New importers increase their conditional survival rate and import share over time. However, a model of multi-input firms with an import entry cost and stochastic import costs cannot replicate these dynamics. I show that an extended model can be reconciled with the data. I calibrate both models and use them to identify the effects of trade shocks. The simulations show that a decrease in import prices with the new importer dynamics generates lower productivity gains, but these gains are more widespread across firms.