Because R&D is expensed immediately, a company investing heavily looks less profitable today than one harvesting old products, even if the investor is building more value. This is one reason pe-ratio comparisons across companies at different investment stages mislead.
R&D intensity, meaning R&D divided by revenue, varies enormously: low single digits for industrials, twenty percent or more for software and biotech. Compare within an industry, not across.
Example: Northwind Tools spends $90M of R&D on $840M of revenue, or 10.7%. If it froze R&D at the prior year's $72M, reported operating-income would be $138M instead of $120M, a 15% flattering effect with no change in demand.
Related: operating-expenses, capitalised-software