It depends. I found myself in that situation and just started doing the job, but this requires your boss' approval, since it will take time away from coding and cost money. To do this, you have to have a current gap in product management (as opposed to PMs making bad decisions). This was my path into product management.
If you can't get your boss' approval or there is existing product management, you can try sitting down informally with them. Explain your concerns, but be tactful (this part is hard for many engineers who are completely data driven). You have to be careful not to get people defensive. Go in with an attitude of "I see this as suboptimal, teach me why I'm wrong" instead of "You're an idiot, you should be doing this thing instead". Make sure you are being objective, though. I've had engineers fight me on what "should" be done because they were only looking at technology instead of the whole business.
If neither of those work, you are probably going to get frustrated quickly. Few things did more damage to my team's morale than feeling like they were working on something useless.
If you can't get your boss' approval or there is existing product management, you can try sitting down informally with them. Explain your concerns, but be tactful (this part is hard for many engineers who are completely data driven). You have to be careful not to get people defensive. Go in with an attitude of "I see this as suboptimal, teach me why I'm wrong" instead of "You're an idiot, you should be doing this thing instead". Make sure you are being objective, though. I've had engineers fight me on what "should" be done because they were only looking at technology instead of the whole business.
If neither of those work, you are probably going to get frustrated quickly. Few things did more damage to my team's morale than feeling like they were working on something useless.