Purpose – Entrepreneurs operate in conditions of dynamic uncertainty; identifying and exploiting opportunities presented by the business environment. Opportunistic search is core to entrepreneurial activity, but its dynamics are rarely explored. Groups of entrepreneurs are attracted to the same potential business opportunities. They have no incentive to cooperate, they may not even know of the existence of others. However, over time, clusters of entrepreneurs interested in the same opportunities develop. Aims to discuss the issues.

Design/methodology/approach – Ant colony optimisation modelling is used to simulate the activities of entrepreneurs in an opportunity rich environment. The entrepreneurs must identify the locations of the appropriate resources. Three simulations were run to observe entrepreneurial success in different environments.

Findings – A random search of the business environment for resources by individual entrepreneurs was unproductive. Once the entrepreneurs learned to read the business environment and so refine their search, they were increasingly efficient. This was even more pronounced when time allowed for search was constrained and weaker entrepreneurs had little influence.

Research limitations/implications – The computer simulations demonstrate how a cluster of entrepreneurial activity may begin. The results raise questions about the appropriateness of policies supporting entrepreneurial activity and about the path dependency of cluster development. Empirical research is now needed to test these research implications.

Originality/value – Focusing on the little explored dynamics of opportunistic search by would-be entrepreneurs in a spatially defined business environment combines previous research in the fields of entrepreneurial outcomes and cluster development. Using a multi-agent search model to simulate the dynamic interaction of a number of entrepreneurs in the same business environment demonstrates early cluster formation without the protagonists relying on cooperative, competitive or value chain interaction.