I often go through an innovation rut while at a company. It’s often something like this:

Some product you’re working has the daily stream of bugs, feature requests, and things to build. These items are known and easier to quantify both in cost and benefit.

If you want to do something new, you usually need permission. To get that, you often have to do a few things:

1. Convince yourself this is better than the roadmap
2. Convince others this is better than the roadmap
3. Convince others it is better than all other possible ‘new’ things that have been proposed already.

Even in lean, high trust organizations, this is a lot. Often, I’ll just park it and noodle on it in my own time.