A way of visualizing your r&d investments is with 3 buckets. Having the right mix is critical. They work together. A healthy investment distribution is 50% platform, 40% features, and 10% experiments. You can skew for short periods, but over the long term its fatal.

Elsewhere I’ve heard it recommended to spend up to 50% on platform (maybe Platform engs cost more $$, so this doesn’t map perfectly to the model in the above article) https://buildrightside.com/book/chapter2-platform-investments.html Regardless of which model you go with, I think it’s an interesting exercise to capture the CURRENT platform/“feature” ratio and ask questions from there. (Are there signs of underutilized platform tools? Signs of over-worked feature engineers? Etc)