparable AI value (equifinality).
• Dynamic Perspective: It shifts the focus from static technological benchmarks to dynamic capability development and learning.

Practical Implications:

• For SME Managers: Provides a diagnostic tool to assess AI readiness without benchmarking against unrealistic enterprise standards. It helps identify specific capability gaps and choose appropriate archetypal pathways.
• For Policymakers: Highlights the need to support shared infrastructures, advisory services, and ecosystem orchestration rather than just funding firm-level technology adoption.
• For Vendors/Consultants: Encourages the tailoring of AI solutions to specific SME maturity profiles, avoiding over-engineered implementations that exceed organizational capacity.

Future Research:
The study establishes a foundation for empirical validation, comparative research on AI competitiveness in SMEs, and further investigation into how different archetypes influence long-term adaptability and performance.