Over the past 23+ years, I have read, learnt and experienced various strategies for how to invest in product development. While I can't/won't take credit for most of these concepts, I have sculpted my own consolidated view of these approaches - one that I believe is solid and sustainable.
I call it the 3 dimensions of product investment: (1) time horizon, (2) risk, and (3) life-cycle.
First dimension - consider dividing your investments into buckets along time horizons of how long before they materialize. E.g., short-term, medium-term and long-term. In the 2nd dimension, you do the same, but look at the risk/probability of the project. For example, you may have little-risk, bigger-risk/reward, and moon-shots. Finally in the 3rd dimension, split your investment based on the life-cycle of your portfolio. Some resources into your near-end-of-life products, some into sustaining your current/installed-base products, and some into new products/technologies.
I don't think there is a right or wrong distribution of these dimensions, and how these work for your specific case may vary based on your company, strategies, risk-approach, etc. But I think the fundamental concept remain the same. Look at your investments (annually or multi-year) and deliberately put them into the different buckets along the 3 dimensions. Then, as a business, make clear decision on what the relative proportion of these buckets ought to be. E.g., 5% of our investments go into moon-shot ideas, or 20% will be in installed-base products, etc. Even if there is no clear-cut criteria, just the visibility in this way would likely add value in your funding approach. Perhaps it will lead to rich dialogs and debates on what the appropriate split is. Over time, you will end up with a distribution that works well for you.