📈 economics

When Does AI Adoption Enhance SME Innovation Performance? The Roles of Absorptive Capacity, Entrepreneurial Agility, and Digital Maturity

This study demonstrates that in SMEs, AI adoption enhances innovation performance not directly, but indirectly through a serial mediation pathway where adoption builds absorptive capacity to foster entrepreneurial agility, a process that is significantly strengthened by the firm's digital maturity.

Original authors: Ali Zokaei Kuhbanani, Léo-Paul Dana, Mohammad Hadi Bolandian, Zahra Najafi Kalyani

Published 2026-07-02

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Ali Zokaei Kuhbanani, Léo-Paul Dana, Mohammad Hadi Bolandian, Zahra Najafi Kalyani

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The Big Question: Does Buying a Ferrari Make You a Better Driver?

Imagine you are a small business owner (an SME). You hear that Artificial Intelligence (AI) is the new "Ferrari" of business tools. It can predict trends, automate boring tasks, and help you make decisions. So, you buy the Ferrari (you adopt AI).

The paper asks a simple question: Does just buying the Ferrari automatically make you a faster, better driver who wins more races (innovation)?

The short answer from the study: No. Just having the car doesn't make you win. You need to know how to drive it, you need a good map, and you need to be ready to react to the road.

The Three Key Ingredients

The researchers found that for AI to actually help a small business innovate (create new products, services, or better ways of working), three specific things must happen in a specific order. Think of it like a relay race or a cooking recipe.

1. The Fuel: Absorptive Capacity (The "Learning Engine")

What it is: This is the company's ability to learn. It's not just about seeing new information; it's about understanding it, mixing it with what you already know, and figuring out how to use it.
The Analogy: Imagine AI gives you a library of 1,000 new books. Absorptive Capacity is your ability to actually read those books, understand the stories, and remember the lessons. If you buy the books but can't read them or don't understand the language, the books are useless.
The Finding: The study shows that using AI heavily helps build this "learning engine." The more you use AI, the better you get at learning from it.

2. The Driver's Reflexes: Entrepreneurial Agility (The "Quick Reaction")

What it is: This is the ability to act fast. It's about spotting an opportunity and changing your plan immediately without getting stuck in red tape.
The Analogy: Once you've read the books (Absorptive Capacity), you now know a shortcut exists. Entrepreneurial Agility is the driver who sees that shortcut and immediately turns the wheel to take it, rather than sticking to the old route. It's the ability to pivot quickly.
The Finding: The study found that good "learning engines" (Absorptive Capacity) lead to better "reflexes" (Agility). If you understand the data, you can move faster.

3. The Finish Line: Innovation Performance (The "Race Win")

What it is: This is the result: new products, better services, or a more successful business model.
The Analogy: This is crossing the finish line first.
The Finding: The study found that Agility is what actually gets you to the finish line. You can have the car (AI) and the books (Learning), but if you don't turn the wheel and drive fast (Agility), you won't win.

The Surprising Twist: The Direct Link is Broken

Here is the most important part of the paper: Buying the Ferrari (AI) does not directly make you win the race.

The study found that if you just look at "AI use" and "Innovation," there is no direct connection.

Why? Because AI is just a tool. It produces data and insights. If you don't have the Learning Engine to understand that data, and the Quick Reflexes to act on it, the AI sits there doing nothing.
The Path: AI $\rightarrow$ Learning $\rightarrow$ Quick Action $\rightarrow$ Innovation.
The Lesson: You cannot skip the middle steps. You can't just buy AI and expect new products to appear.

The Secret Weapon: Digital Maturity (The "Garage Condition")

The study also looked at a factor called Digital Maturity.

What it is: This is how "ready" the company is for technology. Do they have good internet? Do employees know how to use computers? Is the boss supportive? Is the data clean?
The Analogy: Imagine you have a Ferrari (AI) and a great driver (Agility). But your Garage (Digital Maturity) is a muddy, broken-down shed with no fuel pumps. Even with a great car and driver, you can't get the car out of the mud.
The Finding: If a company has high Digital Maturity (a clean, well-equipped garage), the link between using AI and learning from it becomes much stronger.
- High Maturity: AI use $\rightarrow$ Super fast learning.
- Low Maturity: AI use $\rightarrow$ Slow, confused learning.

Summary for the General Audience

Think of AI adoption like buying a high-tech kitchen appliance.

Buying the appliance (AI Adoption) doesn't automatically make you a great chef.
You need Absorptive Capacity: The ability to read the recipe, understand the ingredients, and learn how the machine works.
You need Entrepreneurial Agility: The ability to quickly taste the food, adjust the spices, and serve the dish before it gets cold.
Digital Maturity is your kitchen setup. If your kitchen is messy, has no electricity, and you don't know how to use the stove, the fancy appliance won't help you cook a great meal. But if your kitchen is well-equipped and organized, that appliance helps you cook amazing food very fast.

The Bottom Line: Small businesses shouldn't just buy AI hoping for magic. They need to build their ability to learn from it, train their teams to act fast on that knowledge, and make sure their digital "kitchen" is ready for the tools. Only then does AI turn into real innovation.

Technical Summary: When Does AI Adoption Enhance SME Innovation Performance?

