Identifying industry structure and industrial relevance of commercial historic cultural districts using functional co-location networks

Old city centers are more than just collections of old buildings; they are living neighborhoods where history, daily life, and commerce intersect. For decades, urban planners and preservationists have struggled with a difficult question: how do you keep these historic places alive without turning them into generic shopping malls that lose their soul? The answer often lies in the mix of businesses allowed to operate there. If a historic street is filled only with souvenir shops, it feels hollow. If it is filled only with offices, it feels dead after hours. The challenge is finding the right combination of restaurants, hotels, shops, and cultural spaces that works together naturally. Until recently, decisions about what businesses to allow in these districts were often made based on gut feeling or by copying what worked in other famous places. This approach frequently led to crowded, repetitive streets where the unique character of the neighborhood faded away.

A new study by researchers at Tianjin University and Beijing University of Civil Engineering and Architecture offers a different path. Instead of relying on intuition, they used a massive amount of digital data to map out exactly how businesses in China's historic districts actually arrange themselves. By analyzing the locations of nearly 80,000 specific businesses across 50 different historic districts in 29 cities, the team discovered that these areas naturally fall into seven distinct patterns. They also found that the businesses within these patterns are not just randomly scattered; they cluster together in specific ways because of four underlying reasons that drive them to be neighbors. This research moves the conversation from guessing what might work to understanding the invisible rules that already govern these vibrant places, offering a clear guide for how to manage them in the future.

The researchers began by gathering data from a major Chinese mapping service, which provided a detailed list of every business, park, and landmark in the 50 selected historic districts. They looked at everything from the number of restaurants to the density of hotels, creating a digital snapshot of each neighborhood. Using a method called hierarchical clustering, which groups similar things together based on their features, they sorted the 50 districts into seven clear categories. The first group, which they called "balanced development," includes famous spots like Beijing's Nanluogu Lane. These areas have a healthy mix of shopping, dining, culture, and lodging, with no single type of business dominating. The second type is "shopping and leisure dominant," seen in places like Guangzhou's Beijing Road, where retail stores make up the vast majority of activity. The third type is the "overnight accommodation ancient town," such as Lijiang and Pingyao, where hotels and guesthouses are far more common than in other districts, supporting visitors who stay for days rather than just a few hours.

The study also identified districts driven by specific local flavors. The "specialty catering dominant" type, found in places like Chengdu's Kuanzhai Alley, is defined by a high concentration of restaurants serving local cuisine, which draws the crowds. Another category, the "scenic attractions pulling" type, relies heavily on natural landscapes and historic sites to bring people in, with commerce serving the tourists rather than leading the way. Two other types serve different primary functions: the "business and living services" districts, like Tianjin's Five Avenues, where offices and daily life services coexist with tourism, and the "administrative and cultural services" type, such as Qingdao's Badaguan, which is dominated by government buildings, schools, and museums rather than shops. The researchers found that each of these seven types has a unique "fingerprint" of business proportions, proving that there is no single perfect model for a historic district; the right mix depends entirely on the specific character and resources of the place.

To understand why these businesses group together the way they do, the researchers built what they call a functional co-location network. Imagine looking at a map of a neighborhood and drawing a line between any two types of businesses that frequently appear in the same small block. By connecting all these lines, they could see which businesses naturally form tight-knit groups. They found eight distinct clusters that appeared across the different district types. For example, in the balanced districts, shopping, dining, and cultural venues formed a dense, interconnected web where each supported the others. In the ancient towns, the hotel cluster was tightly linked with sightseeing spots, creating a loop where tourists could easily move from a museum to their hotel. In the business-focused districts, office buildings were closely tied to local services like laundries and cafes, reflecting the daily rhythm of workers rather than tourists.

The study went a step further to explain the logic behind these connections, identifying four main reasons why certain businesses end up next to each other. The first is "consumption behavior relevance," which simply means that people tend to do certain things in a specific order. Visitors often walk, then eat, then get entertained, so shops, restaurants, and theaters naturally cluster together to follow this flow. The second reason is "environmental quality relevance." High-end cafes, art galleries, and boutique shops tend to group together in areas with beautiful historic architecture or nice parks, while businesses that might be noisy or messy are pushed away. The third driver is "planning and administrative relevance," where government rules and the presence of public institutions shape what businesses can open nearby. Finally, there is "production chain relevance," where businesses that rely on each other for their work locate close together, such as travel agencies sitting near hotels and ticket offices.

These findings suggest that the most successful historic districts are not the ones with the most businesses, but the ones where the businesses understand their role in the local ecosystem. The researchers argue that trying to force a "one-size-fits-all" mix of shops and restaurants into every historic street is a mistake. Instead, planners should first identify which of the seven types their district naturally fits into and then support the specific clusters that belong to that type. For instance, a district that functions as a food destination should focus on extending the dining experience with more leisure options, while an ancient town should ensure its hotels and sightseeing spots work in perfect sync. By using data to see these natural patterns, city managers can make decisions that respect the unique identity of each place, ensuring that these historic neighborhoods remain vibrant and authentic rather than becoming uniform copies of one another.