An iteration is a single pass through the Build-Measure-Learn cycle: build something, test it, learn, adjust. The term describes the basic unit of agile work. Instead of completing a product in one big push, it is incrementally improved through many short loops. The speed of iterations determines how fast a team learns.
In practice, this looks like: version 1 of an MVP is released and tested. Results show that users cannot find a specific function. In the next iteration, navigation is adjusted and tested again. Version 3 shows significantly better numbers. Or the data show that the entire approach is not working — then a pivot follows. What matters is that every iteration has a concrete learning objective and does not merely deliver incremental improvements to the existing product.
The iteration principle comes from agile software development and Lean methodology. The most common mistake: running iterations without a clear hypothesis. The result is constant change without systematic learning.