I had never considered that death could or would be off-limits in D&D but I think that I will change my approach somewhat. As others have said, I would gauge the players' opinions in a Session 0 to the idea of having no death until the average party level is 5 and Revivify becomes available. Until then, PCs that 'die' are saved from actual death but will lose an ability score permanently depending on how they 'died'. Through bad luck - player chooses; through stupidity - wisdom or intelligence; big mouth brought on an unnecessary fight - charisma; hit by massive damage - strength, dex or con. That sort of thing, so that death has lasting consequences but doesn't require a new character. After Level 5, the consequences are more narrative-based - something happens to the PC in the nexus between life and death that affects them when they are brought back/revived, malevolent interest from the planes etc. My basic premise is that 'death' should have lasting consequences but not necessarily a new character.
In my home campaign raising someone from the dead is not as simple as casting a spell. Revivify pretty much works as written (but still has a chance of attracting unwanted attention) because the soul takes a minute to start crossing over. However, after that minutes souls go to the Shadowfell/Nifleheim where they meet a spirit guide that takes them to their final reward.
In order to raise dead someone has to actually travel to Nifleheim and find the spirit and convince it to return, the raise dead portal just opens the door. Sometimes people bring back a stowaway or the soul of the person that died becomes tied to Nifleheim and constantly attracts attention. Even the people that go into Nifleheim can have negative consequences.
Once people cross over to their final destination, it's nearly impossible to retrieve them because I loosely base my world on Norse mythology where even gods can die. Souls that refuse to move on (or manage to evade their guide) eventually become ghosts, specters or other undead.