Just adding features is a deadend. Look for things Trello wouldn't want to do. For example, features specific to a subset of customers. Larger companies are not able to focus on specific smaller segments because people in charge won't have the attention span to handle that many, but you can survive with just one. Another example is addressing some scenarios with fewer / simpler features. Established products cannot remove features, hence even simple uses cases end up over-complicated over time.