Determining how much to spend on R&D and product development each year is an issue that has plagued management for decades. It is a difficult question.
One of the big challenges has been tying R&D spending to results. Projects are funded, development occurs, products are launched, and commercialization ensues. Years pass before the data is in place to tie spending to results. Companies have tens to hundreds of projects going on at the same time. The relationship between spending and results is unknown, so many executives modulate R&D spending to mitigate variations elsewhere in the business without knowing the effect on R&D over time.
Since the 1970s, academics and practitioners have experimented with various formulas and approaches to tie spending to results. In small sample sizes, such as a single product, the numbers can be assembled and crunched. But even that is not definitive for many. Is success measured by revenues? Profits? Units sold? Market share? Technology leverage? Customer satisfaction? Something else? There will likely never be a perfect solution, but we need something better than what we currently have.