In cities, economic activity is spatially concentrated, with firms producing similar goods often clustering together. Much of the urban and spatial economics literature attributes this to lower production costs from shared knowledge, suppliers, or workers. However, in low-income countries, firms are typically small, integrating production and sales in a single location, with face-to-face interactions dominating transactions. For these firms, access to customers becomes a critical driver of performance.

This study combines a theoretical model and original data from 600 garment firms and their customers in Kampala, Uganda, to quantify the role of an additional agglomeration force: consumer information frictions. When consumers have limited information about the variety of goods sold in the market, they are forced to visit firms in person to learn about product characteristics and availability—particularly in settings where both firms and customers lack digital access. These costly visits lead consumers to favor dense firm clusters that allow them to reduce the cost of gathering information about products. This creates a location trade-off for firms: dense areas attract more customers due to easier search, but also mean more competition. If the first effect dominates, firms face demand-side incentives to agglomerate.

The data reveals two key consumer behaviors consistent with this mechanism. First, customers incur roughly three times higher transport costs when traveling to the city’s dense core compared to the periphery. Yet once in the core, they visit 31% more firms, indicating lower within-location search costs. Second, consumers searching in the core find higher quality products and form longer relationships, suggesting that more intensive search leads to better matches.

I develop and estimate a model of consumer search and firm location that matches these patterns, to assess how much information frictions contribute to urban firm clustering. This is important for two reasons. First, understanding consumer search in low-income cities sheds light on demand-side constraints that limit firm growth. Second, with urban populations projected to grow 75% by 2050, understanding location drivers is key for evaluating policies shaping city development.

Two key findings emerge. First, consumer information frictions account for 21% of the observed firm agglomeration within the city—on par with traditional production-side agglomeration forces. Second, the high search costs that customers incur to obtain product information suppress demand. Low demand, combined with the high commuting costs firm owners sustain to operate in denser locations, significantly reduces firm profitability.

I use the model to evaluate two policies being considered in Kampala to decongest the city center: (i) the introduction of an e-commerce platform, and (ii) caps on the number of firms in the core, enforced through evictions. The distinction is key: the platform targets the root cause—information frictions—while the caps treat the symptom—firm density—without addressing the cause. The analysis reveals that both policies reduce agglomeration, but evictions disproportionately harm high-quality firms. This occurs because the increased spatial dispersion reduces the visibility of high-quality firms, preventing them from exploiting their competitive advantage in denser areas.

The next step in this research agenda is to understand why, despite the benefits of proximity, small businesses in Uganda continue to operate independently rather than merging into larger entities.