Summary of the Key Points
This article focuses on the challenges faced by small and medium-sized enterprises (SMEs) in China, particularly those in Zhejiang, in implementing AI technologies. It examines their unique characteristics, the underlying concerns regarding AI adoption, and the specific difficulties they encounter, while proposing a step-by-step approach to overcome these obstacles. The main argument is that AI for SMEs is not merely about adding additional tools; rather, it requires a reengineering of organizational processes and the integration of existing expertise. The key challenges in implementing AI lie in issues such as talent availability, data management, and pacing. To address these, four steps should be taken: strengthening digital infrastructure, identifying business pain points, integrating AI throughout the entire process, and restructuring governance structures.
Why Do SMEs Have Anxiety About AI?
Let's first look at the inherent characteristics of SMEs in Zhejiang. Most of these enterprises started as family-owned workshops, specializing in the production and wholesale of small commodities, which inherently come with a fast-paced approach and a preference for quick results, along with a willingness to experiment with different approaches. Their strengths include well-established supply chains, swift decision-making by owners, and strong adaptability. However, they also have significant weaknesses, such as highly homogeneous products, thin margins, lack of brand influence, and varying levels of internal management.
The anxiety SMEs feel about AI is not due to its ineffectiveness, but rather a misunderstanding of its true nature. Just as the Industrial Revolution required factories to be relocated near water sources to take advantage of hydro-powered spinning machines, implementing AI requires a fundamental reconfiguration of organizational processes. SMEs want to embrace AI but are unsure where to begin and fear that their investments might be in vain or that they could fall behind their competitors.
The Challenge of Integrating AI Talent into the Workplace
What SMEs lack is not AI scientists, but individuals who understand both algorithms and the practical aspects of production. Why are AI professionals from large cities reluctant to work in SME workshops? It's not about the hardship; rather, there are experienced workers who have mastered all the hidden nuances of production over the years and are resistant to change.
When AI experts do join, they face a dilemma: if they follow the old ways, new AI methods cannot be effectively applied; if they try to impose their approaches, they may not be listened to. The core issue is the conflict between outdated and modern practices. Enterprises have not yet learned how to combine AI technology with traditional expertise. Simply providing employees with tools like ChatGPT does not equate to true AI integration; a synergistic approach is necessary.
Four Major Barriers to AI Implementation
1. Rapid Homogenization: In small commodity markets in Zhejiang, if one business succeeds, others quickly copy its strategies, often faster than AI can evolve. This leads to a lack of competitive advantages and steadily declining profits, leaving no funds for AI investment.
2. Mismatch Between IT Project Timelines and Business Needs: IT companies typically plan AI projects on a yearly basis, expecting results after several rounds of development. However, SME owners are concerned with immediate needs, such as scheduling production for upcoming orders. The time and financial pressures mean they end up using only basic AI features, which do not address their real problems.
3. Lack of a Digital Foundation: Critical data is scattered across WeChat groups, isolated Excel files, and the memories of experienced workers, making it impossible to train effective AI models. Without a solid digital foundation, even the best tools are useless.
4. Incompatible Tools and Business Processes: Many enterprises purchase AI tools but use them only for basic tasks like timekeeping and approval, failing to address crucial issues such as cost estimation, inventory management, and root causes of quality issues. These tools do not touch on the core business operations, rendering their investment ineffective.
Four Steps to Overcome These Barriers
1. Strengthen Digital Infrastructure: Start by organizing detailed data such as product samples, defective product images, and customer feedback. Don't let this information remain in unstructured formats like WeChat groups or individual memories.
2. Target Pain Points: Focus on areas where immediate benefits can be achieved. For example, textile companies can use AI to predict the design preferences of customers in the Middle East and produce in advance, avoiding delays caused by waiting for trade shows.
3. Integrate AI Throughout the Process: Apply AI across all business stages, from production to order fulfillment and procurement. This way, competitors cannot simply copy your data and processes; you will have a comprehensive dataset that gives you a competitive advantage.
4. Restructure Roles and Responsibilities: Owners should step back from daily routine tasks and focus on managing exceptions, while standardizing knowledge and processes. Clearly define who is responsible for making AI-related decisions and who will handle issues when they arise. This ensures a clear division of labor between humans and AI.
In conclusion, the application of AI must be tailored to each industry and region. The key is to use AI to deliver unique value to customers—perhaps not by lowering prices, but through innovative design or exceptional services. SMEs should start with their actual problems, build a solid foundation, and gradually develop competitive advantages.