When to Push Out New Product Features

Continuing my post from yesterday that Usage is Like Oxygen for Ideas in Products I want to talk a bit more about the endless debate on when to push out new product features. Product, in this context, is Software-as-a-Service (SaaS) applications on the web. Generally, with installed software, new features need to be tested 10x more thoroughly before shipping as there is a much higher cost to make a mistake. As for SaaS products, here are some reasons to push out features fast and frequently:

• No feature survives intact with contact in the real world
• Customers appreciate innovation, even if it isn’t always perfect
• Customers provide more input when they see a product is actively enhanced vs one that doesn’t change often
• Programmers have a tendency to overcomplicate features, so pushing out bite-sized chunks forces a simplification of the functionality

Now, how is fast and frequently defined? There’s a movement to do continuous deployment where every piece of code checked in goes straight to production if all the tests pass (see IMVU pushing out code 50 times a day). We don’t do continuous deployment but we do look to push code anywhere from daily to bi-weekly depending on the maturity of the product and the impact of the feature. My recommendation is to err slightly on the side of pushing out too fast knowing the trade-off that some features will not be enough for users, but will provide a strong foundation for feedback.

What else? What are some other considerations in pushing out new product features?