<p>The rapid increase in the construction of commercial complexes in large and medium-sized cities in China is due to their ability to meet various consumer needs. Therefore, the decision of where to locate these complexes has become increasingly important. According to market research, customers tend to follow the partial proportional rule with a threshold when selecting commercial complexes. However, determining the exact threshold or its stochastic distribution for each customer is an impossible task. The utilization of interval data for the expression of customer thresholds represents a relatively reasonable methodology for the resolution of uncertainty. This paper presents a two-level robust optimization model that uses interval data to represent customer threshold uncertainty. The model assists in making location decisions for a new company entering the competitive market for commercial complexes. After converting the lower-level model into a mixed integer linear programming model that is easy to solve, an improved harmony search algorithm is proposed to solve this two-level model. The effectiveness of the improved harmony search algorithm was assessed on small-scale problems. The study investigates the location decision of a new entrant, using the commercial complex market in Hangzhou’s main city as an illustrative case. Finally, sensitivity analysis of key parameters provided some valuable insights.</p>