Yea, textures are the biggest thing (unless there's video). But don't underestimate vertices, even when using strips. Unity, i think, just ships textures as BCn, meaning 1MB per 1k texture (would be 3-4MB raw). It's even better for the gpu then raw. Then there's normal maps, etc.

Another thing is lighting data, be it some textures, probes, or whatever. That can also take up plenty of space.