Problem Statement
While Artificial Intelligence (AI) is increasingly recognized as a driver of innovation, a significant "adoption–value gap" persists, particularly among Small and Medium-sized Enterprises (SMEs). Many firms adopt AI tools without successfully converting them into tangible innovation outcomes. Existing literature often treats AI adoption as a direct antecedent of innovation or firm performance, failing to adequately explain the organizational mechanisms that translate technical deployment into value. Furthermore, the specific context of SMEs—characterized by resource constraints, less formalized innovation processes, and reliance on managerial cognition—remains underdeveloped in AI research. This study addresses the critical question: When and through what capability pathways does AI adoption intensity enhance SME innovation performance?

Theoretical Framework
The study integrates Dynamic Capabilities Theory (Teece et al., 1997; Teece, 2007) and Absorptive Capacity Theory (Cohen & Levinthal, 1990; Zahra & George, 2002). It proposes a sequential capability-building model where:

AI Adoption Intensity acts as a technological input that strengthens Absorptive Capacity (the ability to identify, acquire, assimilate, transform, and exploit knowledge).
Enhanced Absorptive Capacity fosters Entrepreneurial Agility (the ability to rapidly recognize opportunities, experiment, and reconfigure resources).
Entrepreneurial Agility translates these capabilities into SME Innovation Performance.
Digital Maturity is posited as a boundary condition that strengthens the relationship between AI adoption and absorptive capacity, thereby influencing the downstream indirect pathway.

Methodology

Design: Quantitative, explanatory, cross-sectional survey design.
Context: SMEs operating in Iran, an emerging economy characterized by resource and institutional constraints.
Sample: 294 valid responses from SMEs that had adopted or experimented with AI tools. Respondents included owners, CEOs, innovation managers, and IT managers.
Data Collection: A purposive sampling strategy was used to ensure respondents possessed direct knowledge of AI usage and innovation activities.
Measurement: Five reflective constructs were measured using adapted scales (7-point Likert): AI Adoption Intensity, Absorptive Capacity, Entrepreneurial Agility, SME Innovation Performance, and Digital Maturity.
Analysis: Partial Least Squares Structural Equation Modeling (PLS-SEM) was employed using SmartPLS. The analysis followed a two-step approach: assessment of the measurement model (reliability, validity) and the structural model (path coefficients, mediation, moderation).
Validation: The study utilized bootstrapping (5,000 subsamples) for significance testing and PLSpredict to assess out-of-sample predictive relevance. Common method bias was addressed through procedural remedies and collinearity diagnostics.

Key Results
The empirical analysis yielded the following findings:

Indirect-Only Mediation: The direct effect of AI Adoption Intensity on SME Innovation Performance was not statistically significant ( $\beta = 0.079, p = 0.142$ ). However, the total effect and the serial indirect effect were significant. This indicates an indirect-only mediation mechanism.
The Capability Chain:
- AI Adoption Intensity positively affects Absorptive Capacity ( $\beta = 0.391, p < 0.001$ ).
- Absorptive Capacity strongly enhances Entrepreneurial Agility ( $\beta = 0.555, p < 0.001$ ).
- Entrepreneurial Agility significantly improves SME Innovation Performance ( $\beta = 0.484, p < 0.001$ ).
Serial Mediation: The serial path (AI Adoption $\rightarrow$ Absorptive Capacity $\rightarrow$ Entrepreneurial Agility $\rightarrow$ Innovation Performance) was significant ( $\beta = 0.105, p < 0.001$ ), confirming the proposed sequential mechanism.
Moderating Role of Digital Maturity: Digital maturity positively moderates the relationship between AI Adoption Intensity and Absorptive Capacity ( $\beta = 0.128, p = 0.008$ ). The interaction plot shows that the positive effect of AI adoption on absorptive capacity is significantly stronger for firms with high digital maturity.
Moderated Serial Mediation: Digital maturity strengthens the entire indirect pathway, meaning the capability-building process is more effective in digitally mature SMEs.
Predictive Relevance: PLSpredict results indicated satisfactory indicator-level predictive relevance for SME Innovation Performance, with the PLS-SEM model outperforming a linear benchmark in terms of Root Mean Square Error (RMSE).

Key Contributions
The paper claims four primary theoretical contributions:

Shift in AI-Innovation Logic: It moves beyond the direct-effect perspective, demonstrating that AI adoption is not a standalone predictor of innovation. Instead, it functions as an input into a capability-building process.
Extension of Absorptive Capacity Theory: It positions absorptive capacity as the critical "knowledge-processing bridge" between AI adoption and entrepreneurial action, showing that AI strengthens the firm's ability to process external knowledge.
Entrepreneurial Agility as a Mechanism: It identifies entrepreneurial agility as the action-oriented mechanism through which absorbed knowledge is converted into innovation performance, linking AI research with entrepreneurship theory.
Digital Maturity as a Boundary Condition: It theorizes and empirically supports digital maturity not merely as a background factor, but as a critical condition that determines whether AI adoption can be converted into absorptive capacity and, subsequently, innovation.

Significance and Claims
The study concludes that the innovation value of AI in SMEs depends less on the act of adoption itself and more on the organizational capabilities that allow firms to learn from AI and act on AI-enabled insights. The authors assert that AI does not innovate on behalf of SMEs; rather, SMEs innovate with AI only when they possess the necessary digital maturity, absorptive capacity, and entrepreneurial agility.

The findings suggest that for AI to enhance innovation, SMEs must focus on building complementary capabilities—specifically, the ability to interpret AI-generated data and the agility to act on it—rather than viewing technology acquisition as a sufficient condition for success. The study emphasizes that in resource-constrained emerging economies, these capability conditions are even more critical for bridging the adoption-value gap.